journal of management journal/uose v_11_book_final.pdf · issn 1391-8230 volume xi no. 1 october...

ISSN 1391-8230

October 2014Volume XI No. 1

JOURNAL OF MANAGEMENT

Published by the Faculty of Management & Commerce

South Eastern University of Sri Lanka

Oluvil # 32360

Sri Lanka

The Journal of the Faculty of Management and Commerce


Journal of Management

EDITORIAL BOARD MEMBERS – JOURNAL OF MANAGEMENT

Editor in Chief: Dr. MIM. Hilal, Senior Lecturer, SEUSL

Associate Editors: Mrs. Sulaiha Beevi Athambawa, Senior Lecturer, SEUSL

Mr. KM. Mubarak, Senior Lecturer, SEUSL

Managing Editor: Ms. MMM. Mashroofa, Senior Assistant Librarian, SEUSL

Editorial Advisory Dr. SMM. Ismail, Snr. Lecturer, SEUSL

Board: Prof. Mahalliyawarachchi, Sabragamuwa University of Sri Lanka.

Prof. Sampath Amaratunge, University of Sri Jayawardenepura.

Dr. (Mrs.) FHA. Rauf, Snr. Lecturer, SEUSL

Dr. A. Jahufer, Snr. Lecturer, SEUSL

Dr. S. Gunapalan, Snr. Lecturer, SEUSL

Mrs. Sareena UMA Gaffoor, Snr. Lecturer, SEUSL

Ms. MAC. Sulaiha Beevi Athambawa, Snr. Lecturer, SEUSL

Mrs. A. Inun Jariya, Snr. Lecturer, SEUSL

Mr. ALMA. Shameem, Snr. Lecturer, SEUSL

Publication: The Journal of Management is published two times in a year in April

and October by the Faculty of Management & Commerce (FMC),

South Eastern University of Sri Lanka (SEUSL).

Subscription: Foreign: US $ 12 (SAARC Countries)

US $ 25 (Other Countries)

Local: Rs.500.00 inclusive of postage

Subscription will be accepted as standing orders. Single issues also will be provided on request

for Rs. 700.00 per copy. All orders must be accompanied with pre-payment in favour of the

“South Eastern University of Sri Lanka”. Sales request should be addressed to Assistant

Registrar, FMC, and SEUSL.

Editorial Policies: Manuscript of research papers, reviews, and short communications in the field

of management sciences should be accepted for publication on the understanding that they have

not been published or submitted simultaneously for publication elsewhere. Manuscripts should

be clearly typed in single-spaced; length may be usually limited to 20 pages in A4 size paper.

Every manuscript must include a 100 – 150 words long abstract. By submitting an article, the

author transfers the copyright to Editorial Board of the Journal.

All papers submitted for publication are examined and evaluated and by referees, to assess the

suitability for publication. The Editorial Board reserves the right to edit the materials to confirm

to the style and format of the journal.

The Editorial Board is not responsible for any shortcomings in the form of inaccurate or

misleading data, option or statement appearing in the Journal leading to confusion and criticism.

So, the author is entirely responsible for the contents of the article subject to the above

consideration.

Correspondence: All correspondence regarding the Journal articles should be sent to:

Editor in Chief

Editorial Office

“Journal of Management”

Faculty of Management & Commerce


Oluvil #32360

Sri Lanka.

TP / Fax : 0094-67-2255069

E-mail : [email protected]

Copyright © Faculty of Management and Commerce, South Eastern University of Sri Lanka.

The Copyright covers the exclusive right to reproduce and distribute the article, including

reprints, photographic reproductions, microforms or any other reproductions of similar nature

and translations. No part of the article may be reproduced by the reader without permission in

writing from the copyright holder.

JOURNAL OF MANAGEMENT – Volume XI No. 1 April 2014

JOURNAL OF MANAGEMENT

Volume XI No. 1 April 2014

Contents Page No

Small and Medium Entrepreneurs’ Intension to Use Cloud Computing: Reference to Eastern Province of Sri LankaS. Sabraz Nawaz and M.H. Thowfeek

Brand Switching for Three Sarong Brands in Maruthamunai: A use of theory of stochastic processM.B.M. Ismail

Organizational Justice as an Antecedent of Job Satisfaction of Administrative Staff in Sri Lankan State UniversitiesM.I. Nawfer and F.H. Abdul Rauf

Sociopolitical and Economic Causes for Corruption in Developing Countries: A Cross Country Empirical StudySelvarathinam Santhirasegaram

Relationship of Demographic Factors to Ability Utilization of Job Satisfaction ofGovernment and Private Bank Employees in Ampara Region – Sri Lanka Aboobacker Jahufer and Rifkhan Ahamed

Impact of Post Purchase Experiences on Customer Loyalty: An Empirical InvestigationV.K. Hamza

Modelling Colombo Consumer Price Index: A Vector Autoregressive ApproachM.C. Alibuhtto

Students’ Attitudes Towards Information and Communication Technology (ICT) EducationSubathini Priyadharsan and S. Sivasenthuran

1

11

23

36

58

67

79

87

ABSTRACT

Cloud computing has become one of thekeywords in the industry today. Although beingnot new concept in computing because of theproliferating growth of internet with amazingbandwidth, arrival of mobile devices andrequirements of end-users. This paper focuses onthe small medium entrepreneurs’ intension to usecloud computing services, particularly theEastern province of Sri Lanka. This paperpresents seven factors influencing entrepreneurs’intension. Instruments were developed based onpreviously studies and study used quantitativemethod. Data were collected conveniently.Instrument was tested for reliability,regressionanalysis was performed to see the strengthindependent variables on the dependent and itwas found technological as well asorganizational characteristics significantlyinfluence small medium entrepreneurs’ intensionto use cloud computing services, which isconsistent with many previous studies.

Key words: Cloud computing, intension touse, small medium entrepreneurs, technology,organization.

Introduction

Because of severe market competition and adramatically changing business environment,

firms have still prompted to adopt various high-tech information technologies (IT) to improvetheir business operations (Pan and Jang, 2008;Sultan, 2010). Recently, the term CloudComputing has been critical in the world of IT.The use of internet-based technologies toconduct business is recognized as an importantarea for IT innovation and investment (Armbrustet al., 2010; Goscinski and Brock, 2010). Cloudcomputing has spread out through the main areasrelated to information systems (IS) andtechnologies, such as operating systems,application software, and technological solutionsfor firms (Armbrust et al., 2010). Cloudcomputing is a kind of computing applicationservice that is like e-mail, office software, andenterprise resource planning (ERP) and usesubiquitous resources that can be shared by thebusiness employee or trading partners. Thus, auser on the internet can communicate with manyservers at the same time, and these serversexchange information among themselves (Hayes,2008). Moreover, telecommunication andnetwork technology have been progressing fast;they contain 3G, 4G, etc. so the high speedinfrastructures are integrated strongly. Cloudcomputing services can provide the userseamlessly, the convenience, and the qualitystable technological support that can develop theenormous potential demand (Buyya et al., 2009;Pyke, 2009). Thus, cloud computing provides theopportunity of flexibility and adaptability toattract the market on demand.

1

SMALL AND MEDIUM ENTREPRENEURS’ INTENTION TO

USE CLOUD COMPUTING: REFERENCE TO EASTERN

PROVINCE OF SRI LANKA

S. Sabraz Nawaz

Department of MIT, Faculty of Management and Commerce,South Eastern University of Sri Lanka,

Email: [email protected]

M.H. Thowfeek

Department of Management, Faculty of Management and Commerce,South Eastern University of Sri Lanka,

Email: [email protected]

Literature Review

Cloud Computing

McAfee (2011) observe that cloud computingcomprises three services: Software-as-a-Service(SaaS): Instead of installing software on theclient’s machine and updating it with regularpatches, frequent version upgrades etc.,applications like Word processing, CRM(Customer Relationship Management), ERP(Enterprise Resource Planning) are madeavailable (hosted) over the internet for theconsumption of the end-user. It can achieveeconomies of scale. This is the biggest and mostmature cloud model. Commercial vendors areYahoo Mail, Gmail, Hotmail, TurboTax Online,Facebook, Twitter, Microsoft Office Live,Google Apps, Salesforce.com, Cisco WebEx webconferencing, antivirus, Success Factors (HRMtool) etc., Platform-as-a-Service (PaaS): Insteadof buying the software licenses for platforms likeoperating systems, databases and middleware,these platforms and the software developmentkits (SDKs) and tools (like Java, .NET, Python,Ruby on Rails) are made available over theInternet. Commercial vendors include MicrosoftAzure Services, Amazon Web Services (AWS),Sales-force’s Force.com, Google App Engineplatform, IBM Cloud burst, Amazon’s relationaldatabase services, Rackspace cloud sites,Infrastructure-as-a-Service (IaaS): This refers tothe tangible physical devices (raw computing)like virtual computers, servers, storage devices,network transfer, which are physically located indata centers but they can be accessed and usedover the internet using the login authenticationsystems and passwords from any dumb terminalor device. Commercial vendors include AmazonEC2 (Elastic Compute Cloud), Elastic BlockStorage (EBS) and Simple Storage Service (S3),Rackspace cloud servers, etc.

There are four different cloud deploymentmodels within organizations namely (Marston etal., 2011; Rath, 2012): Public cloud: It isavailable from a third party service provider via

Internet and is very cost effective for universitiesto deploy IT solutions. For example, GoogleApps.

Private cloud: It is managed within anorganization and is suitable for large enterprises.For example, the US government cloud productis in a segregated environment, both physicallyand logically and is being handled by a thirdparty provider, Google. Private clouds providethe advantages of public clouds but still incurcapital expenditures. Community cloud: It isused and controlled by a group of enterprises,which have shared interests. For example, the USfederal government using community cloud forUSA.gov, cars.gov, Apps.gov. Hybrid cloud: It isa combination of public and private cloud.

TOE Framework

Tornatzky and Fleischer (1990) summarizedprevious studies on innovative informationtechnology adoption and proposed theTechnology-Organization-Environment (TOE)framework to understand the critical factorsaffecting the introduction of new informationtechnology. Their framework contains threemajor elements that affect the process ofadopting innovation technology and it includesthe organization dimension, technologydimension and environment dimension(Tornatzky & Fleischer, 1990).

This study focuses on Technology andOrganization factors and proposes a twodimensional model (Fig. 1), which incorporatesthe variables of technology and organizationfactors in understanding the decision to adoptcloud computing by small medium entrepreneursin Eastern province of Sri Lanka.

Research Model and Hypotheses

The research model of this study incorporatestechnologicaland organizational contexts asimportant determinants of cloud computingadoption.

JOURNAL OF MANAGEMENT – Volume XI No.1 - October 2014

2

Small and Medium Entrepreneurs’ Intension to Use Cloud Computing: Reference to Eastern Province of Sri Lanka

Technology dimension

The technology dimension entails the internaland external influences of adopting specificinformation technology in the organization. Dueto the nature of cloud computing technology,data security and privacy are the major concernsfor adoption (Kuo, 2011). As a result,establishing a secure environment for cloudbased enterprise resource planning system dataintegration and sharing is indeed critical (Chen,Lu, & Jan, 2012). For this reason, data securityis included as one of the critical key variables inthe technology dimension.

Rogers (1983) defined relative advantage as thedegree to which a technological factor isperceived as providing greater benefit for firms.It is reasonable that firms take into considerationthe advantages that stem from adoptinginnovations (To and Ngai, 2006). Cloudcomputing services, which allow operations to begeneralized and mobilized through internettransactions, can substitute for or complementERP software. The study of Premkumar andRoberts (1999) indicated that relative advantageswill affect businesses and push them to adoptnew information technologies.

Firms may not have confidence in a cloudcomputing system because it is relatively new tothem (Buyya et al., 2009). It may take users along time to understand and implement the newsystem. Thus, complexity of an innovation can actas a barrier to implementation of new technology;

complexity factor is usually negatively affected(Premkumar et al., 1994). Compatibility refers tothe degree to which innovation fits with thepotential adopter’s existing values, previouspractices and current needs (Rogers, 1983).Compatibility has been considered an essentialfactor for innovation adoption (Cooper and Zmud,1990; Wang et al., 2010). When technology isrecognized as compatible with work applicationsystems, firms are usually likely to consider theadoption of new technology. When technology isviewed as significantly incompatible, majoradjustments in processes that involve considerablelearning are required.

Previous studies have also indicated that IScomplexity (Changet al., 2007) and compatibilitywill affect IT adoption decision positively (Lin etal., 2012; Liu, 2011) for those in the highereducation industry. Therefore, perceived systemcomplexity will be a key criterion when makingan adoption decision. Furthermore, highereducation ERP modules that handle students’enrollment, examination and researches by staff,etc. are rather unique by nature. How to migratethese data with the cloud computing platformwill also be a critical factor these organizationsneed to consider. Consequently, the level ofsystem compatibility is another key factor in thetechnical dimension. If cloud computingtechnology can be compatible with the existingsystems or applications of the university, then itwill be more helpful and also more feasible forthe adoption of cloud computing technology.

3

�Fig.: Research Model

Because of the varied and extensive nature of thecosts, organizations can find the expensesassociated with this type of project to be verysizable. For this reason, costs will also be acritical factor for the adoption decision.

Based on the above discussions, this dimensionis composed of four variables. These variablesare Data Security, Relative Advantage,Complexity and Compatibility.

From the aforementioned theory, the followinghypothesis has been developed:

H1: There will be a positively significantrelationship between Technology Factors and theAdoption Decision of cloud computing.

Organization dimension

Organizational factors will affect the intention toadopt new information systems technology(Chang, Hwang, Yen, &Lian, 2006; Hsiao et al.,2009). In this study, the organizational dimensionrepresents different organizational conditionsincluding variables such as Top Management’sSupport, Adequate Resources, and Benefits foradoption.

Top management’s support refers to whether ornot the executives understand the nature andfunctions of cloud computing technology andtherefore fully support the development of it. Thestudy of Chang et al. (2006) found that topmanagement’s support will affect new ISadoption. Top management support is critical forcreating a supportive climate and for providingadequate resources for the adoption of newtechnologies (Lin and Lee, 2005; Wang et al.,2010). As the complexity and sophistication oftechnologies increase, top management canprovide a vision and commitment to create apositive environment for innovation (Lee andKim, 2007; Pyke, 2009).

If a given organization has a sufficient budget,adequate human resource support, ample time,and good top management’s involvement, thenthe adopting of cloud computing technology will

be met in a positive manner. To this end,adequate resources are also critical to thesuccess of adoption (Chang et al., 2007).

Finally, potential benefits such as improvinganorganization’s image, gaining strategicadvantage over others, improving their servicequality, and enhancing the efficiency of internaloperations will also be critical. The study ofChang et al. (2006) found that the benefits of ISwill lead to a positive adoption. Themeasurements used for this dimension wereadapted from the studies of Premkumar andRoberts (1999), Chang et al. (2007), and Kuanand Chau (2001) respectively.

Based on the above discussions, this dimensionis composed of three variables. These variablesare Top Management’s Support, AdequateResource and Benefits.

From the aforementioned theory, the followinghypotheses have been developed:

H2: There will be a positively significantrelationship between Organizational Factors andthe adoption of cloud computing.

Research Method

The research is a quantitative study based onquestionnaire survey. Quantitative methodenables the researcher to test the relationshipsbetween the variables identified in the model andthereby let him provide evidence to support ordisprove the hypotheses (Carter and Belanger,2005). The population of this study included allsmall and medium sized organizations in theEastern province of Sri Lanka. As the Samplingframe was not available the sampling methodbecame non-probabilistic convenient sampling.According to Hair et al. (1998) as cited byRehmanet al. (2012), “each independent variableis expected to have ten data records”; since thisstudy had seven independent variables, 70respondents would have sufficed. According toSekaran and Bougie (2010), “sample sizes largerthan 30 and less than 500 are appropriate for


4

most research”. Therefore, the size of the samplewas decided to be at least 250 small and mediumentrepreneurs. Although there are many datacollection methods available, because ofconstraints in terms of time, costs and humanresources, the questionnaires were administeredpersonally using drop-off and pick-up approachfor printed questionnaires and published on‘web-forms’.

The data collected were examined for outliers,coded and scored appropriately. The initialanalysis included an examination of descriptivestatistics of demographic variables. Reliabilitytest was conducted to see the consistency of dataand Principal Component Analysis was alsoperformed. The resultant variables of the factoranalysis were used to see the strength ofrelationship as well as strength of explainingability of the variables. These regression testsinvolved calculating and comparing to gaininsight into the nature of the relationshipbetween independent variables and dependentvariable. MS Excel 2010 and SPSS 20 were usedfor analysis.

Data Analysis

Scale Reliability and Factor Analysis

The constructs were tested for reliability bycalculating Crobach’s Alpha. A total of 30 itemswere developed to capture the eight constructsunder investigation. In order to improve thereliability, two items were removed, one in TopManagement’s Support and another in Benefits,thereby making the Alpha value well above 0.7(Hair et al., 1998). Table 1 shows the variables,Alpha values and number of items for eachvariable.

Results and Discussion

A regression analysis was undertaken based onthe research model which included independentvariables and dependent variable. As there weretwo categories of variables namely Technologyand Organization, the regression analysis was runfor each category separately. The analysis wasfacilitated by SPSS software, which investigatedthe relationship of predictors to outcomevariable.

The first analysis was run between Technologyvariables, namely Data Security, RelativeAdvantage, Complexity and Compatibility, andIntension to Use variable. The resultant ANOVA(Table 2), showing Significance of .001 or onechance in 1000 of incorrect rejection of nullhypotheses, confirms that the data betweenTechnology-predictor variables and Intensionoutcome variable are strongly correlated andthere is a good model.


5

Variable Cronbach’s No.

Alpha Items

Data security 0.728 2

Relative advantage 0.872 3

Complexity 0.899 5

Compatibility 0.818 3

Top management’s 0.923 3

support

Adequate resource 0.896 5

Benefits 0.659 4

Intension to Use 0.746 3

The coefficient of determination of thecontribution of Technology variables to Intentionto Use, the R2, value from Table 3 which is0.779 (Adjusted R2 .776) indicates a sharedvariation of about 78% between Technology

variables data and Intention to Use data. That is,approximately 78% of the variances in Intentionto Use can be accounted for by knowledge ofTechnology variables alone.

According to Table 4, the Beta values of DataSecurity (.617), Relative Advantage (.788),Complexity (.756) and Compatibility (.678)show positive correlations. While the constant isinsignificant, all predictor variables showsignificance of .000, indicating a probability ofless than one in 1000 of Type-I error;significantly positive relationships betweenpredictor variables and outcome variable. With a

unit increase in Data Security factor, Intensionfactor will increase by .604 units. The Intensionfactor will increase by .773 units when there isone unit increase in Relative Advantage factor. Aunit increase in Complexity factor andCompatibility factor will cause .740 unit and.665 unit increase in Intension factorrespectively.


6

Table 2: ANOVAa

Model Sum of df Mean Square F Sig.

Squares

1 Regression 191.513 4 47.878 222.436 .000b

Residual 54.242 252 .215

Total 245.755 256

a. Dependent Variable: IntensionToUse, b. Predictors: (Constant), Dat_Security, Re_Advantage, Compatibility, Complexity

Table 3: Model Summary

Model R R Square Adjusted Std. Error of the

R Square Estimate Sig.

1 .883a .779 .776 .536 .000

a. Predictors: (Constant), Dat_Security, Re_Advantage, Compatibility, Complexity

Table 4

Unstandardized Coefficients Standardized Coefficients Sig.

Model B Std. Error Beta

1 (Constant) .059 .034 .088

Dat_Security .604 .048 . 617 .000

Re_Advantage .773 .038 . 788 .000

Complexity -.934 .040 -.954 .000

Compatibility .665 .045 .678 .000

a. Dependent Variable: Intension to Use

The second analysis was run betweenOrganization variables, namely Topmanagement’s support, Adequate resource andBenefits, and Intension to Use variable. Theresultant ANOVA (Table 5), showing

Significance of .001 or one chance in 1000 ofincorrect rejection of null hypotheses, confirmsthat the data between Organization-predictorvariables and Intension outcome variable arestrongly correlated and there is a good model.

The coefficient of determination of thecontribution of Organization variables toIntention to Use, the R2, value from Table 5which is 0.745 (Adjusted R2 .742) indicates ashared variation of about 75% between

Organization variables data and Intention to Usedata. That is, approximately 75% of thevariances in Intention to Use can be accountedfor by knowledge of Technology variables alone.


7

Table 7

Unstandardized Coefficients Standardized Coefficients Sig.

Model B Std. Error Beta

1 (Constant) -.014 .037 .708

Top_Man_Support .783 .037 .799 .000

Resources .729 .041 .744 .000

Benefits .801 .035 .818 .000

a. Dependent Variable: Intension to Use

Table 5: ANOVAa

Model Sum of df Mean Square F Sig.

Squares

1 Regression 183.083 3 61.028 246.361 .000b

Residual 62.672 253 .248

Total 245.755 256a. Dependent Variable: IntensionToUse, b. Predictors: (Constant), Top_Man_Support, Benefits, Resources

According to Table 6, the Beta values of TopManagement’s Support (.799), AdequateResources (.744) and Benefits (.818) showpositive correlations. While the constant isinsignificant, all predictor variables showsignificance of .000, indicating a probability ofless than one in 1000 of Type-I error;significantly positive relationships between

predictor variables and outcome variable. With aunit increase in Top Management’s Supportfactor, Intension factor will increase by .783units. The Intension factor will increase by .729units when there is one unit increase in AdequateResources factor. A unit increase in Benefitsfactor will cause .801 units increase in Intensionfactor.

Table 6: Model Summary

Model R R Square Adjusted Std. Error of the

R Square Estimate Sig.

1 . 863a . 745 . 742 . 498 .000

a. Predictors: (Constant), Top_Man_Support, Benefits, Resources

The nature of cloud computing is very much inclose relationship to the core of the businessprocesses. The purpose of this study is toenhance the understanding of intension to usecloud computing by small medium entrepreneursin the Eastern province of Sri Lanka. We foundthat there are seven factors that drive theintension to use cloud computing. They are DataSecurity, Relative Advantage, Compatibility,Complexity, Top Management Support, AdequateResources and Benefits.

Data Security was seed to have significant andpositive influence on firms’ intension to usecloud computing services; the higher the securitythe more intended the entrepreneurs. Relativeadvantage was observed to have significantlypositive influence on the intension to use cloudcomputing by small medium entrepreneurs in theEastern province. This finding is consistent withprevious studies by Tan et al. (2008) and Wanget al. (2010). The relative advantage of cloudcomputing services usage would improve thespeed of organizational communication,coordinating efficiency among firms,communication with clients as well as access tomarket information, etc. (Armbrust et. al., 2010).Expectedly, complexity and compatibility werefound to be significant discriminators whichmean that small medium entrepreneurs think thatcloud computing usage has technologicalcomplexity and compatibility. If the complexityin usage and charging looks complex thenentrepreneurs would not intend to go for suchsystems and also if the new system would not gohand in hand with existing legacy system thenthey would not risk such new systems, and thisfind is consistent with earlier studies by Wang et.al., (2010) and Oliveria and Martins (2010).

Organizational characteristics would play amajor role in decision making process (Cho,2006) of entrepreneurs. This study have foundthat top management’s support, adequateresources and benefits to significant factors indetermining small medium entrepreneursintension to use cloud computing services.Benefits offer motivation for people to use an

innovating technology if they expect relativeadvantages from a new system to enhance workefficiency Wang et al. (2010). It is clear that iftop management’s support and adequateresources available and potential benefits areunderstood, then intension to use an innovativetechnology would increase.

Conclusions

In order to promote the intension to use andadopt cloud computing it is vital to delineate thefactors that influence this intension. Like the prosof this innovative technology there are cons aswell, which hinders or defames its usefulness.One such feature is the downtime which isdifferent for provider to provider. Thecomplexity and compatibility of implementingcloud computing could be another barrier to theintension to use such technology. Therefore, it isimportant to understand the impact of factorsinfluencing the intension to use cloud computingservices in small medium industries. Smallmedium firms that would like to use cloudcomputing services could start with incrementalimplantation by means of slowly adding thenumber of process by establishing more internetinfrastructures. These firms can start implementcloud services into businesses by beginning withaccounting information systems and customerrelationship management systems, etc. and thesesystems are of high benefits and make the firmsto compete with rivals competently.

This study presents key finding that influence thesmall medium entrepreneurs intension to usecloud computing services. The findings revealthat whether small medium entrepreneurs intendto use cloud computing services depend onfirm’s technological factors; data security,relative advantage, complexity and compatibilityas well as organizational factors; topmanagement’s support, adequate resources andbenefits.

Further researches can include environmentalfactors and the scope can be extended to includewhole country.


8

References

Armbrust, M., Fox, A., Griffith, R., Joseph, A.D.,Katz, R., Konwinski, A., Lee, G., Patterson, D.,Rabkin, A., Stoica, I. and Zaharia, M. (2010),“A view of cloud computing”, Communicationsof the ACM, Vol. 53, pp. 50-8.

Buyya, R., Yeo, C.S., Venugopa, S., Broberg, J. andBrandic, I. (2009), “Cloud computing andemerging it platforms: vision, hype and realityfor delivering computing as the 5th utility”,Future Generation Computer Systems, Vol. 25,pp. 599-616.

Buyya, R., Yeo, C.S., Venugopa, S., Broberg, J. andBrandic, I. (2009), “Cloud computing andemerging it platforms: vision, hype and realityfor delivering computing as the 5th utility”,Future Generation Computer Systems, Vol. 25,pp. 599-616.

Chang, I. C., Hwang, H. G., Hung, M. C., Lin, M.H., & Yen, D. C. (2007), Factors affecting theadoption of electronic signature: Executives’perspective of hospital information department.Decision Support Systems, 44(1), 350–359.

Cooper, R.B. and Zmud, R.W. (1990), “Informationtechnology implementation research: atechnological diffusion approach”,Management Science, Vol. 36, pp. 123-39.

Goscinski, A. and Brock, M. (2010), “Towarddynamic and attribute based publication,discovery and selection for cloud computing”,Future Generation Computer Systems, Vol. 26,pp. 947-70.

Hair, J.F., Anderson, R.E., Tatham, R.L. and Black,W.C. (1998), Multivariate Data Analysis,Prentice-Hall, Upper Saddle River, NJ.

Hayes, B. (2008), Cloud computing [Electronicversion]. Communications of the ACM, 51(7),9-11.

Kuan, K.K.Y. and Chau, P.Y.K. (2001), “Aperception-based model for EDI adoption insmall businesses using a technology-organization-environment framework”,

Information and Management, Vol. 38, No. 8,pp. 507-521.

Lee, S. and Kim, K. (2007), “Factors affecting theimplementation success of internet-basedinformation systems”, Computers in HumanBehavior, Vol. 23, pp. 1853-80.

Liu, I.F., Chen, M.C., Sun, Y.S., Wible, D. andKuo, C.H. (2010), “Extending the TAM modelto explore the factors that affect Intention toUse an Online Learning Community”,Computers & Education, Vol. 54 No. 2, pp. 600-10.

Liu, I.F., Chen, M.C., Sun, Y.S., Wible, D. andKuo, C.H. (2010), “Extending the TAM modelto explore the factors that affect Intention toUse an Online Learning Community”,Computers & Education, Vol. 54 No. 2, pp. 600-10.

Marston, S., Li, Z., Bandyopadhyay, S., Zhang, J.,&Ghalsasi, A. (2011), Cloud computing—Thebusiness perspective. Decision SupportSystems, 51(1), 176–189.

McAfee, A. (2011), What every CEO needs toknow about the cloud. Harvard BusinessReview, 89(11), 124–132.

Oliveira, T. and Martins, M.F. (2010),“Understanding e-business adoption acrossindustries in European countries”, IndustrialManagement & Data Systems, Vol. 110, pp. 1337-54.

Pan, M.J. and Jang, W.Y. (2008), “Determinants ofthe adoption of enterprise resource planningwithin the technology-organization-environment framework: Taiwan’scommunications industry”, Journal ofComputer Information Systems, Vol. 48, pp. 94-102.

Premkumar, G., and Margaret Roberts. “Adoptionof new information technologies in rural smallbusinesses.” Omega 27.4 (1999): 467-484.

Pyke, J. (2009), “Now is the time to take the cloudseriously”, White Paper, available


9

at:www.cordys.com/cordyscms_sites/objects/bb1a0bd7f47b1c91ddf36ba7db88241d/time_to_take_the_cloud_seroiusly_online_1_.pdf(accessed April 2014).

Pyke, J. (2009), “Now is the time to take the cloudseriously”, White Paper, availableat:www.cordys.com/cordyscms_sites/objects/bb1a0bd7f47b1c91ddf36ba7db88241d/time_to_take_the_cloud_seroiusly_online_1_.pdf(accessed April 2014).

Rath, A. (2012), Cloud computing: Facing thereality. Bhubaneswar (India): Batoi.

Rogers, E.M. (1983), Diffusion of Innovation, TheFree Press, New York, NY.

Sultan, N. (2010), “Cloud computing for education:a new dawn?”, International Journal ofInformation Management, Vol. 30, pp. 109-16.

To, M.L. and Ngai, E.W.T. (2006), “Predicting theorganizational adoption of B2C e-commerce:an empirical study”, Industrial Management &Data Systems, Vol. 106, pp. 1133-47.

Tornatzky, L. G., & Fleischer, M. (1990).Theprocess of technological innovation. Fleisher:Lexington Books.

Uma Sekaran and Roger Bougie. (2010), ResearchMethods for Business: A skill buildingapproach, Wiley India.

Wang, Y.M., Wang, Y.S. and Yang, Y.F. (2010),“Understanding the determinants of RFIDadoption in the manufacturing industry”,Technological Forecasting & Social Change,Vol. 77, pp. 803-15.


10

11

ABSTRACT

Brand switching occurs in consumer productssince they can be substituted for each other. Mostof the people in Maruthamunai wear sarongsspecially, three brands such as white brand,normal colour brand and full colour brand.Previous studies witness that switchingbehaviour of consumer products have beenanalysed in different setting. However, studies inSrilankan context are limited to researchers’search. Thus, there is a need to study aboutswitching behaviour in different industries.Objectives of this study are to know brandinsistence for selected three brands; to know theswitching rate i.e. switching behavior forselected three brands and to know the brandloyalty of selected three brands. This study used385 sarong users in Maruthamunai. They wereselected on the basis of random sample selection.Research instrument was a questionnaire. Thisstudy used stockastic matrix, Markov Chains andnotations of Markovian Model. Results revealedthat there is high brand insistency for whitebrand. Moderate brand insistency exists fornormal colour brand. Low brand insistencyexists for full colour brands. Findings revealedthat the highest switching rate would happen inwhite brand. Switching rate would be moderatefor normal colour brand. The lowest switchingrate would happen in fullcolour brands.Further, it was found that the highest brandloyalty exists for white brands than normalcolour brand and full colour brand. It concludedthat of the three brands, white brand has thehighest brand insistence, brand switching and

brand loyalty. Normal colour brand has themoderate in all these three aspects. Full colourhas the lowest in all these three aspects. Thisstudy has the implication for the producers of allthree brands.

Keywords: Brand switching, Maruthamunai,sarong brands, stochastic process.

Introduction

Brand is a name, sign, symbol or design used toidentity a product and distinguish it from anotherproduct, service or business. Brand insistencerefers to tendency to be continued with existingbrand in future. Brand switching is the process ofchoosing to switch from routine use of oneproduct or brand to steady use of a different butsimilar product. Brand switching occurs inconsumer products since they can be substitutedfor each other. Brand loyalty refers toconsumer’s behaviour of repeatedly purchasing aspecific brand over a certain period of time.Sarong usage is common in Ampara Coastal BeltAreas. That customers evaluate brands, developimages of brands with varying degrees of loyaltyis well established (Keller, 1993; Park, Jun &Shocker, 1996; Arvind Nivedita, 2010). Productattributes are the image building features of aproduct which may include packaging, branding,labeling, design, colouring, quality, price,warranty and servicing ( Biodun, 2002).Maruthamunai is one of the villages situated inAmpara Coastal Area that produce and marketsarongs too other parts of the Island. Most of the

BRAND SWITCHING FOR THREE SARONG BRANDS IN

MARUTHAMUNAI: A USE OF THEORY OF STOCHASTIC

PROCESS

M. B. M. Ismail

Senior Lecturer in Management, Department of Management, Faculty of Management and Commerce, South Eastern University of Sri Lanka.

E- mail: [email protected].

people in Maruthamunai wear sarongs specially,three brands such as white brand, normal colourbrand and full colour brand on many occasionssuch as off- office hours, village functions,festivals, weddings, home visits, and so on.However, sarong users switch from one brand toanother time to time. Chaarlas & Rajkumar(2012) indicated that if a consumer’s propensityto switch is known, the market can be modeledto indicate future market share and the relativepositioning of the competing brands.

Statement of the problem

Different studies have been conducted indifferent countries and in different industries.Empirical evidences shows that switchingbehaviour has been modeled in numerousstudies. Awogbemi, Oloda and Osama (2012)studied about modeling brand switching inconsumers’ products in Nigeria. In this study, itwas examined the relevance of product attributesto switching rates with reference to three brandsof soft drinks. Markov chains were employed todetermine the brand loyalty of the consumers ofthe soft drinks and the future market shares inthe long run. Sequel to the balance vectorgenerated, it was discovered that the consumersexhibited the most brand loyalty towards Fanta.Similarly, Joe, Alice and Sternthal (1978) studiedabout brand switching. They used panel data fortwo consumer packaged goods indicates thatmedia-distributed coupons and cents-off dealsinduce brand switching and result in less loyaltywhen retracted than if no deal is offered. Incontrast, package coupons stimulate brandloyalty which is maintained when they areretracted. It was concluded that the extent towhich economic utility theory and self-perception theory follows the brand switching.Previous studies witnessed that switching costresults in sales decline for products. Some otherstudies have proved that brand switching causesvarying competition. Due to these reasons, theremay be chances for sales fluctuation andincreased competition among three brands suchas white brand, normal colour brand and fullcolour brand. These previous studies witness that

switching behaviour of consumer products havebeen analysed in different setting. However,studies in Srilankan context are limited toresearchers’ search. Thus, there is a need to studyabout switching behaviour in different industries.In addition to these empirical findings ofprevious search of literature, evidence fromindustrial experts termed as owners of weavingcentres were also not in same opinion regardingthe switching behaviour of sarong users amongthree brands i.e. white brand, normal colourbrand and full colour brand. Thus, researcherselected weaving products for studying aboutswitching behaviour.

Objectives of this study

Objectives of this study are to:

1. To know brand insistence for selected threebrands

2. To know the switching rate i.e. switchingbehavior for selected three brands

3. To know the brand loyalty of selected threebrands

Significance for the study

Number of studies has been conducted time totime since they play an important role inmarketing for marketer, customer and thecommunity. Customers evaluate brands, developimages of brands with varying degrees of loyaltyis well established (Keller, 1993; Park, Jun &Shocker, 1996; Arvind Nivedita, 2010). Productattributes are the image building features of aproduct which may include packaging, branding,labeling, design, colouring, quality, price,warranty and servicing (Biodun, 2002). Everyconsumer of a product is expected to have utilityfunction for each of these attributes. Utilityfunction enables a consumer of a product tostudy how product satisfaction varies withalternative levels in each of the attributes. A goodbrand induces competitive advantage andprevents competition. For instance, Lake (2014)stated that brand brand is intended to identify thegoods and services of one seller or group ofsellers and to differentiate them from those of


12

other sellers. Therefore it makes sense tounderstand that branding is not about gettingyour target market to choose you over thecompetition, but it is about getting yourprospects to see you as the only one thatprovides a solution to their problem. Therefore,marketers have to prevent customers from brandswitching. Branding has many advantages suchas delivering clear message to customers,confirming marketers’ credibility, connectingmarketers’ target prospects emotionally,motivating the buyer and generates user loyalty.When marketer allows customers to switch fromone brand to another, gaining these advantageswould become questionable. Thus, a marketerhas to be concentrated on brand switching abouttheir brands. Brand switching is a loss to thecompany. To succeed in branding marketer mustunderstand the needs and wants of theircustomers and prospects. Thus, brand switchinganalysis should be conducted time to time for thebetterment of the company. Brand resides withinthe hearts and minds of customers, clients, andprospects. It is the sum total of their experiencesand perceptions. Maruthamunai is one of sarongproducing city in the Island. Sarong productionfulfills the local demand for men who wearsarongs. Thus, when studying about brandswitching behaviour of sarong users consumersmay be sustained for long term for localconsumption of sarongs. Sometimes, brandswitching can lead sarong users to shift from oneto another. Thus, a particular manufacture mayface switching cost in due course. This is whythis study is undertaken to make awareness aboutthe switching behaviour among sarongmanufacturers.

Review of Literatures

Brand switching can be studied using thestochastic theory. Properties of stochastic theoryrelates to probability and the time series. This iswhy a number of researchers have started theirstudies on brand switching using stochastictheory. Research has been conducted from theearliest. Frank (1974) studied about the theory ofstochastic preference and brand switching. Studyprovided strong evidence using brand choice

behavior. Study found that brand choicebehaviour is substantially stochastic. This studytested a general theory of stochastic preference.Study concluded that brand switching data werein substantial agreement with the theory.Similarly, another study was conducted by Givon(1984) who studied about variety seekingthrough brand switching. Research analysed theconcept of variety seeking behavior is modeledas a stochastic brand choice model that yields ameasure of variety seeking for each individualconsumer. Study used panel data for 28 products.Study concluded that each household and thepossibility of market segmentation are consistentwith by variety seeking behavior.

Following these studies, there were few recentstudies in the areas of brand switching. Forexample, Harald, Gupta and Dick (2003) studiedabout sales promotion bump due to brandswitching. Research argued that severalresearchers have decomposed sales promotionelasticities based on household scanner-paneldata. study offer a complementary decompositionmeasure based on unit sales. The measure showsthe ratio of the current cross-brand unit sales lossto the current own-brand unit sales gain duringpromotion. Study reported empirical results forthis measure. Study also derived analyticalexpressions that transform the elasticitydecomposition into a decomposition of unit saleseffects. These expressions show the nature of thedifference between the two decompositions.Fudenberg and Tirole (2000) studied aboutcustomer poaching and brand switching. Studyexamined switching and repeat purchase effectsof advertising in mature, frequently purchasedproduct categories. Study draw on consumerbehavior theories of framing and usagedominance to formulate a logit choice model formeasuring these effects. Study estimate themodel using single-source scanner data. Resultssuggest that advertising induces brand switchingbut does not affect the repeat purchase rates ofconsumers who have just purchased the brand, aresult consistent with usage dominance ratherthan framing. It was found that the switchinginfluence to be largely confined between thecurrent and previous purchase occasions. Study

Brand Switching for Three Sarong Brands in Maruthamunai: A use of theory of stochastic process

13

concluded that the magnitude of this effect andexplore potential profitability. In the first study,scanner-panel data were used. Similarly, in thesecond study, logit choice model were used.These two studies used different methodologicalaspects for their study.

Some other studies have been conducted foranalysing the relationship between two variablesalong with different methodological views. One ofthe study conducted by Deighton, Caroline andScott (1994) studied about the effects of advertisingon brand switching and repeat purchasing. Studydetermined the multiple effects of retail promotionson brand loyal and brand switching segments ofconsumers. Segments were determined by aniterative Bayesian procedure. The variations inwithin-segment brand shares within a store arerelated to promotional variables by a logit modelestimated by nonlinear least squares. Store sharewas modeled as a function of store attractiveness, asummary measure of the store's promotionalactivity on the multiple brands. Finally, categoryvolume is related to overall product categoryattractiveness in a model that includes both currentand lagged effects. The research approach isapplied to the IRI ground coffee data. Result of thefirst finding disclosed that the market can becharacterized by brand loyal segments, each ofwhich buys mostly their favorite brand, andswitching segments, each of which switches mainlyamong different brands of the same type (e.g., drip,percolator). The second finding was thatpromotional variables have significant effects onwithin-segment market shares.

Some other studies have been conducted aboutmodeling Richard (1989) studied about brandswitching model with implications for marketingstrategies. Study collected switching data in theareas of durables and services. A two-class“hard-core loyal” and “potential switcher” latentmodel for the analysis of brand switching data isproposed. Some previously unpublishedautomobile data will be presented and analyzedalong with another data set for frequentlypurchased packaged goods. Study showed howsimple model can be easily estimated using a

standard log-linear modeling approach. Naufel,and Dipak (1991) modelled purchase-timing andbrand-switching behavior incorporatingexplanatory variables and unobservedheterogeneity. This study used a continuous-timesemi-Markov approach to analyze in a singleframework the purchase-timing and brand-switching decisions of households for afrequently purchased product. Study found thatthe probability distribution of interpurchase timesis not the same for various switching betweenbrands, revealing extra information about thepurchase-timing decisions. Further, it was foundthat though the marketing mix and householddemographic variables explain a large part of thevariation in the brand-switching rates, theyaccount for only a small part of the variation inthe repeat purchase rates. Another finding fromthe analysis is that the rates of switching betweenbrands due to promotional activities such asspecial displays and price reductions are inreverse order to the share of purchases of thevarious brands.

Methodology

Sample selection

This study used 385 sarong users inMaruthamunai. They were selected on the basisof random sample selection. Major sarongweaving centres, Textiles, retailer outlets andwholesale outlets were considered for selectingthese samples.

Material

Research instrument was a questionnaire thatwas distributed with final year Undergraduates,Faculty of Management and Commerce, SouthEastern University of Sri Lanka, Oluvil. Items inthe questionnaire was tested using satisfactoryCronbach alpha values.

Method and procedure

This study uses stockastic matrix, MarkovChains and notations of Markovian Model foranalyzing objectives of this study.


14

Stockastic matrix

In mathematics, a stochastic matrix (also termedprobability matrix, transition matrix, substitutionmatrix, or Markov matrix) is a matrix used todescribe the transitions of a Markov chain. Eachof its entries is a nonnegative real numberrepresenting a probability. It has found use inprobability theory, statistics and linear algebra, aswell as computer science and populationgenetics. There are several different definitionsand types of stochastic matrices. They are (1) Aright stochastic matrix is a square matrix ofnonnegative real numbers, with each rowsumming to 1. (2) A left stochastic matrix is asquare matrix of nonnegative real numbers, witheach column summing to 1.

In the same vein, one may define stochasticvector (also called probability vector) as a vectorwhose elements are nonnegative real numberswhich sum to 1. Thus, each row of a rightstochastic matrix (or column of a left stochasticmatrix) is a stochastic vector.

Markov Chains

Markov model is a stochastic process used interms of a random variable indexed with respectto time. Its analysis also takes cognizance of a

sequence of events. The state probabilities at afuture instant given the present state of theprocess do not depend on the states occupied inthe past. The behavior of the system in each statememorizes i.e the future state of the system attn+1depends on its present state at tn (Dilip,Rupam & Anupawa, 2009). Markov chains havebeen used in many applications; see Jarrow,David and Stuart (1997), Zipkin (1993), White(1993), Sandman (2005), Guedon (1993),Glennon, Dennis and Peter (2005) among others.

Notations of Markovian Model

Suppose Xn with n denotes random variable ondiscrete space S. The sequence X = (Xn : n ) iscalled a stochastic process. If P is a probabilitymeasure of X such that P(Xn+1 = j/X0 = i0…,Xn = in) = P(Xn+1 = j/ Xn = in) for all i0,… in,and , then the sequence X is a Markov chain onS. The probability measure P is the distributionof X, and S is the state space of X. If theconditional probability P(Xn+1 = j/Xn = in) areindependent of time index n , then the Markovchain X is homogeneous and denoted by P(Xn+1= j/Xn = i) = Pij for all i,jϵS. Pij describes theprobability of movement from state i to state jduring a specified or discrete time interval. Table3.3.31 tabulates the probability Pij.


15

Table 3.3.31: The probability Pij

From ……To S1 S2 ….. Sn

S1 P11 p12 ….. P1n

S2 P21 P22 ….. P2n

. . . . .

Pij . . . . .

. . . . .

. . . . .

. . . . .

Sn Pn1 Pn2 . Pnn

Where; Σ Pij = 1, Pij ≥ 0 for all i,j and S1 S2, .. . Sn are discrete states. However, if a Markovchain has initial probability vector X0 = (i1, i2, .

. . in) and transition matrix Pij, the probabilityvector after n repetition is X0 . p n

I,j whichdefines the future state probabilities.

Data Analysis, Results and Presentation

Collected data were tabulated in Table 4.1.

From ……To

White Brand

Normal Colour Brand

Full Colour Brand

White Brand

100 (White Brand toWhite Brand)

40 (Normal ColourBrand to WhiteBrand)

10 (Full Colour Brandto White Brand)

Normal Colour Brand

50 (White Brand toNormal Colour Brand)

80 (Normal ColourBrand to NormalColour Brand)

15 (Full Colour Brandto Normal ColourBrand)

Full Colour Brand

35 (White Colour toFull Colour Brand)

15 (Normal ColourBrand to Full ColourBrand)

40 (Full Colour Brandto Full Colour brand)

Table 4.1: Collected data


16

Brands Number of sarong users

White Brand to White Brand 100

White Brand to Normal Colour Brand 50

White Brand to Full Colour Brand 35

Normal Colour Brand to Normal Colour Brand 80

Normal Colour Brand to White Brand 40

Normal Colour Brand to Full Colour Brand 15

Full Colour Brand to Full Colour brand 40

Full Colour Brand to Normal Colour Brand 15

Full Colour Brand to White Brand 10

Total 385

(Source: From Field survey, 2014)

Brand preferences are shown in Table 4.2.

Table 4.2: Brand Preference


17

Brand insistence and switching rate are shown in Table 4.3.

Table 4.3: Brand insistence and switching rate

Transition frequency is tabulated in Table 4.4. Transition matrix is shown in Table 4.5.

Table 4.4: Transition Frequency

White Brand toWhite Brand

White Brand toNormal Colour Brand

White Brand to FullColour Brand

Sub total

Normal Colour Brandto Normal ColourBrand

Normal Colour Brandto White Brand

Normal Colour Brandto Full Colour Brand

Sub total

Full Colour Brand toFull Colour brand

Full Colour Brand toNormal Colour Brand

Full Colour Brand toWhite Brand

Sub total

Total

100 54 0.540541

50 27 0.27027 45

35 18 0.189189

185 1 0.480519

80 59 0.592593

40 29 0.296296 40

15 11 0.111111

135 1 0.350649

40 62 0.615385

15 23 0.230769 38

10 15 0.153846

65 1 0.168831

385 1

From ……To Frequency Percentage Probability for

sub total

Probability for

grand total

White Brand

Normal Colour Brand

Full Colour Brand

Total

100 50 35 185

40 80 15 135

10 15 40 65

385

From ……To White Brand Normal ColourBrand

Full ColourBrand

Frequency

(Source: From Field survey, 2014)


Table 4.5: Transition Matrix

Transition diagramme is shown in Figure 4.1.

White Brand

Normal Colour Brand

Full Colour Brand

0.540541 0.27027 0.189189 1

0.592593 0.296296 0.111111 1

0.615385 0.230769 0.153846 1

From ……To White Brand Normal ColourBrand

Full ColourBrand

Probability

Figure 4.1: Transition diagramme

18


Brand loyalty using forecasting

Brand loyalty is measured using general rules forforecasting for next period with the support offormulas stated in 4.1.1 and 4.1.2.

X(n) = X (n - 1) P…………………Equation (4.1.1)

X(n) = X (0) Pn…………………Equation (4.1.2)

Forecast for the 0th period is predicted byfrequency for transition matrix for each brandsuch as white brand, normal colour brand andfull colour brand divided by total frequency.Table 4.1.1 shows the frequency ad probabilityfor forecasting 0th period.

Table 4.1.1:

The frequency and probability for

forecasting for 0th period

Frequency Probability

185 185 / 385 = 0.480

135 135/ 385 = 0.350

65 65 / 385 = 0.168

385

Forecast for the 1st period is predicted by X(n) =X (0) Pn. i.e. X (0) is 0.480, 0.350 and 0.168. Thesevalues are multiplied by Pn found in transitionmatrix. Table 4.1.2 shows forecasting for 1st

period.

19

White Brand

Normal Colour Brand

Full Colour Brand

0.480, 0.350 &0.168

0.540541

0.592593

0.615385

0.480 * 0.540541= 0.2594

0.480 * 0.592593= 0.2844

0.480 * 0.615385= 0.2953

0.8391

0.27027

0.296296

0.230769

0.350 * 0.27027= 0.0945

0.350 * 0.296296= 0.1037

0.350 * 0.230769= 0.08076

0.2789

0.189189

0.111111

0.153846

0.168 * 0.189189= 0.0317

0.168 * 0.111111= 0.01866

0.168 * 0.153846= 0.0232

0.0736

From ……To X (0) White Brand Normal Colour

Brand

Full Colour

Brand

Table 4.1.2: Forecasting for 1st period

Table 4.1.3 shows the forecasting for 0th and 1st period

Table 4.1.3: Forecasting for 0th and 1st period

0

1

0.480 0.350 0.168

0.8391 0.2789 0.0736

From ……To White Brand Normal Colour Brand

Full Colour Brand

Conclusions

From Table 4.3, brand insistency varies brand tobrand. It could be understood that majority 100users of the sarong users insist on white brand.80 users insist on normal colour brand users.Low number (40) of sarong users insists on fullcolour brand. There is high brand insistency forwhite brand. Moderate brand insistency exists fornormal colour brand. Low brand insistency existsfor full colour brands. In case of brandswitching, 85 white brand sarong users wish toswitch to normal colour brand (50) full colourbrand (35). 55 normal colour brand sarong userswish to switch to white brand (40) and fullcolour brand (15). 25 full colour brand sarongusers wish to shift their brand as normal colourbrand (15) and as white brand (10). The highestswitching rate would happen in white brand.Switching rate would be moderate for normalcolour brand. The lowest switching rate wouldhappen in fullcolour brands. With regard tothe brand loyalty, forecasting values for 0th andthe 1st period, values of white brand, normalcolour brand and full colour brand are 0.480 &0.8391; 0.350 & 0.2789 and 0.168 & 0.0736.Thus, the highest brand loyalty exists for whitebrands than normal colour brand and full colourbrand. In toto, of the three brands, white brandhas the highest brand insistence, brand switchingand brand loyalty. Normal colour brand has themoderate in all these three aspects. Full colourhas the lowest in all these three aspects.

Managerial Implications

White brand has highest brand insistence, brandswitching and brand loyalty among users. Thereis a significant number of switching brand users.Thus, it is not a good sign for producers of whitebrand sarongs. Awareness has to be given toprevent the brand switching by way of newdesigns, new decorations and so on. Normalcolour brand has the moderate in all these threeaspects among users. Brand switching rate isnearing to white brand. Thus, similar efforts haveto be taken by producers of normal colourbrands. Full colour has the lowest in brandinsistence among users. This is danger to theproducers. Thus, full colour is not preferableamong users. Thus, production of full colourshould be limited. Full colour brands cannot beinventoried.

Value addition

This study applies the stochastic probabilitymethod in this study for brand switchingbehaviour in weaving industry in AmparaCoastal Area.

Limitations and further researchavenues

This study is limited among the sarong users ofMaruthamunai. This geographical limitation canbe extended to some other places in futurestudies.


20

References

Arvind, S. and Nivedita, S. (2010), “BrandRelationships and Switching Behaviour forHighly Used Product in Young Consumers,VIKALPA, Vol. 35(1).

Asmussen, S. R. (2003), “Markov Chains”.Applied Probability and Queues. StochasticModelling and Applied Probability, Vol. 51, pp. 3–8. doi:10.1007/0-387-21525-5_1. ISBN978-0-387-00211-8. edit

Awogbemi, C. A., Oloda, F. S. and Osama, C. K.(2012), “Modeling Brand Switching inConsumers’ Products”, Journal of Economicsand Sustainable Development, Vol. 3, No.12,pp. 82-92.

Biodun, T.A. (2002), “A Survey of ProductAttributes on Brand Switching”. Ilorin Journalof Arts and Social Sciences”, Vol. 3(1) pp 138-147.

Chaarlas, I.J and Rajkumar, R. (2012), “BrandSwitching – A conceptual Analysis”,THAVANIJRMM, Vol. 1(2) pp 1-5.

Dilip, R., Rupam, T. and Anupawa, T. (2009), “TheLearning Effects of Brands, Journal ofManagement Research, Vol. 1(2) pp 1-13.

Deighton, J., Caroline, M. H. and Scott, A. N.(1994), “The Effects of Advertising on BrandSwitching and Repeat Purchasing”, Journal ofMarketing Research, Vol. 31, No. 1, pp. 28-43

Frank M. Bass (1974), The Theory of StochasticPreference and Brand Switching, Journal ofMarketing Research, Vol. 11, No. 1, pp. 1-20

Fudenberg, D. and Tirole, J. (2000), “CustomerPoaching and Brand Switching”, The RANDJournal of Economics, Vol. 31, No. 4 pp. 634-657

Givon, M. (1984), Variety Seeking Through BrandSwitching, Journal of Marketing, Vol. 3 Iss. 1,pp. 1-22

Glennon, Dennis and Peter, N. (2005), “Measuringthe Default Risk of Small Business Loans:Survival Approach, Journal of Money, Creditand Banking, 37: 109 - 125

Grover, R. and Srinivasan, V. (1992), “Evaluatingthe Multiple Effects of Retail Promotions onBrand Loyal and Brand Switching Segments”,Journal of Marketing Research, Vol. 29, No. 1pp. 76-89

Guedon, Y. (2003), “Estimating Hidden Semi-Markov Chains from Discrete Sequences,Journal of Computational and GraphicalStatistics, Sept., 03.

Harald, J. He, Gupta, S. Dick, R. W. (2003), “Is 75% of the Sales Promotion Bump Due toBrand Switching? No, Only 33%”, Journal ofMarketing Research, Vol. 40, No. 4, pp. 481-491.

Jarrow, R. David, L. and Stuart, T. (1997), “A Markov Model for the Term Structure ofCredit Risk Spreads”, the Review of FinancialStudies, Vol. 10, pp. 231-239.

Joe A. Dodson, Alice M. Tybout and BrianSternthal (1978), brand switching, Journal ofMarketing Research, Vol. 15, No. 1, pp. 72-81

Kettler, K.J (1993), “Conceptualizing, Measuringand Managing Customer Based Brand Equity”,Journal of Marketing, Vol. 57(1) pp. 11-22.

Matrix Algebra V2.1 (2006), An Utility for FiniteMathematics and Applied Calculus. Accessedfrom http://infstra.ed/stefan-waner/real/world/matrixalgebra/fancymatrixalg2.htm

Lake, L. (2014), “What is branding and howimportant is it to your marketing strategy?”,Accessed from http://www.marketing.about.com/cs/brandmktg/a/whatisbranding.htm. Accessed on:18/ 06/ 2014.

Naufel, J. V. and Dipak, C. J. (1991), “ModelingPurchase-Timing and Brand-SwitchingBehavior Incorporating Explanatory Variables


21

and Unobserved Heterogeneity”, Journal ofMarketing Research, Vol. 28, No. 1, pp. 29-41.

Park, C.W, Jun, S.Y and Shocker, A.D.(1996),“Composite Branding Alliances: Aninvestigation Extension and Feedback Effect”,Journal of Marketing Research, Vol. 33(4), pp. 453-467.

Richard, A. C. (1989), “A Brand Switching Modelwith Implications for Marketing Strategies”,Journal of marketing, Vol. 8, Iss. 1, pp. 89-99

Sandman, W. (2005), “On Optimal ImportanceSampling for Discrete–Time Markov Chains,

Quantitative Evaluation of System, Vol. 19, issue 22.

Sun,B., Scott, A. N. and Srinivasan, K. (2003),“Measuring the Impact of Promotions onBrand Switching When Consumers AreForward Looking”, Journal of MarketingResearch, Vol. 40, No. 4, pp. 389-405.

Umeshanand, G. (2008), “Study on BrandSwitching in Consumer Products”, MBAThesis, Department of Management andTechnology Chhattisgarh VivekanandaTechnical University India.

White, J. (1993), “A Survey of Applications ofMarkov Decision Process, Journal of theOperational Research Society, Vol. 44,pp. 11-20.


22

ABSTRACT

The construct of job satisfaction is widelyacknowledged as an important requirement forthe effective functioning of any organization. Oneimportant antecedent of Job satisfaction isjustice perception. Universities in Sri Lanka arefacing challenges in improving job satisfaction ofadministrative staff and thus, their commitmentto gain competitive advantage and retention ofthe efficient administrative staff. Further thelevel of justice perception is found to be lowamong the administrative staff. Therefore,objective of this study was to find out theinfluence of justice perception on the jobsatisfaction. Self administered questionnaire wasused to collect the data from a sample of 250administrative staff members employed in 15national universities in Sri Lanka. Analysis ofthis study revealed that the state universityadministrative employees were generallysatisfied with their job. However, satisfactionwith opportunity for independent thought,feedback on performance, pay for job, promotionopportunities, benefits received were very lowamong the staff. The findings of the study alsoshowed positive relationship of justice perceptionwith job satisfaction. Implications of findings,limitations and areas for future research alsodiscussed.

Keywords: Job satisfaction, Organizationaljustice perception, distributive justice, proceduraljustice, interactional justice, administrative staffin universities.

Introduction

The concept of job satisfaction according toRobbins (2001) can be described as a generalattitude towards one’s job; the differencebetween the rewards received and what theyactually believe they should have received. Inorganizational studies, Currivan (2000) claimedthat job satisfaction is widely studied as workoutcomes in organizational settings. Numerousresearchers (for instance Goris et al., 2000)conceptualized job satisfaction as multifacetedinstrument that consisted of “work itself”,“quality of supervision”, “relationships withcoworkers”, “promotion opportunities”, and“pay”. The job satisfaction can therefore bedefined as a collection of attitudes, feelings,beliefs and behavior one has towards his or herjob.

Employees with higher job satisfaction areimportant as they believed that the organizationwould be of tremendous future for them. Hencethey are more committed to the organization,have higher retention rates and tend to havehigher productivity (Fatt, Khin & Heng, 2010).Therefore, organizations must strive to identifyantecedents that influence the job satisfaction ofemployees. One such antecedent isorganizational justice which describes theindividual’s perception of the fairness oftreatment received from an organization and theirbehavioural reactions to such perceptions(Fernandes & Awamleh, 2006). It has shown tobe associated with several outcomes such as jobsatisfaction, work motivation (Suliman, 2007;

23

ORGANIZATIONAL JUSTICE AS AN ANTECEDENT OF JOB

SATISFACTION OF ADMINISTRATIVE STAFF IN SRI

LANKAN STATE UNIVERSITIES

M.I. Nawfer

Deputy Registrar, South Eastern University, Sri Lanka

F.H. Abdul Rauf

Senior Lecturer, Department of Management, South Eastern University, Sri Lanka

Fernandes & Awamleh, 2006; Cropanzano et al.,2001; Moorman, 1991), intention to turnover(Colquitt et al., 2001), work performance(Suliman, 2007; Fernandes & Awamleh, 2006;Phillips et al., 2001), commitment (Folger &Konovsky, 1989), organizational citizenshipbehaviour (Moorman, 1991).

Organizational justice is typically conceptualizedwith three components: distributive, procedural,and interactional justice (Cropanzano et al.,2001; Masterson et al., 2000; McDowall &Fletcher, 2004). Distributive justice, recognizedas first sub-dimension of organization justice,mainly considered with the workers’ perceptionin the fairness of outcomes, such as monetaryrewards obtained by the workers from theorganization (Elovainio et al., 2005;Ramamoorthy & Flood, 2004; Greenberg, 2006;Aryee et al., 2002). This is defined by Moorman(1991) as “the fairness of outcomes an employeereceives such as pay and promotions”. This typeof justice is based on equity theory whichemphasized on the judgments made by theemployees about the outcomes (for example,promotion, pay) offered by the organizationagainst their effort by which they work or inaccordance with given criteria (Cropanzano & Greenberg, 1997; Hubbel & Chory-Assad,2005; Blakely et al., 2005; Alder & Ambrose,2005). Distributive justice is also considered asthe leading factor towards organizationaleffectiveness (Tang & Sarsfield-Baldwin, 1996).

Procedural justice refers to the perceived fairnessof the processes used to determine organizationaloutcomes (Colquitt et al., 2001; Folger &Konovsky, 1989). It derives from the perceivedequity of organizational policies and proceduresdetermining resource allocation and othermanagerial decisions (Peele III, 2007). Employeesjudge the equity of procedures by the amount ofbias, the breadth and accuracy of informationgathering, number of relevant parties given voicein the decisions, ethical standards applied, and theconsistency and universality of decisionimplementation (Stecher & Rosse, 2005).

Interactional justice focuses on employees’perceptions about the fairness of the interpersonaltreatment received during implementation (Bies& Moag, 1986). It refers to the quality ofinterpersonal processes and treatment ofindividuals (i.e. being treated with dignity andrespect), as well as the extent to which reasonsbehind the outcomes are explained (Bies &Moag, 1986). Perceptions of interactional justiceresult from supervisor trust-building behaviourssuch as “availability, competence, consistency,discreetness, fairness, integrity, loyalty, openness,promise fulfillment, receptivity and overall trust”(Deluga, 1994, p. 317).

Effects of Organizational Justice

on Job Satisfaction

The influence of different dimensions oforganizational justice (procedural, distributive,interactional) on job satisfaction is a widelyresearched topic and hence explains theimportance of organizational justice in anorganization (Cohen-Charash & Spector, 2001;Colquitt et al., 2001; Viswesvaran & Ones,2002). A basic element in employee’s satisfactionand organizational productivity is organizationaljustice (Aydin & Kepenekci, 2008).

Organizational justice puts stronger impact ondifferent attitudes of the employees like jobsatisfaction and organizational commitment (e.g.Bakhshi & Kumar, 2009; Colquitt et al., 2001).Folger and Konovsky (1989) found positiveassociation among the dimensions oforganizational justice with job satisfaction. Thestudies of Colquitt et al., (2001), Folger andCropanzano (1998), Cohen-Charash and Spector(2001) also found that justice dimensions havepositive and significant impact on jobsatisfaction.

Fernandes and Awamleh (2006) conducted aresearch to find the impact that three dimensionsof organizational justice (procedural, distributiveand interactional justice) have on job satisfactionand self assessment performance. The results of


24

the study revealed that all three dimensions oforganizational justice were significantlyinfluenced job satisfaction. A meta-analysisfound that distributive justice is a crucialpredictor of job satisfaction (Colquitt et al., 2001;Lambert, 2003). Distributive justice is amongvarious determinants of job satisfaction (Feinstein& Vondrasek, 2001). Folger and Konovsky(1989) and McFarlin and Sweeney (1992) foundthat perceptions of distributive justice aresignificantly correlated with pay raise satisfactionas well as with job satisfaction (Martin & Bennet,1996). Contemporary studies noted thatemployees have a tendency to display feeling ofdissatisfaction as they perceive an unfair contentof rewards (Cropanzano & Greenberg, 1997).

Past researches showed that procedural justicealso has a relationship with employee satisfaction(Konovsky & Cropanzano, 1991; Lambert,2003), because when employees observe thatperformance rating and chances of promotion arenot based on justice practices but on political andbiased motives, and their performance is nottruly considered, they become de-motivated andtheir satisfaction with job decreased. In anotherstudy, Bakhshi, Kumar and Rani (2009) reportedthe positive and significant association ofdistributive justice and procedural justice withorganizational commitment and job satisfaction.Consistent with the prior findings, Najafi et al.(2011) also concluded that educational experts ofdifferent universities reported higher jobsatisfaction on the provision of organizationaljustice. Fatt et al. (2010) reported that “thehigher level of employee’s perception towardsprocedural justice and distributive justice tendedto increase the level of employees’ jobsatisfaction and organizational commitment”(p.13). Findings from empirical studies in avariety of settings provide evidence that the threetypes of organizational justice have significanteffects on the job satisfaction. Based on theabove literature, this study seeks to investigatehow significance is the perceptions oforganizational justice on the job satisfaction ofadministrative staff in Sri Lankan stateuniversities.

Study Context

Although academic staff and students areundoubtedly integral to the higher educationenterprise, another relevant stakeholder group inhigher education is administrative staff members.Administrative staff in universities undertakes awide variety of duties, including those performedby analysts, secretaries, personal assistants andexecutives.

Universities in Sri Lanka are facing challenges inimproving the job satisfaction of theiradministrative staff and thus, their commitmentto gain competitive advantage and at the sametime retention of the efficient administrativestaff. There are many changes occurring in SriLankan higher education sector today. Accordingto the proposals for the university reformsintroduced by the government in 1998, and someof the potential activities such as expansion ofuniversity education, diversification of universitycourses and curriculum reforms, private sectorlinkage, staff training and development, qualityimprovements and the growth of student intakefor universities etc., have been identified to bestrengthened in the university education in thefuture.

Therefore, in this rapidly changing environment,all administrators in the system should beencouraged and motivated to face the newchallenges associated with their jobs, which willbe helpful to increase their job satisfaction.Because all administrative functions depend onadministrative staff to perform them whereby thesatisfied administrators can provide good serviceto the institution and thus the institution canfunction smoothly.

The findings of the survey conducted by Luo,Fan and Premakumara (2008) on Job Satisfactionof University Administrators: EmpiricalEvidence from Sri Lanka show that most of theuniversity administrators were dissatisfied withtheir present salary, promotions and careeradvancement. List of present pay related factorsexplaining the job dissatisfaction of the

Organizational Justice as an Antecedent of Job Satisfaction of AdministrativeStaff in Sri Lankan State Universities

25

respondents included but not limited to thefollowing: not reflecting their work performanceand responsibility, not matching with the marketvalue, same salary scales offered to officers whodo not bear any responsibility or accountability,not matching with the cost of living, insufficientwhen compared with the private sector, and notmatching with their qualifications. The factorswhich they were dissatisfied on promotion are asno performance based system, long delays in thepromotions, no proper procedure and scheme,more weight given for experience thanqualifications, third party interventions whichopen avenues for unsuitable persons to getpromotions, and no uniformity and notransparency in circulars and the policies. Anexamination of the respondents’ responsesrevealed the list of factors, which contributemostly to dissatisfaction with personal careergrowth and advancement: limited prospects forcareer advancement, no formalized training, noequal opportunities in training, favourites gettingopportunities even for unwanted training, notraining need analysis, and superiors consideringtheir own advancement.

According to a primary survey conducted by theresearchers among the administrative staff in SriLankan state universities in August 2013, it wasfound that there have been a number of problemsexisting among them which contribute to jobdissatisfaction related to justice perceptions.Some of them are: extra working hours withoutadditional payment, work pressure, low level oftreatment by top management, bad workingenvironment, less promotion opportunities, workunfairness, low salary level, etc. It was alsofound in the survey that many administrativestaff members have shown intentions to exitfrom their administrative positions and join otherbetter jobs if they could gain. The job requireshigher standard of qualifications and a toughrecruitment system. However, the benefits arenot much attractive when compared to someother comparable positions which require thesame level of qualifications.

Further, according to the existing scheme, thepromotional opportunities for administrative staffin Sri Lankan state universities seem to be unfair.For example, an Assistant Registrar aftercompletion of the necessary requiredqualifications to become a Senior AssistantRegistrar has to wait until a cadre falls vacant inthe university system. The promotion does notdepend only on merit. This is the case for aSenior Assistant Registrar to become a DeputyRegistrar. This unfair situation of less promotionopportunities tends to push the administrativestaff into job dissatisfaction. Therefore, it isnecessary to study the effects of justiceperception on the job satisfaction of theadministrative staff and find solutions toeliminate or minimize these problems for thesmooth functioning of the universities.

Therefore, objectives of the study are to find outthe general influence of justice perception on thejob satisfaction of administrative staff in SriLankan state universities and also to find out therelationship between dimensions of justiceperception and job satisfaction. It is also aimedto determine which dimension of the justiceperception has more effects on the jobsatisfaction of administrative staff in Sri Lankanstate universities. Although many researcheshave been carried out on employee jobsatisfaction in general, no research has been doneso far related to the effects of justice perceptionon the job satisfaction of administrative staff inSri Lankan universities. In this research, throughcombining theoretical and empirical research, itis aimed to find out the effects of justiceperception dimensions on universityadministrators’ job satisfaction, thereby layingthe theoretical foundation for the practice and forfuture research.

Based on the review of literature on the effectsof organizational justice on the job satisfactionand the existing problem of administrative staffin Sri Lankan state universities the followingconceptual framework (Figure :1) is developed.


26

Figure 1: Conceptual Model

Based on the aforementioned conceptual model,the following hypotheses have been developed.

H1. Distributive justice perception is positivelyassociated with the job satisfaction ofadministrative staff in Sri Lankan state universities.

H2. Procedural justice perception is positivelyassociated with the job satisfaction ofadministrative staff in Sri Lankan state universities.

H3. Interactional justice perception ispositively associated with the job satisfaction ofadministrative staff in Sri Lankan stateuniversities.

Methods

The study used a correlational design to examinethe relationship between two quantitativevariables. The aim of this study is to determinethe degree of the relationship between justiceperception and job satisfaction.

Sample characteristics

Sampling method employed was stratified. Datawere collected from 250 administrative staffmembers employed in 15 national universitiescoming under the purview of the UniversityGrants Commission of Sri Lanka through aquestionnaire.


27

Organizational

Justice

v Distributive Justice

v Procedural Justice

v InteractionalJustice

Job

Satisfaction

Table 1: Characteristics of Respondents (N=139)

Gender Male 83 59.7Female 56 40.3

Age groups Below 30 years 09 6.531 – 40 years 52 37.441 – 50 years 57 41.051 – 60 years 21 15.1

Education Level G.C.E (A/L) 08 5.8Bachelor Degree 47 33.8Postgraduate Diploma 35 25.2Master Degree 49 35.3

Position Registrar 05 3.6Bursar 02 1.4Deputy Registrar 13 9.4Deputy Bursar 04 2.9Senior Asst. Registrar 32 23.0Senior Asst. Bursar 12 8.6Asst. Registrar 51 36.7Asst. Bursar 20 14.4

Year of Experience Less than 02 years 27 19.402 – 05 years 33 23.706 – 10 years 30 21.611 – 15 years 27 19.416 – 25 years 15 10.8More than 25 years 07 5.0

Frequency Percentage

Source: Survey Data

139 questionnaires were returned, comprising aresponse rate of 58.4%. Seven responses wereeliminated due to excessive missing data. Table.1presents the profile of the respondents withregard to gender, education, employment status,job title, and years of experience. Most of therespondents (60.43%, n=84) had completedpostgraduate Diploma and Master Degree. Ofthis sample, 83 (59.7%) were male and 56(40.3%) were female. With regard to years ofemployment, most of the respondents (n=60,43.1%) had 0 - 5 years of experience.

Data were collected using questionnaire. Whilethe first part of the questionnaire intended tocollect the demographic variables of therespondents the second part was intended tocollect the data about independent and dependentvariables.

Measures

Perceptions of distributive justice were measuredwith the Distributive Justice Index, developed byPrice and Mueller (1986). This five-item scalemeasures the degree to which rewards receivedby employees are perceived to be related toperformance inputs. Each item asks for thedegree to which the respondent believes that heor she is fairly rewarded on the basis of somecomparison with responsibilities, education andtraining, effort, stresses and strains of job, andperformance. Items were re-worded toaccommodate the use of a 5-point scale rangingfrom (1) "strongly disagree" to (5) "stronglyagree." For example, the item, "How fair has thecompany been in rewarding you when youconsider the responsibilities you have?" waschanged to "I am rewarded fairly in view of theresponsibilities I have".

Perceptions of procedural justice were measuredusing six items. This scale is based on one usedby Moorman (1991). These six items weredesigned to measure the fairness of procedures inthe organization as revealed by procedures whichpromote consistency, bias suppression, accuracy,

correctability, representativeness, and ethicality.For example, “My organization’s proceduresgenerate standards so that decisions can be madewith consistency”. Items measure the responseusing a 5-point scale ranging from (1) "stronglydisagree" to (5) "strongly agree”.

Perceptions of interactional justice weremeasured using 15 items. This scale is based onone used by Moorman (1991). The items weredesigned to measure supervisor consideration ofemployee rights, treatment of employees withrespect and kindness, and provision ofexplanations and justification for decisions. Forexample, “My immediate supervisor treats mewith kindness and consideration”. Items measurethe response using a 5-point scale ranging from(1) "strongly disagree" to (5) "strongly agree”.

Job satisfaction was measured using twenty threeitems. This scale is based on MinnesotaSatisfaction Questionnaire (MSQ). This scaleuses multiple items. There are two good reasonsto use multiple items. First, multiple item scalesare more reliable than single items. This isbecause respondents can make mistakes whenfilling out questionnaires. Errors can be madewhen a respondent interprets a questiondifferently than intended. Second, multiple itemsallow for a more complete assessment of a facet.A single item may not do a good job of coveringall aspects. For example, an employee may beable to indicate their overall satisfaction with payin a single item, but pay includes many aspectsthat would take several items to cover. Thequestionnaire includes a 23-item scale tomeasure seven specific satisfactions; pay andbenefits, job security, social, supervisory, growthsatisfaction, work environment and nature of thejob. The format for the facet items is a five-pointscale ranging from (1) "very low" to (5) "veryhigh".

The questionnaires were prepared in Englishsince the sample were able to understand Englishlanguage which is a compulsory component forrecruitment of administrative staff in Sri Lankanuniversity system. Contact information of the


28

researcher was provided to the participants incase any questions or concerns arise. Theresearcher visited universities and distributedsurveys to participants. Also the questionnaireswere sent to participants and received backthrough email and fax. A covering letter from theresearcher accompanied each survey.

Data Analysis

In order to achieve acceptable levels ofmeasurement reliability and validity, a pilot studywas conducted among 38 respondents to refinethe original survey instrument.

The collected questionnaires were then subjectedto primary analysis in order to ascertain whetherthe respondents understood the instructionsclearly and they followed them simply and easilyand to judge the time taken to complete thescale, and specially, whether the scale items wereappropriate for the target respondent population.Tests of internal consistency (Cronbach's alpha)were conducted to assess the reliability of eachof the scales used. All of the measures includedin the questionnaire showed adequate levels ofinternal consistency reliability. Normality waschecked with data collected and was found that

it had distributed normal in the Z curve. Table 2reports the descriptive statistics for the measuresused, including Mean, standard deviation, andinternal consistency reliability for each measure.These values show that the constructs achievedare all excellent and high reliabilities and thealpha values indicated that the study’s instrumentand data were reliable.

Correlations between the independent anddependent variables of this study are also givenin Table 2. To examine the relationship betweendistributive justice perception (DJP) and overallsatisfaction (OAS), this study performedcorrelation and regression analysis. A significant

correlation was found between distributivejustice perception (DJP) and overall satisfaction(OAS), r = 0.456, p = 0.000. A significantcorrelation between procedural justice perception(PJP) and overall satisfaction (OAS), r = 0.448,p = 0.000, between interactional justiceperception (IJP) and overall satisfaction (OAS),r = 0.521, p = 0.000, and between distributivejustice perception (DJP) and intrinsic satisfaction(INS), r = 0.493, p =0.000 were also found inthis study.


29

Table 2: Correlations among Variables (N=139)

Distributive JusticePerception (DJP) 2.72 1.06 0.94 1

Procedural JusticePerception (PJP) 2.97 0.89 0.90 0.11 1

Interactional Justice Perception (IJP) 3.29 0.91 0.87 0.33** 0.44** 1

IntrinsicSatisfaction (INS) 3.19 0.93 0.90 0.49** 0.36** 0.50** 1

ExtrinsicSatisfaction (ENS) 3.13 0.93 0.89 0.45** 0.48** 0.55** 0.79** 1

OverallSatisfaction (OAS) 3.12 0.92 0.90 0.45** 0.44** 0.52** 0.87** 0.93** 1

Source: Survey Output

M SD Alpha DJP PJP IJP INS ENS OAS

A significant correlation between proceduraljustice perception (DJP) and intrinsic satisfaction(INS), r = 0.360, p = 0.000, betweeninteractional justice perception (IJP) and intrinsicsatisfaction (INS), r = 0.507, p =0.000, betweendistributive justice perception (DJP) and extrinsicsatisfaction (ENS), r = 0.453, p =0.000, betweenprocedural justice perception (PJP) and extrinsicsatisfaction (ENS), r = 0.483, p = 0.001. andbetween interactional justice perception (IJP) andextrinsic satisfaction (ENS), r = 0.550, p = 0.000were also found from the analysis.

Discussion of Findings

Organizational justice perceptions i.e. distributivejustice perception, procedural justice perceptionand interactional justice perception are positivelycorrelated with the overall satisfaction of theadministrative staff in Sri Lankan stateuniversities. These findings are consistent withthe studies of Folger and Konovsky (1989),Colquitt et al., (2001), Folger and Cropanzano(1998), Cohen-Charash and Spector (2001) whofound positive and significant association amongthe dimensions of organizational justice with jobsatisfaction.

The results of the regression analysis suggest thatthe interactional justice was the best predictor ofoverall satisfaction (β = 0.204, t = 3.712, p =0.00) as well as intrinsic satisfaction (β = 0.252,t = 3.841, p = 0.00) and extrinsic satisfaction (β= 0.250, t = 4.133, p =0.00) of the administrativestaff in Sri Lankan state universities, followed byprocedural and finally the distributive justice.This finding indicate that the quality of treatmentthat administrative staff receives from their topmanagement when policies and procedures areimplemented at the workplace are the mostimportant predictor of their level of satisfaction.Honesty, courtesy, timely feedback, respect forrights and the chances to express viewpoints arethe most critical components for securingsatisfied administrative members in the SriLankan university system. Administrative staffmembers are satisfied when their immediate

supervisors interact with them in a transparentmanner and the procedures are clear ingenerating standards to make consistentdecisions, and being able to request forclarification or additional information about thedecision. The more the universities improve theinteractional justice in the university system, themore the administrative staff will be satisfiedboth intrinsic and extrinsic.

Procedural justice perception also influences theoverall, intrinsic and extrinsic satisfaction of theadministrative staff significantly. Overallsatisfaction (β = 0.150, t = 3.860, p = 0.00),intrinsic satisfaction (β = 0.113, t = 2.436, p =0.00) and extrinsic satisfaction (β = 0.186, t =4.368, p =0.00).

Distributive justice perception also plays animportant role in the overall, intrinsic andextrinsic satisfaction of the administrative staff.Overall satisfaction (β = 0.118, t = 4.685, p <0.05), intrinsic satisfaction (β = 0.156, t =5.217, p = 0.00) and extrinsic satisfaction (β =0.128, t = 4.638, p = 0.00). The results of thisstudy support previous research conducted toexplain the importance of the allocationphenomenon in organizations (Alexander &Ruderman, 1987; Cropanzano & Greenberg,1997; Folger & Konovsky, 1989). For example,people tend to be more satisfied with outcomesthey perceive to be fair than with those theyperceive to be unfair. In addition, people maycompare the adequacy of the rewards theyreceive to their expectations, or referentstandards. Thus, if employees feel discontentwith what they receive compared to those doingsimilar jobs, they are more likely to quit. Thissupports the contemporary studies which notedthat employees have a tendency to displayfeeling of dissatisfaction as they perceive anunfair content of rewards (Cropanzano &Greenberg, 1997).

Administrative staff members in Sri Lankan stateuniversities are generally satisfied with their job.However, satisfaction with opportunity forindependent thought, feedback on performance,


30

pay for job, promotion opportunities, benefitsreceived were low among the staff. The study’sfindings indicate that administrative staff whotend to show positive feelings towardsdistributive, procedural and interactional justiceare likely to report higher level of job satisfaction.These findings support the studies of somescholars in this field (Cohen-Charash & Spector,2001; Colquitt, et al., 2001; Lambert, 2003;Fernandes & Awamleh, 2006; Suliman, 2007).

Generally, all the administrative staff membershave indicated their dissatisfaction with theirlevel of pay. Among them, the administrativestaff with more than 25 years of service haveshown very low satisfaction with their pay (M =1.43), the next group which is less satisfied withpay is those who have less than 02 years ofexperience. With respect to supervision, theadministrative staff have generally indicated theirsatisfaction, however, the new recruits ofadministrative members have showndissatisfaction (M = 2.81). The most experiencedstaff members are the most satisfied ones withthe supervision facet.

Conclusions and Recommendations

This study explores perceptions of theadministrative staff towards organizationaljustice in the form of distributive justice,procedural justice, and interactional justice. Thisstudy was conducted to examine the relationshipbetween organizational justice perceptions andthe job satisfaction of the administrative staff inSri Lankan state universities.

In view of the essential role of administrativestaff members in the effective functioning of theuniversities, one might ask whetheradministrative staff members’ perception oforganizational justice and their level of jobsatisfaction should be an issue of concern toleadership of universities. As the findings of thisstudy revealed, the answer is yes. Interestingly,the use of pay as the main source of motivationto increase the level of job satisfaction amongadministrative staff members might not work.

Through the analysis of this study it can beconcluded that the state university administrativeemployees were generally satisfied with theirjob. However, satisfaction with opportunity forindependent thought, feedback on performance,pay for job, promotion opportunities, benefitsreceived were very low among the staff. Thefindings of the study showed positiverelationship of distributive justice, proceduraljustice and interactional justice with overall jobsatisfaction as well as intrinsic and extrinsicsatisfaction of administrative staff in Sri Lankanstate universities. The findings were consistentwith the prior researchers that the organizationaljustice dimensions foster the overall jobsatisfaction of the employees (Bakhshi et al.,2009; Zaman et al., 2010; Ramamoorthy &Flood, 2004; Aryee et al., 2002; Najafi et al.,2011). These findings supported the studies ofFolger and Konovsky (1989), Colquitt et al.,(2001), Folger and Cropanzano (1998), Cohen-Charash and Spector (2001) who found positiveand significant association among the dimensionsof organizational justice with job satisfaction.

Interactional justice was found to have moreeffect on overall job satisfaction ofadministrative staff members serving in stateuniversities. The results of this study supportprevious researches on the impact of the qualityof the supervisor-subordinate relationship on thefairness perceptions of subordinates (Dansereau,Graen, & Haga, 1975; Podsakoff, MacKenzie,Moorman, & Fetter, 1990). Thus, theadministrative staff would probably receive morejustification for procedural justice as well asdistributive justice due to the relative advantageof higher quality interactions and a closerrelationship with the higher level management.Administrative staff’s perceptions of fairness areenhanced when they feel that they are valuedmembers of the university system.

As this study found perceived fairness as apredictor of job satisfaction, a higher level ofperception leads to a high level of jobsatisfaction, which in turn contribute toindividual and organizational performance /


31

outcomes. With the understanding of theinfluence of organizational justice perception onjob satisfaction, leadership of universities caneffectively improve administrative employees’perception of organizational justice through thecommunication of critical institutional values,needs and expectations and on how these values,needs, and expectations align with that of theadministrative staff.

This study would recommend that the topmanagement of the universities and policymakers like University Grants Commission couldboost the overall job satisfaction of theiradministrative staff members by promotingdistributive, procedural and interactional justicepractices in their respective universities.

Implications of findings

The results of this study provide both theoreticaland practical implications. Despite the fact thatorganizational justice is an important factor as abasic requirement for the effective functioning oforganizations, there have been no empiricalresearch of organizational justice in the SriLankan university sector. As anticipated, thisstudy revealed the importance of justiceperception on job satisfaction in Sri Lankan stateuniversities. Thus, this study provides furthersupport for the past findings on these constructsfrom the Sri Lankan sample. The present studyfound that organizational justice emerged as thestronger predictor of the job satisfaction.Therefore, the management concerned with theeffectiveness and vitality of their institution,should be concerned with this phenomenon.

Although this study was conducted in auniversity environment the results are applicableto organizations such as non profit, and businessentities, because the issues of justice perceptionand job satisfaction are relevant to employee andorganizational level outcomes in all types oforganizations. Policy makers may institute

policies or programs to recognize and reward theemployees, and increase their overall jobsatisfaction.

The results have several valuable practicalimplications for the managers. Managers need toapply rules fairly and consistently to allemployees, and reward them based onperformance and merit without personal bias inorder to create a positive perception of justice.The perceptions of unfairness can result innegative reactions to the organization, due topoor job satisfaction. The present findingssuggest that procedural fairness has more effecton their job satisfaction than distributive justicedoes. Hence, management should pay moreattention to the means or the process of decisionmaking for the distribution as it will lead tosubstantial pay-offs in individual job satisfaction.Therefore, management can influence importantwork attitudes through creation and maintenanceof a procedurally fair climate.

Limitations and Areas for FutureResearch

As with any research, this study too has somelimitations that should be acknowledged. First,the study examined administrative staff membersonly in Sri Lankan state universities, andtherefore the generalizability of the findings islimited to state universities of Sri Lanka comingunder the purview of the University GrantsCommission. There are many other universitiesestablished under various acts and also there aresome private universities currently operating inSri Lanka. Future research should extendcovering these universities too. Further, all thesefactors constrain the ability to make causalstatements about the examined relationships, andthe exclusive use of self-reported data may create the potential for common method biases. In addition, qualitative data could becollected to provide further explanations for thefindings.


32

References

Alder, G. S., & Ambrose, M. L. (2005), Towardsunderstanding fairness judgments associatedwith computer performance monitoring: Anintegration of the feedback, justice, andmonitoring research. Human ResourceManagement Review, Vol. 15, pp. 43–67.

Alexander, S., & Ruderman, M. (1987), The role ofprocedural and distributive justice inorganizational behaviour. Social JusticeResearch, Vol. 1, No. 2, pp. 177-198.

Aryee, S., Budhwar, P. S., & Chen, Z. X. (2002),Trust as a mediator of the relationship betweenorganizational justice and work outcomes: test of a social exchange model. Journal of Organizational Behaviour, Vol.23, pp.267–285.

Aydin, I. & Kepenekci, Y.K. (2008), Principals’opinions of organizational justice in elementaryschools in Turkey, Journal of EducationalAdministration, Vol. 46, No. 4, pp. 497- 513

Bakhshi A, Kumar K, Rani E (2009),Organizational justice perceptions as predictorof job satisfaction and organizationalcommitment. International Journal of BusinessManagement, Vol. 4, No. 9, pp. 145-154

Bies, R. J., & Moag, J. (1986), Interactional justice:communication criteria of fairness, in Lewicki,R., Sheppard, B., and Bazerman, M. (Ed.).Research on Negotiation in Organizations, Vol. 1, pp. 43-55, JAI Press, Greenwich, CT.

Blakely, G., Andrews, M., & Moorman, R. (2005),The moderating effects of equity sensitivity onthe relationship between organizational justiceand organizational citizenship behaviour.Journal of Business Psychology, Vol. 20, No. 2, pp. 259-273.

Cohen-Charash, Y., & Spector, P. (2001), The roleof justice in organizations: A meta analysis.Organizational Behaviour and HumanDecision Processes, Vol. 86, pp. 278-232.

Colquitt, J. A., Conlon, D. E., Wesson, M. J.,Porter, C.O. L. H., & Ng, K. Y. (2001), Justiceat the millennium: a meta-analytic review of 25years of organizational justice research.Journal of Applied Psychology, Vol. 86, No. 3, pp. 425–45.

Cropanzano, R., Byrne, Z., Bobocel, D. R., &Rupp, D. (2001), Moral virtues, f a i r n e s sheuristics, social entities, and other denizens oforganizational justice. Journal of VocationalBehaviour, Vol. 58, No. 2, pp. 164–209.

Cropanzano, R., & Greenberg, J. (1997), Progressin organizational justice: tunneling t h r o u g hthe maze, in Cooper, C. L. and Robertson, I. T.(Ed.). International Review of Industrial andOrganizational Psychology, Vol. 12, pp. 317–72, Wiley, New York.

Currivan, D.B. (2000), The causal order of jobsatisfaction and organizational commitment inmodels of employee turnover, HumanResource Management Review, Vol. 9, No. 4,pp. 495-524.

Dansereau, F., JR., Graen, G., & Haga, W. J.(1975), A vertical dyad linkage approach toleadership within formal organizations: Alongitudinal investigation of the role makingprocess. Organizational Behaviour and HumanPerformance, Vol. 13, pp. 46-78.

Deluga R. J. (1994), Supervisor trust building,leader-member exchange and organizationalcitizenship behaviour. Journal of Occupationaland Organizational Psychology, Vol. 67, No. 4, pp. 315-26.

Elovainio, M., Van den Bos, K., Linna, A.,Kivimaki, M., Ala-Mursula, L., Pentti, J., &Vahtera, J. (2005), Combined effects ofuncertainty and organizational justice onemployee health: Testing the uncertaintymanagement model of fairness judgmentsamong Finnish public sector employees. SocialScience and Medicine, Vol. 61, No. 12, pp. 2501-2512.


33

Fatt CK, Khin EWS, Heng TN (2010), The Impactof Organizational Justice on Employee’s JobSatisfaction: The Malaysian CompaniesPerspectives. Am. J. Econ. Bus. Adm., Vol. 2, No. 1, pp. 56-63.

Feinstein, A.H. & Vondrasek, D. (2001), A study ofrelationships between job satisfaction andorganizational commitment among restaurantemployees, Journal of Hospitality, Tourism,and Leisure Science, available at:http://hotel.unlv.edu/pdf/jobSatisfaction.pdf(accessed February 15, 2013).

Fernandes, C., & Awamleh, R. (2006), Impact oforganizational justice in an expatriate workenvironment. Management Research News,Vol. 29, No. 11, pp. 701 –12.

Folger, R., & Konovsky, M. A. (1989), Effects ofprocedural and distributive justice on reactionto pay raise decisions. Academy ofManagement Journal, Vol. 32, No. 1, pp. 115-130.

Goris, J. R., Vaught, B. C., & Pettit, J. D. (2000),Effects of communication direction on jobperformance and satisfaction: a moderatedregression analysis. Journal of BusinessCommunication, Vol. 37, No. 4, pp. 348-368.

Greenberg, J. (2006), Losing Sleep OverOrganizational Injustice: AttenuatingInsomniac Reactions to UnderpaymentInequity With Supervisory Training inInteractional Justice. Journal of AppliedPsychology, Vol. 91, No. 1, pp. 58-69.

Hubbel, A. P., & Chory-Assad, R. M. (2005),Motivating Factors: Perceptions of Justice andTheir Relationship with Managerial andOrganizational Trust. Communication Studies,Vol. 56, No. 1, pp. 47-70.

Konovsky, M.A. & Cropanzano, R. (1991),Perceived fairness of employee drug testing asa predictor of employee attitudes and jobperformance, Journal of Applied Psychology,Vol. 76, pp. 698-707.

Lambert, E. (2003), Justice in corrections: Anexploratory study of the impact oforganizational justice on correctional staff.Journal of Criminal Justice, Vol. 31,pp. 155−168.

Luo, Fan, & Premakumara, P. (2008), JobSatisfaction of University Administrators:Empirical Evidence from Sri Lanka. The 5th

International Conference on Innovation andManagement, Wuhan University ofTechnology, Netherlands, pp. 1901-1906.

Martin, C. L., & Bennett, N. (1996), The role ofjustice judgments in explaining the relationshipbetween job satisfaction and organizationalcommitment. Group & OrganizationalManagement, Vol. 21, No. 1, pp. 84-104.

Masterson, S., Lewis, K., Goldman, B. M.,&Taylor, M. S. (2000), Integrating justice andsocial exchange: the differing effects of fairprocedures and treatment of work relationships.Academy of Management Journal, Vol. 43, No. 4, pp. 738-48.

McDowall, A., & Fletcher C. (2004), Employeedevelopment: An organizational justiceperspective. Personnel Review, Vol. 33, No. 1, pp. 8-29.

Moorman, R. H. (1991), Relation betweenorganizational justice and organizationalcitizenship behaviours: does fairness perceptioninfluence employee citizenship? Journal ofApplied Psychology, Vol. 76, No. 6, pp. 845-55.

Najafi, S., Noruzy, A., Azar, H. K., Nazari-Shirkouhi, S., & Dalvand, M. R. (2011),Investigating the relationship betweenorganizational justice, psychologicalempowerment, job satisfaction, organizationalcommitment and organizational citizenshipbehaviour: An empirical model. AfricanJournal of Business Management, Vol. 5, No. 13, pp. 5241-5248.


34

Peelle III, H. E. (2007), Reciprocating perceivedorganizational support through citizenshipbehaviour. Journal of Managerial Issues, Vol. 19, No. 4, pp. 554-575.

Phillips, J., Douthitt, E., & Hyland, M. (2001), Therole of justice in team member satisfactionwith the leader and attachment to the team.Journal of Applied Psychology, Vol. 86, No. 2, pp. 316-25.

Podsakoff, P.M., MacKenzie, S.B., Moorman,R.H., & Fetter, R. (1990), Transformationalleader behaviours and their effects onfollowers' trust in leader, satisfaction, andorganizational citizenship behaviours. LeaderQuarterly, Vol. 1, No. 2, pp. 107-142.

Price, J.L., & Mueller, C. W. (1986), Handbook ofOrganizational Measurement. Marshfield, MA:Pittman.

Ramamoorthy, N., & Flood, P. C. (2004), Genderand Employee Attitudes: The Role ofOrganizational Justice Perceptions. BritishJournal of Management, Vol. 15, pp. 247–258.

Robbins, S.P. (2001), Organizational behaviour.(9th ed.). New Jersey: Prentice Hall

Stecher, M.D., & Rosse, J.G. (2005), Thedistributive side of interactional justice: Theeffects of interpersonal treatment on emotionalarousal. Journal of Managerial Issues, Vol. 17, No. 2, pp. 229-46.

Suliman, A.M.T. (2007), Links between justice,satisfaction and performance in the workplace:A survey in the UAE and Arabic context.Journal of Management Development, Vol. 26, No. 4, pp. 294-311.

Tang, T.L., & Sarsfield-Baldwin, L.J. (1996),Distributive and procedural justice as related tosatisfaction and commitment. SAM AdvancedManagement Journal, Vol. 61, No. 3, pp. 25-31.

Viswesvaran, C. & Ones, D.S. (2002), Examiningthe construct of organizational justice: a meta-analytic evaluation of relations with workattitudes and behaviours, Journal of BusinessEthics, Vol. 38, No. 3, pp. 193-203.

Zaman, G., Ali, N., & Ali, N. (2010), Impact ofOrganizational Justice on EmployeesOutcomes: An Empirical Evidence. AbasynJournal of Social Sciences, Vol. 3, No. 1.


35

36

ABSTRACT

Corruption in developing countries has beenhigher than developed countries. Most importantobjective of this study is to investigate theimpacts of economic and political freedom oncorruption in developing countries. Thisempirical work engages 70 developing countriesduring 2000-2004 from all regions over theworld. It examines the relationship betweencorruption, measured by corruption perceptionindex and economic political causes ofcorruption, measured by direct and proxymeasures. Results of regression analysis showthat economic freedom significantly andpositively affects corruption. Even democraticfreedom positively affect corruption, it does nothave statistical significant. Economic variables,income level of nations, economic freedom,capability of information accesses and tertiaryeducation have negative effects on degree ofcorruption in developing countries. Corruptionreduces income inequality and increase povertyin developing countries. Dependency ofagriculture, unemployment and government’slong procedure to start economic activitiesnegatively related with corruption in developingcountries. Democracy, multi party politicalsystem and weak majority of government(unstable regime) reduce the corruption. Humanright violations, conflicts and violence and nonpeace environment increase the corruption.Economic freedom is more powerful determinantin reduction of corruption than democraticfreedom. This study concludes that increasing ofeconomic freedom and openness, less control ofgovernment’s rule and regulations to startbusiness activities and increasing of people’s

information capability via extensive developmentof information accesses such as education,internet, telephones and institutions will leads toreduction of corruption in developing countries.In a political view, even thought moredemocratization, human security and capacity ofpeace building environment reduce corruption;democratic freedom has less power to reduce thecorruption than economic freedom. However,consequences of these determinants differcountry to country based on country’s leadership,people’s attitudes, sociopolitical institutions andreligious and cultural ethos.

Introduction

Corruption is generally regarded as one of themost serious obstacles to development. One ofthe recent surveys has shown this to be acommon perception of developing countries inworldwide. Brunetti surveyed 3600 entrepreneursfrom 59 countries, 58 of which considered lessdeveloped. Respondents were asked to choosefrom a menu of fifteen potential obstacles todoing business. Corruption was indicated to be aserious obstacle by a majority of businessman in35 countries, all less developed. Corruption wassecond only to “Tax regulation and /or high tax”as an obstacle to do business, outstripping otherfactors such as lack of infrastructure, financing,and various type of regulation (Robert C A,2001). Corruption arises from the incentives ofpublic and private agents to conspire in theconcealment of information from the government(Keithet al, 2006 and Samia C, 2007).

Definition of corruption is broad. Ed and Cloke(2004) show three types of criteria for definition

SOCIOPOLITICAL AND ECONOMIC CAUSES FOR

CORRUPTION IN DEVELOPING COUNTRIES:

A CROSS COUNTRY EMPIRICAL STUDY

Selvarathinam Santhirasegaram

Senior Lecturer, Department of Economics, University of Jaffna

of corruption. First, A basic legal definition (iecorruption is defined through the breaking of thelaw). Second, the articulation of a set of normsfor the formal duties of a public role (iecorruption occurs when those norms are broken).Third one is the identification of when theactions of individuals subvert the public interestfor private gain. According to Robert C A,(2001), corruption is defined as “Power of someto collect and keep fee for operation of anofficial business by others”. World Bank, (1997)defines corruption is “The abuse of the publicoffice for private gain”. It encompasses unilateralabuses by government officials such asembezzlement and nepotism, as well as abuseslinking public and private actors such as bribery,extortion, influence peddling, and fraud.Corruption arises in both political andbureaucratic offices and can be petty or grand,organized or unorganized. Though corruptionoften facilitates criminal activities such as drugtrafficking, money laundering, and prostitution,it is not restricted to these activities. Forpurposes of understanding the problem anddevising remedies, it is important to keep crimeand corruption analytically distinct. behavior onthe part of officials in the public sector, whetherpoliticians or civil servants, in which theyimproperly and unlawfully enrich themselves, orthose close to them, by the misuse of the publicpower entrusted to them. This would includeembezzlement of funds, theft of corporate orpublic property as well as corrupt practices suchas bribery, extortion. These definitions leave outmany high- profile corruption phenomena, suchas rigged privatization scheme and diversion offoreign aids. Corruption is further defined as“The misuse of public office for privategain”(Linda and Aaron Stern,2003).Corruptioninvolves behavior on the part of persons in whichthey improperly enrich themselves or those closeto them by misusing power with which they havebeen entrusted. In short, corruption is a misusingof public power for personal gain. There are twosegments of corruption, ‘Bureaucratic corruption’and ‘political corruption’, but, both areinterrelated in many cases.

Causes for Corruption

There are many sociopolitical and economiccauses for corruption in developing countries.

Political Freedom:

Any mechanism that increases politicalaccountability via political freedom, probablybased on democracy, either by encouraging thepunishment of corrupt individuals or by reducingthe informational problem related to governmentactivities, tends to reduce the incidence ofcorruption. It can be increased by allowingdemocratic freedom and development ofdemocratic institutions. The political science andeconomics literatures have extensively discussedthe role of political accountability in generatinggood governance practices and, particularly, inreducing corruption. According to the Bailey andValenzuela (1997), Persson et al. (2000), RoseAckerman (1999), Djankov et al. (2001), andLaffont and Meleu (2001), the countries whichhave more democratic freedom are able toreduce corruption. The degree of accountabilityin the system is determined by the specificfeatures of the political system. Three maincharacteristics can be identified in this respect ofpolitical system which are degree of competitionin the political system, the existence of checks-and-balances mechanisms across differentbranches of government, and the transparency ofthe system. Any institution or rule that providesa punishment mechanism for politicians, such asthe loss of elections or the possibility of beingforced out of office, can induce politicians toimprove their behavior by aligning their owninterests with those of their constituents. Themore the system forces politicians to face theelectorate, the higher are their incentives to stickto good governance. This would imply, forexample, those political systems that allow for(clean and fair) executive reelections would haveless myopic and more electoral consciouspoliticians, and, therefore, less corruption.

Sociopolitical and Economic Causes for Corruption in Developing Countries: A Cross Country Empirical Study

37

Following Persson and Tabellini (2000) proposethe following channel, which shows the waydemocracy and press freedom may work tocombat corruption. The presence of pressfreedom brings public corruption cases to thevoters while voters in a democracy in turn punishcorrupt politiciansPress freedom Voters state ofknowledge Democracy Selection of politicalpartiesState of corruption Treisman (2000) findsthat democracy is associated with less corruptiononly in countries which have been democraticfreedom for decades, which has not been thecase in Southeast Asia.. Lederman et al (2001)argue that countries with a British legal traditiontend to be less corrupt disappears once politicalvariables are included in the regression. Theeconomists Ades & Di Tella (1999) find“political rights” consistently had no significanteffect on corruption: “If anything, lack ofpolitical rights seems to be associated with lesscorruption” (1999, 987). Alternatively, thepolitical science literature is somewhat moresupportive of the influence of democracy ininhibiting corruption (Montinola & Jackman,2002, Sandholtz & Koetlze, 2000; Treisman,2000). Monopoly of political power allows themonopolist to exploit his or her position withoutfear of losing power to a competitor. Conversely,democracy increases uncertainty about whichofficial to corrupt, lowers the incentive to investin a particular politician, and at the same timeincreases the accountability of officials to theelectorate. Beyond ideological differences, atelection time the corrupt practices of theincumbent or the officials for whom theincumbent is accountable, creates an importantopportunity for the challenger. Democracy andthe consequent accountability raise the costs ofcorrupt behavior and likely deter bribe giving,therefore limiting the number of opportunitiespresented for corruption. The concept ofdemocracy and how it is to be measured is acritical issue in clarifying theoreticalrelationships in empirical analysis. This studymeasures democracy by index based on twoessential dimensions of democracy, electoralparticipation and competition. The concept of

democracy from the concept of liberalism isdifferent. Democracy is a multidimensionalconcept, but electoral competition andparticipation are core elements, theoreticallyimportant to the analysis of corruption, and bothrequire empirical analysis.

For a number of reasons, the risk of exposuremay also be higher in more democratic, openpolitical systems. Greater civic engagement maylead to closer monitoring. In democratic systems,competitors for office have an incentive todiscover and publicize the incumbent’s misuse ofoffice whenever an election beckons. Exposuremay also be more likely in more economicallydeveloped countries. Besides its apparent impacton democracy (Lipset, 1960), economicdevelopment increases the spread of education,literacy, and depersonalized relationships — eachof which should raise the odds that an abuse willbe noticed and challenged

Decentralization of political power leads togreater variety in provision of public goods,which are tailored to better suit local populations.On the other side, Prudhomme (1995) and Tanzi(1996) have argued that there exist manyimperfections in the local provision of servicesthat may prevent the realization of benefits fromdecentralization. For example, local bureaucratsmay be poorly trained and thus inefficient indelivering public goods and services. Fisman andGatti (2000) find a strong negative relationshipbetween decentralization in governmentexpenditure and corruption, that is, moredecentralized polities tend to be less corrupt.. Butas argued by Pasuk and Baker (2002), and in linewith Shleifer and Vishny’s (1993) thesis, it mayin the longer term reduce corruption by curtailingthe power of provincial ‘godfathers’ to extractrents.

Even democracy negatively affect corruption, itis a debatable issues in empirical literatures. Dopeople in developing countries utilize theirvoting power effectively? In the case ofdeveloping countries, people do not haveinformation to recognize the behavior of


38

politicians. People in developing world such asin India, Sri Lanka and Pakistan have interestedon a certain politicians and political parties eventhey have a bad history in their political life. It isnot a mater of illiteracy. Instead, educatedpeople also fall into the pork barrel politics.Politicians fill their pockets by giving a lessvalued and attractive social services to voters.Educated people are misdirected attractive shortrun policies by corrupted politicians indeveloping countries. For instances, voters of theIndian state of Tamil Nadu which is one ofleading state of India with higher literacy rateelected Miss Jayalalitha Jayaram as ChiefMinister though she had several corruption casesagainst her alleged to have been committed whenshe was in power during 1991 -1996. Anotherfeature of institutional accountability is related totransparency. Transparency depends crucially onfreedom of press and expression, and on thedegree of decentralization in the system.

Economic Freedom

Economic institutions generating a competitiveenvironment in the provision of the same publicservice tend to reduce the extraction of rents,thus reducing corruption via economic freedomwhich makes straightforward economiccompetitive mechanism. Economic freedomstimulates the competition and quality of publicgoods with minimizing corruption. When severalgovernment agencies provide exactly the sameservice, and citizens can freely choose where topurchase it hence competition among economicagencies reduce corruption. This is the case whendifferent government agencies compete byproviding substitutable or similar services,without any control over the services providedby each other. The other extreme is whendifferent government agencies providecomplementary services. This occurs, forexample, when several licenses are required fora particular activity or different levels ofgovernment legislate over the same activity. Inthis case, power is shared among differentbureaucracies that extract rents from the same

source. This institutional setup increasescorruption and the inefficiency of the system.

Sociopolitical Instability

Sociopolitical instability made by sociopoliticalcauses such as ethnic divisions and languageheterogeneity, religion, democracy, culture, ethosof people, history of nations and political systemadversely affect corruption in developingcountries. War, violation, coups, assassinationsand genocides affects corruption.

Government’s Size, Rules, Quality

of Public Administration

Increasing government size and activity expandsthe scope of public officials’ discretion andincreases the likelihood of corruption, becausegovernment intervention designed to correctmarket failures requires the use of bureaucrats tomake decisions, it will create opportunities forthese employees to be corrupt and demandbribes. The expectation is that the smaller thesize of government programs and the narrowerthe scopes of regulatory activity lead to lower thelevels of corruption. More and effectiveregulations must be minimized to reducecorruption. Common law developed fromprecedents established by judges, usually alliedwith the property-owning aristocracy against theCrown, while civil law developed from codesdrawn up by jurists at the sovereign’s bidding.Greater protections of property against the stateembodied in common law systems improvevarious aspects of government performance,including reducing corruption (Daniel Treisman,2000). Thus, one might expect countries withdifferent colonial traditions to have differentlegal cultures — and different degrees ofsusceptibility to corruption — irrespective ofwhether they have common law or civil lawsystems. Legal system and colonial experienceare, of course, highly correlated with corruption.


39

Internationalization and

Corruption

External economic factors may influence levelsof corruption. Trade and openness reduces thesize of government programs and thereforeinhibits corruption. In order to encourage tradeand investment, governments have an incentiveto control the transaction costs of corruptpractices that put their countries at adisadvantage. The international institutions thatsupport global economic integration, such as theWorld Bank and the World Trade Organization,create anticorruption pressures (Alan andHeather, 2005) Reforms and liberalization indeveloping countries has positive and negativeeffects on corruption. But recent study showsthat liberalization in politics and economicsreduce the corruption. One hand, reforms makepoliticians accountable to voters as well asintroducing more competition, which shoulddecrease corruption. On the other hand, reformsmay not be credible, which provides for anincentive for corruption. There are numerouscases of political and economic liberalizationsthat occurred in the 1980s and 1990s, which hadbeen undertaking both types of reforms in rapidsuccession leads to a decrease in corruption.Exception of china, many scholars are unable toconclude the relationship between reforms andcorruptions since china has not come fully inpolitical reforms Tomas Larsson (2006)demonstrates that three intervening factorscomparative advantage, the organization ofcorruption, and the nature of rent determines theimpact of corruption on economic performancefor different economic growth in china andRussia with same level of corruption.

While conventional wisdom has it that foreigninvestment is likely to reduce corruption becauseof the preferences of foreign investors fortransparency and no added ‘hidden costs’, andtheir greater efficiency than local competitorsburdened with such costs, In the 1990s, however,multilateral aid agencies like the World Bank andAsian Development Bank launched research and

assistance programs focused on corruption andgovernance and, following the financial crisis,foreign donors including the IMF have begun todemand bureaucratic reforms and transparencythat should reduce corruption.42 President Bushhas also expanded US foreign aid for ‘countrieswith good governance’.43 democracies,parliamentary systems, political stability, andfreedom of the press is all associated with lowercorruption. They conclude that politicalaccountability reduces corruption.

Economic Development and

Corruption

There is evidence to suggest that the relationshipbetween corruption and growth is two-waycausal affects. Corruption leads to slow down ofeconomic development. On other hand,development reduces corruption. Bureaucraticmalpractice not only influences, but is alsoinfluenced by, the level of development. In athorough and detailed study by Treisman (2000),rich countries are generally rated as having lesscorruption than poor countries, with as much as50 to 73 percent of the variations in corruptionindices being explained by variations in percapita income levels. A cursory inspection of thedata reveals that many of the most poor andcorrupt countries in the past are among the mostpoor and corrupt countries today. This conjuresup the idea of poverty traps and the notion thatsome countries may be drawn into a viciouscircle of low growth and high corruption, fromwhich there is no easy escape. Li, Xu and Zou(2000) find a weaker negative relationshipbetween corruption and growth. They also findthat ‘inequality is low when levels of corruptionare high or low, but inequality is high whencorruption is intermediate. Keith et al (2006)investigate the effects of corruption on economicdevelopment. They show the public corruptionnegative affect economic development.According to Mauro (1995), the principalmechanism through which corruption affectsgrowth is a change in private investment. An


40

improvement in the corruption index by onestandard deviation is estimated to increaseinvestment by as much as 3% of output.

Corruption has a negative effect on economicdevelopment. Economic development, throughcompensation and living standards and perhapsthrough education and information, has anegative effect on corruption. The use of thismeasure may bias the estimation of the model asa consequence of the reciprocal relationshipbetween GDP and corruption

Role of Religion, Culture, History

in Corruption

A third way in which historical tradition mightaffect the perceived costs of corrupt actions isthrough the influence of religion. This mightwork in at least two ways. Religious traditionshave often been thought to condition culturalattitudes towards social hierarchy. Where more‘hierarchical religions’ — Catholicism, EasternOrthodoxy, Islam — dominate, challenges tooffice-holders might be rarer than in cultures

shaped by more egalitarian or individualisticreligions, such as Protestantism. Religions mayalso influence how individuals view theirloyalties to family as opposed to other citizens— what Edward Banfield has called ‘familism’— which, in turn, may affect the level ofnepotism. Second pathway by which religioncould affect corruption levels is via the historicalpattern of influence that developed in differentsettings between church and state. In religioustraditions such as Protestantism, which arose insome versions as dissenting sects opposed to thestate-sponsored religion, institutions of thechurch may play a role in monitoring anddenouncing abuses by state officials. In othertraditions — such as Islam — where church andstate hierarchies are closely intertwined, such arole may be absent. As La Porta et al. (1999)have found, the percentage of Protestants in thepopulation is a robust predictor of lowercorruption. At the same time, long-lived aspectsof countries’ cultural or institutional traditionsaffect the level of perceived corruption moresignificantly than current state policies.


41

Study and year

Treisman, D., 2000

Methodology

Cross developed anddeveloping countriesanalysis with OLS and WLSmethods

Main findings

* Democracy undermines thefoundations of corruption.

* Openness to foreign trade apparentlyreduces corruption.

* Percentage of Protestants in thepopulation is a robust predictor oflower corruption.

* British colonies have significantlylower perceived corruption.

* Process of economic developmentreduces corruption

* Corruption does not necessarilyprevent growth when other factors areconducive.

Table 1: Causes of corruption and its impacts on economy

Study and year

Robert C A, 2001

Besley, T., Burgess,R., 2002

Brunetti, A., Weder,B., 2003

Persson and Tabellini(2000)

Methodology

Theoretical andmathematical proof

Time series analysis in India

Cross developed anddeveloping countriesanalysis

Cross country analysisCross developed anddeveloping countriesanalysis

Main findings

* Federal states were robustly perceivedto be more corrupt than unitary ones,

* More demands for bribes that end updriving many private actors out of themarket.

* Time length of determinants isimportant to access the actual effects

* Inverse U shape relationship betweenlevel of corruption and GINIcoefficient (inequality)

* Negative relationship betweencorruption and Regionaldecentralization.

* Democracy and press freedom doeshave positive effect on public policies

Democracy and press freedom does nothave positive effect on public policies

Democracy and press freedom does havepositive effect on corruption


42

Shyamal K.Chowdhury, 2004

Daniel Lederman etal (2005)

Alok K. et al 2004

Cross developed anddeveloping countriesanalysis

Cross developed anddeveloping countries paneldata analysis

Cross developed anddeveloping courtiers analysis

Democracy and press freedomsignificantly reduce corruption

Fiscal decentralization in governmentexpenditure is strongly andsignificantly associated with lowercorruption;

Democracies, parliamentary systems,political stability, and freedom of pressare all associated with lowercorruption.

* Federalism increases corruptionrepetitive participation

* Competitive elections increase thecontrol of corruption.


Patrick M. 2006

Hongyi Li Et al 2000

Samia Costa Tavares2007

Theoretical andmathematical proof

Cross country analysis(developing andindustrialized countries)

Cross countries analysisduring 1984-2001

Competition and Corruption is negativelyrelated.

* Corruption effects income distributionin an inverted U- shaped way

* Corruption have larger proportion ofGINI differential

* Corruption negatively affectseconomic growth.

Reforms in rapid succession leads to adecrease in corruption, while countriesthat liberalized more than 5 years afterdemocratizing experienced an increasein corruption

43

Study and year Methodology Main findings

Hypothesis and Data

According to the literatures review, this study hasfollowing hypothesis which are tested with datafrom selected developing countries.

H1: More democratic freedom will be lowercorruption in developing countries.

H2: Higher economic freedom reduce corruption

H3: Higher sociopolitical instability will behigher corruption

H4: Quality of public administration positivelyaffects reduction of corruption.

H5: Property rights and rule-based governancereduce corruption.

H6: More administrative procedure increasescorruption

H7: Higher education has more effective inreduction of corruption than literacy.

H8: Capability of information accesses haspositive affect in reduction of corruption.

H9: Poverty increases corruption.

H10.Unemployment increases corruption

H11: War increases corruption

H12: Openness reduces corruption.

H13: Corruption will be lower in moreeconomically developed countries withindeveloping countries.

This study selects 70 developing countries fromall regions in the world. Data is collected fromvarious world development reports, world factbooks and polity iv reports during 2000 to 2004.Averages of each variable of countries are usedfor estimation except some variables. Democraticfreedom is measured by index for democracy,taken from polity iv. This index takes a score fromzero to ten. Average score of period 2000-2004 isconsidered as measure of democratic freedom.Score ten means highest democratic freedom andzero means no democratic freedom. Variables ofeconomic freedom are taken from index foreconomic freedom, made by heritage foundation.This index is average of ten variables such astrade, fiscal burden, government intervention,monetary policy, foreign investment, bankingfinance, wages price, property rights, andregulation informal market. The countries withaverage overall score of 1.99 or less are freeeconomies. The countries with average overallscore of 2.00 to 2.99 are mostly free economies.The countries with average score of 3.00 to 3.99

are mostly unfree economies. The countries withaverage score of 4.00 or higher are repressedeconomies. Hong Kong is first rank with around1.5 score and North Korea is last one with 5 score.Index for economic freedom plays two roles inthis model. First, it serves as sole objectivevariables for all measures of economic freedom,second it represents an environmental and controlvariable for economic growth model by includingtrade, foreign investment, property rights andmonetary and fiscal policy etc. Average score ofperiod 2000-2004 is considered as measure ofeconomic freedom as well as economic policyenvironmental variable for economic growth. Thisstudy selects 63 developing counties (less thanUS$ 3000 per capita income in world develop-ment report of 2000) from all regions.

All causes which are root causes forsociopolitical stability have not been includedinto model; instead this study selects one solemeasure for instability. Sociopolitical instabilityis measured through the peace building capacityof selected countries based on sociopoliticalcauses(1) self determination (2) Discrimination(3) Regime types (4) Durability (5) Socialcapacity and (6) Neighborhood. . Accordingthose sociopolitical causes, Centre forinternational development conflict management(CIDCM) categorize all countries into the threecategories in peace building capacity. Frompeace and conflict global survey reports 2001,2003 and 2005, this study takes the peacebuilding capacity of the selected developingcountries during 2000-2004 as a sole measure forsociopolitical instability. From each report, zeroscore is given to more peaceful countries. Twoscores are given to more conflict countries. Onescore is given to countries which have somesociopolitical instability but they are managingsociopolitical conflict without more war. Averageof year 2001, 2003 and 2005 is taken as index ofsociopolitical instability. The index takes a valuebetween zero and two. Zero mean sociopoliticalstability (No instability) and two mean highestsociopolitical instability.(No Stability) One ismiddle level of instability.

Empirical Analysis and

Discussions

Table 2 shows the empirical results. According tothe empirical evidence from selected 70developing countries, overall democraticfreedom positively and significantly correlatedwith corruption. Like previous studies predictedmore democracy reduces corruption indeveloping countries. Within democraticfreedom, competition among political partiesmeasured by political parties representingparliament is positively correlated withcorruption. Competition among political partiesin developing countries reduces the corruption.Countries which have more political parties haveless corruption. Higher power of ruling parties inparliament increases the corruption compara-tively. Proper policy recommendation forreduction of corruption in political arena is thatmaking political competition among the politicalparties. However, this idea is debatable in case ofsome special countries. India has been morepolitical party with liberal democracy. Butcorruption also is higher. In the case ofSingapore, It is very limited democracy withlower corruption. We can not simply concludethat liberal democracy will reduce corruption.For effective functions of democracy andreduction of corruptions, efficient political andeconomic institutions are important. India and SriLanka have more political institutions but theydo not function properly due to the lack ofinformation and income for people. Democracyis necessary condition to reduce corruption but itis not sufficient condition for eradication ofcorruption.

Economic freedom affects corruption positively.Higher economic freedom is associated withlower corruption. Among the all measures,considered in this study as causes for corruptionin developing countries, the measure, economicfreedom has a strong positive correlation withcorruption (Economic freedom has 73% anddemocratic freedom has 46%). It implies thatallowing economic freedom than democratic


44

freedom is a more powerful policy strategy toreduce corruption in developing countries.Controlling other variables, following regressionsshows the effects of economic and democraticfreedom on corruption.

Corruption = f (Democratic freedom,Economic freedom)

CPI = f (DF, EF)

CPI = 5.653 +0 .005DF - 0.847EF

T value (9.26) (0.198) (5.45)

P value (0.00) (0.84) (0.00)

n = 67, R2 = 0.52

Economic freedom significantly reduces thecorruption than democratic freedom indeveloping countries. In comparing China, whyIndia has been more corruption even it has aliberal democracy which around equal to westerncountries? According our finding, the reasonsare: First, China is more economic and lessdemocratic freedom country than India. Chinahas been able charge severe penalties whoseinvolved corruptions including death penalty, butIndia can not make this penalty since its longstanding democracy. Human rights activists,institutions and international community areagainst severe punishments for corruption. Thisargument does not mean that China does nothave corruption; instead it means that politicalleaders in China have been corrected fromcorruption by non- democratic political system.Second, Even India has been a democraticcountry in text book approach, democracy in

India does not function properly. Corruptedleaders have been maintained their politicalpower via democratic means since a largenumbers of people have been living rural withoutinformation accesses to use their voting in aproductive manners. In this view, China’seconomic reforms toward the liberalizationwithout political reforms are one of reason tolower corruption than India. If China will cometo liberal democracy, it will unable to controlcorruption and economic development like Indiahas been faced. For maintaining of goodgovernance like Singapore, democracy has beenchallenge in developing countries. If developingcountries follow a blind copy of democraticpractices from western without consideringground conditions of people, they can not makenot only good government without corruptionbut also economic development. Finally, India’ssociopolitical and ethnic back ground andpeople’s ethos is completely different fromChina. However, visionary leaderships such asLee Kuan Yew in Singapore and MahathirMohamad in Malaysia are also important factorto reduce corruption with suppressingdemocracy. In Many developing countries,particularly in Africa, suppression of democracywith economic liberalization has not reducedcorruption since they have been non- visionaryleadership. A moderate democracy with liberaleconomic freedom and visionary leadership canlead to reduction of corruption in developingcountries.


45


46

Measures

1. Democratic freedom

Index for overallDemocracy

No of political partiesPower of ruling party

2. Economic freedom

Index for economic freedom

3. Sociopolitical instability

Peace Building Capacity

A Defense expenditures as% GDP

4. Government’s size,

rules, quality of public

administration

Quality of budgetary andfinancial management

Quality of publicadministration

Property rights and rule-based governance

Time required to start abusiness

Start-up procedures toregister a business

Procedures to registerproperty

Budget deficit

5. Internationalization

Openness Average time to clear

customs FDIPublic Debt

Explanation of measures

One is lower and 10 is higherNumbers in 2004

Percentage of rulingParliamentarians in 2004

One score means higherfreedom and five score means

lower freedom.

Zero score means peace andtwo scores means war

Average defense expenditures as% GDP

(1=low to 6=high)

(1=low to 6=high) in 2004

(1=low to 6=high) in 2004

(days) in 2004

Numbers

Numbers in 2004

As percentage of GDP

Export plus import divided byGDP

(days) in 2004US $ Billions in 2004

Percentage of GDP

Pearson’s

Correlations

0.468**0.356**

- 0.342**

- 0.737**

-0.486**

-0.142

-0.512**

0.527**

0. 465**

-0.198

-2.00

- 0.267*

-0.275

0.349**

-0.289*0.1430.137

No of

Observations

706670

70

71

68

46

46

46

69

69

69

45

64

536968

Table 2: Correlations between measures and corruption perception index (CPI)

(One score of CPI means higher corruption and 10 scores means lower corruption)


Sociopolitical instability negatively affectscorruption. Countries which experience hugersociopolitical conflicts, violence and war havemore corruption in comparing peace countries.The correlation between war and corruption isaround 50 percent. Quality of budgetary,financial management and public administrationnegatively related with corruption (more than50%). Property rights and rule-based governancenegative related with corruption. Government’slong administrative procedures such as timerequired to start a business, average time to clearcustoms and procedures to register propertypositively related with corruption. Moreopenness reduces the corruption in developingcountries. Both social and economicdevelopment negatively related with corruption.Per capita income, total enrollments ofeducation, literacy and capability of informationaccesses negatively correlated with corruption.But unemployment and poverty positively relatedwith corruption. For further details is annexedwith appendix 1.

Singapore is regarded as model to controlcorruption. Lim (1998) gives an officialexplanation of Singapore’s anti-corruptionstrategies that relies heavily on encouragingvalues, meritocracy, inflexible penalties andinstitutional efficiency. It has made liberalfinancial rewards available to public servantsvitiate any need for enrichment throughcorruption. For example, dominant minister’ssalaries in Singapore are pegged to those of theCEOs in the largest multinational firms in theworld. The Singapore Prime Minister’s pay isseveral times that of the United StatesPresident’s. To quote then-Prime Minister LeeKuan Yew, ‘it’s a simple choice. Pay politicalleaders the top salaries that they deserve and gethonest, clean government or underpay them andrisk the Third World disease of corruption.Haggard and Low (2002) explain the lack ofcorruption in Singapore in terms of the relativeeconomic weakness of the local private sector inthe post-colonial era. Malaysia and Singaporehave ranked among the least corrupt, despite

47

Measures

6. Development

Income levelLiteracy RateSecondary EducationTertiary Education Dependence of AgriculturePoverty

Income inequality

Unemployment

Capability of Informationaccesses

Explanation of measures

Per capita incomePercentage

Percentage of total enrollmentsPercentage of total enrollmentsPercent of agriculture in GDPPercentage of people under the

national poverty available 2000-04

Gini Coefficient available 2000-04

Unemployment available 2000-04

An index computed fromeducation, TV, Telephones and

electronic media

Pearson’s

Correlations

0.557**0.2270.2080.259*

-0.358**-0.132

0.182

-0.161

0.354**

No of

Observations

707163637164

63

55

71

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed)


48

extensive state interventions in the economy.Malaysia and Singapore have also had electedgovernments since independence, whileThailand, Philippines and Indonesia haveexperienced more corruption with long periodsof military rule followed by more recentdemocratization. The important issue in therelation to corruption and type of regimes is thatcharacteristics of leadership. No political systemin any country is immune from corruption.

Whether a country follows democracy ordictatorship is not important, the most importantfact is that leadership’s commitment to punishcorrupted leaders. It is a constant fight to keepthe house clean. As long as the core leadership isclean, any back sliding can be brought undercontrol and the house cleaned up. Constitutionalso is important for cleaning leadership.Singapore amended the constitution to have thepresident popularly elected not by parliament butby whole electorate and has a veto power on thespending of the country’s reserves by the cabinet.The president now also has the power to overruleany prime minister who stops or holds up aninvestigation for corruption against any of hisministers or senior officials or himself. TheDirector of the CPIB (Corrupt PracticesInvestigation Bureau) has two masters to backhim, the elected prime minister, and if he refusesto move, the elected president, who can actindependently of the elected prime minister, toorder that investigations proceed. The presidentalso has the veto on appointments to importantpositions like the chief justice, chief of defenseforce, commissioner of police, the attorneygeneral, auditor general and other key positionsthat uphold the integrity of the institutions ofgovernment. They are key officers, essential forthe government to function without beingsubverted. Lee Kuan Yew allowed prime misterto investigate corruption made by his wife andson in Singapore in 1995. In the case of somedeveloping countries, some politicians neverallow to investigate even his/her labors who havebeen involved in corruptions and illegalactivities. To clean up may require some key

members of the core leadership to be removed.In democratic regime in developing countries, itis impossible to remove third class politiciansfrom regime since he/she has been elected bythird class people and getting safety from firstclass politicians who depends on their supportsfor forming government In the case ofcommunist countries, cleaning up would lead toa split in the party leadership, a serious problem.The outcome of communist regime dependsupon whether the top leader is strong enough totackle other powerful leaders without disastroussplit in the political leadership or a rebellionamong party stalwarts who support the offendingleader. The fear of a collapse of the governmentmay cause the leader to hold his hand.

A very corrupt country like Indonesiaexperienced rapid economic growth for threedecades under a centralized authoritarian regime,whose nature eventually brought about its owncollapse; the subsequent democratization anddecentralization has been accompanied by lowergrowth and no apparent reduction in corruption.Democratization and decentralization in Thailandhave also not significantly reduced corruption asyet, and have been accompanied, since thefinancial crisis, by slower growth.Democratization in the Philippines may havereduced corruption but it remains high andeconomic growth low, regardless of the degree ofcentralization of power. Malaysia and Singaporehave been good economic performers with‘democratic authoritarian’ governments whichhave been elected to power for over 40 yearswithout liberalizing their political systems; ifanything, judicial independence and freedom ofthe press have declined in Malaysia in recentyears, and not changed much in an opendirection in Singapore. Perhaps because of thehigh degree of centralized state power in bothcountries, control of the state continues toprovide channels for legal capital accumulation,without necessitating the illegal acts so commonin their neighbors, and apparently withoutjeopardizing economic growth so long as opentrade and investment regimes are maintained.

Our (weak) conclusion is that while economicliberalization, democratization and centralizationof state power influence the forms of corruptionand its impact on national economicperformance, and may reduce its extent, they areneither necessary nor sufficient for its declineand disappearance. As Lee Kuan Yew statedsuccesses and failures of anti-corruptionactivities depend on leaderships and system ofthe country.

“We have to keep our own house clean. No oneelse can do it for us” (SGPR, 2005).

Conclusion

Studies on causes of the corruption have beentaken placed by various scholars. Apart fromsuch studies, this paper has investigated thesociopolitical and economic causes forcorruption in developing countries. Many factorsdetermine the corruption in developing countries.Among them, democratic freedom and economicfreedom play an important role in developingcountries. Economic freedom is more powerfuldeterminant of corruption in developingcountries. More economic freedom with limiteddemocracy leads to reduce and eradicate poverty.

Even democracy played an important positiverole in reduction of corruption in developingcountries; it is more controversial issue in aspecific case study in developing countries.Democracy is necessary condition to reducepoverty but it is not sufficient condition. Thereare many sufficient conditions such asleadership, constitutions, people’s capability forinformation accesses, sociopolitical andeconomic environment for competition, peacebased on ethnic, religious and languagehomogeneities between and within communitiesand people’s sociopolitical ethos. All anti-corruption measures are related with incomelevel of people. Increasing income of the nationdepends on various factors. Among them,leadership is playing a predominant role inreduction of corruption as well as achievingeconomic prosperity. Unfortunately, developingcountries have failed to gain visionary and cleanleaderships to overcome many sociopoliticalobstacles, particularly corruption. They have tosuffer by this harmful virus until a pragmaticleader imposes policies which are necessary andsufficient conditions for reduction of corruptionand economic development.


49

References

Alan Doig and Heather Marquette, (2005),Corruption and democratization: the litmus testof international donor agency intentions?Futures, Vol. 37, pp 199–213

Alok K. Bohara, Neil J. Mitchell, and Carl F.Mittendorff, (2004), Compound Democracyand the Control of Corruption: A Cross-Country Investigation, The Policy Studies,Journal, Vol. 32, pp. 481 -99

Ades, A., and R. Di Tella. (1999), “Rents,Competition, and Corruption.” AmericanEconomic Review 89 (September): pp. 982-93.

Bailey, J. and A. Valenzuela, (1997), The shape ofthe future. Journal of Democracy Vol. 8, pp. 43–57.

Besley, T., Burgess, R., (2002), The politicaleconomy of government responsiveness:theoryand evidence from India. Quarterly Journal ofEconomics, Vol. 117, pp. 1415–1451.

Brunetti, A., Weder, B., (2003), A free press is badnews for corruption. Journal of PublicEconomics, Vol. 87, pp. 1801–1824.

Daniel Lederman, Norman V. Loayza, and RodrigoR. Soares, (2005), Accountability andcorruption: political institutions matter,economics & politics, Blackwell publishing.

Djankov, S., C. McLiesh, T. Nenova, and A.Shleifer, (2001), Who owns the media? WorldBank Policy Research Working Paper No.2620, World Bank, Washington, DC.

Daniel Treisman, (2000), The causes of corruption:a cross-national study, Journal of PublicEconomics, Vol. 76 (2000) pp. 399–457

Ed Brown and Jonathan Cloke, (2004), NeoliberalReform, Governance andCorruption in theSouth: Assessing the International Anti-Corruption Crusade, Editorial Board ofAntipode

Haggard, S., and Low, L Haggard, S., (2002),‘State, politics, and business in Singapore’, inE.T. Gomez (ed.), Political Business in EastAsia, Routledge, London: pp. 301–23.

Hongyi Li , Xilin Colin, and Heng Fu Zou,(2000), Corruption, income distribution andgrowth, Economics and politics, pp. 155-82

Keith Blackburna, Niloy Boseb,_, M. EmranulHaque, (2006). The incidence and persistenceof corruption in economic development,Journal of Economic Dynamics & Control, pp. 2447-67

Lederman, D., Loayza, N. and Soares, R.R.,(2001), Accountability and Corruption:political institutionsmatter, Working Paper No.2708, World Bank, Washington, DC.

Linda Y.C. Lim and Aaron Stern, (2003), StatePower and Private Profit: the political economyof corruption in Southeast Asia, Faculty ofInternational Business, University of Michigan.

Lin Zhaomu, (2000), The relations among reform,development and stability in China, (WangMengkui edi) China’s economictransformations over 20 years, foreignlanguage press, Beijing.

Mya Maung (1996), The Burmese approach todevelopment: Economic growth withoutdemocratization, Journal of Asian economics,Vol. 7, No. 1, 1996, pp. 97-129.

Mauro, P., (1995), Corruption and growth,QuarterlyJournal of Economics, Vol. 110, pp. 681–712.

Montinola, Gabriella R., and Robert W. Jackman.(2002), “Sources of Corruption: A Cross-Country Study,”British Journal of PoliticalScience 32 (January): pp. 147–70.

Persson, T., Tabellini, G., (2000), PoliticalEconomics: Explaining Economic Policy. MITPress, Cambridge, MA. Sachs, J.D., Warner,A., 1995. Economic reform and the process ofglobal integration. Brookings Papers onEconomic Activity 1, pp. 1–95.

Patrick M. Emerson, (2006), Corruption,competition and democracy, Journal ofDevelopment Economics, pp. 193– 212

Robert Christian Ahlin, (2001), Corruption :political determinants and macroeconomiceffects, unpublished PhD thesis, University ofChicago.


50


51

Raymond Fisman, Roberta Gatti, (2002),Decentralization and corruption: evidenceacross Countries, Journal of Public Economics,pp. 325–345

Rose-Ackerman, S., (1999), Corruption andGovernment: Causes, Consequences, andReform (Cambridge University Press,Cambridge, UK).

Samia Costa Tavares, (2007), Do rapid politicaland trade liberalizations increase corruption?European Journal of Political Economy, Vol. 23, pp. 1053–76

Sandholtz, Wayne, and William Koetzle. (2000),“Accounting for Corruption: EconomicStructure, Democracy, and Trade,”International Studies Quarterly, Vol. 44, pp. 31–50.

Shyamal K. Chowdhury, (2004), The effect ofdemocracy and press freedom on corruption:an empirical test, Economics Letters, pp. 93–101

Singapore Government Press Release (2005),Media Relations Division, Ministry ofInformation, Communications and the Arts,Singapore.

Tomas Larsson, (2006), Reform, corruption, andgrowth: Why corruption is more devastating inRussia than in China, Communist and Post-Communist Studies, Vol 39, pp. 265-81

Treisman, D., (2000). The causes of corruption: across-national study. Journal of PublicEconomics, pp. 399–457.

World Bank, (1997), Poverty R,eduction andEconomic Management, helping countriescombat corruption: role of world bank,Washington, DC


52

Asia

1. Nepal2. India3. Indonesia4. Pakistan5. Sri Lanka6. Philippines7. Cambodia8. Bangladesh9. China10. Laos11. Thailand12. Vietnam13. Mongolia

Europe

14. Russia15. Azerbaijan16. Georgia17. Tajikistan18. Kyrgyzstan19. Turkmenistan20. Ukraine21. Uzbekistan22. Latvia23. Lithuania24. Romania25. Kazakhstan

Latin& Caribbean

26. Bolivia27. Colombia 28. Guatemala29. Peru30. Dominican31. El Salvador32. Honduras33. Jamaica34. Nicaragua35. Ecuador 36. Haiti

Middle East and

Africa

37. Algeria38. Iran39. Egypt40. Jordan41. Morocco42. Tunisia43. Burundi44. D.R.Congo45. Angola46. Eritrea47. Nigeria48. Uganda49. Central50. Rep.Congo51. Ethiopia52. Rwanda53. Benin54. Guinea55. Cameroon56. Chad57. Ghana58. Kenya59. Lesotho60. Mozambique61. Niger62. Senegal63. Tanzania64. Togo65. Zambia66. Malawi67. Mali68. Namibia69. Madagascar70. Mauritania

Appendix 1: Selected Countries

53

ABSTRACT

The purpose of this research study is to check theintrinsic factor ability utilization of jobsatisfaction associate with demographic factors:Gender, Age, Ethnicity, Civil Status, Experience,Educational Qualification and Bank Types inAmpara region government and private bankemployees.

Reliability test, Principle Component Analysis,Independent Samples t-test, ANOVA and MeanComparison test were used for analyzing thedata. The independent sample t-test result revealsthat, there is no significant different betweengender variable (Male & Female) and job abilityutilization factor, whereas there is significantdifferent between type of banks (private &government) and job ability utilization factor at5% level. Furthermore, ANOVA result concludesthat, there is no significant different between thedemographic factors: Age, Ethnicity, Civil Statusand Experience variables and job abilityutilization factor. There is significant differentbetween Educational Qualification variable andjob ability utilization factor at 10% level. Meancomparison test was used to differentiate theability utilization associated with educationalqualification variable.

Keywords: Bank employees, Job satisfaction,Ability utilization, ANOVA, Reliability test.

Introduction

Employee is one of the key factors of theorganization success. No organization cansucceed without a certain level of satisfactionand effort from its employees. Job satisfactioncan be influenced by a variety of factors. Abilityutilization is one of the intrinsic job satisfactionfactors.

In today’s competitive world, management needsto continuously emulate practices that will attractand retain a highly qualified and skilledworkforce. Work, when too difficult or easy canlead to dissatisfaction. In order to enhanceemployees’ general ability, manager should givemore opportunities for employees to solve workproblems with the full support of the bank. Thisis important as employees could develop theirability better within the bank, would build aloyalty to the bank, since they has gained goodexperience within the bank.

For the success of banking, it is very importantto manage human resource effectively and tofind whether its employees are satisfied or notworkforce of any bank is responsible to a largeextent for its productivity and profitability.According to Thakur (2007), efficient humanresource management and maintaining higher jobsatisfaction level in banks determine not only theperformance of the bank but also affect thegrowth and performance of the entire economy.

RELATIONSHIP OF DEMOGRAPHIC FACTORS TO

ABILITY UTILIZATION OF JOB SATISFACTION OF

GOVERNMENT AND PRIVATE BANK EMPLOYEES IN

AMPARA REGION – SRI LANKA

Aboobacker Jahufer

Department of Mathematical Sciences, Faculty of Applied Sciences,South Eastern University of Sri Lanka

[email protected]

M.I. Rifkhan Ahamed

Faculty of Applied Sciences, South Eastern University of Sri [email protected]


54

This research investigates how ability utilizationintrinsic job satisfaction involving bank employeesdemographic factors, in the light of currentrealities. This research paper is composed into fivesections. Section 2 derives review of literature,section 3 describes research methodology, section4 explains results and discussions and in the lastsection conclusions are given.

Review of Literature

Many researches have been conducted to analyzethe job satisfaction of bank employees withvarious issues. Some important and very recentresearch findings are:

Rahman et al., (2009) conducted research for jobsatisfaction of Bangladesh bank employees. Thisstudy found that remuneration and reward,recognition, pride in work and talent utilizationare the most important ones for improving jobsatisfaction and also, factors like job security,relation with colleagues and bureaucracy are notsignificant for job satisfaction.

Eliyana et al., (2012) conducted a research foremployees job satisfaction for productiondepartment at JRC. This study concluded thatability utilization, compensation, relationship withco-workers, working conditions, recognition andachievement simultaneously have a significanteffect on the organizational commitment.

According to Mallika and Ramesh (2010) higherjob satisfaction has been linked with employeeswho are able to exercise autonomy and withthose who have a higher level of jobinvolvement. Women have been found to reportsignificantly higher job satisfaction than menalthough this gender gap appears to benarrowing. The correlation coefficient shows apositive relationship existing among.Organizational commitment, job involvement,quality of work life, organizational climate, jobcontent, income and job satisfaction perceived bypublic and private bank employees. Researcherfound that private bank employees perceived lowlevel of job satisfaction.

Scott et al., (2012) the results of this study canonly be inferred to extension agents inMississippi. Low relationships were observedbetween gender and the job satisfactionconstructs of growth satisfaction, satisfaction withjob security and satisfaction with pay. Femalesrated all three of these constructs higher thanmales, indicating a higher level of satisfactionwith personal learning and growth opportunitiesat work, job security, and compensation.Education was not related to any of the jobsatisfaction constructs for extension agents.

According to Malik (2011) faculty members inUniversity of Balochistan were generallysatisfied with their jobs. However, male facultymembers were less satisfied than female facultymembers. This survey reveals that demographicfactors such as age, academic rank, and degreeno significant impact on job satisfaction; whichimplies that based upon age, total years teaching,and academic rank faculty are stable with regardto their overall level of job satisfaction.

According to Shafiq and Ramzan (2013) thestudy was to investigate the impact of bothpersonal and organizational variables on jobsatisfaction of employees of an industrial sectorwithin the vicinity of Lahore. t-test results showthat there is no evidence of a systematicdifference between males, females; single,married; and permanent, contract employees onjob satisfaction. However the regression analysisstudies show that income and gender aresignificant predictors of job satisfaction forindustrial employees of the studied area.

Research Methodology

Research Problem

Many research have been carried out on the topicof job satisfaction of employees in banking sectorsand the impact of various factors were seen on itwhich affected it both positively and negatively.

Employees’ ability utilization is an importantintrinsic job satisfaction factor to improve the

Relationship of Demographic Factors to Ability Utilization of Job Satisfaction ofGovernment and Private Bank Employees in Ampara Region – Sri Lanka

55

organization. But less number of research wascarried out for job satisfaction factor abilityutilization with employees’ demographic factors. Inthis research relationship of demographic factors toability utilization job satisfactions are analyzed.

Research Objectives

The main objective of this research study is toanalyze the relationship between demographicfactors (Gender, Age, Ethnicity, Civil Status,Experience, Educational Qualification and BankTypes – Government and Private) and abilityutilization job intrinsic satisfaction of bankemployees in Ampara region.

Questionnaire and Data Collection

Study area includes all employees of selectedbranches in Ampara region government andprivate banks. 180 questionnaires werepersonally administered among the respondentsbut received only 105 questionnaires whichindicate the 58.33% response of the respondents.Simple random sampling method was used in thestudy to select the sample.

Respondents provided the required informationon a structured questionnaire based on the

pertinent research objectives, classified into twosections. The first category consists ofdemographic information such as respondents’gender, age, marital status...etc. In the secondcategory consists of five-point Likert scale. Thestructure of the scale was based on the followingcategories: 1-Highly not satisfied, 2-Notsatisfied, 3-Satisfied, 4-Very satisfied and 5-Extremely satisfied.

Data Analysis

For data analysis purpose SPSS-20 was used. Thecollected likert scale data (qualitative data) wereconverted into quantitative data using principlecomponent and factor analysis for the purpose ofstatistical analysis. Reliability test, independentsample t-test, ANOVA and mean comparison testwere carried out for the converted data.

Results and Discussion

Descriptive Statistics for Demographic

Factors

Frequency distribution table for demographicfactors are given below table:

Demographic factors Frequency Percent

Gender Male 79 75.2

Female 26 24.8

Type of Bank

Private 50 47.6

Government 55 52.4

EthnicityMuslim 54 51.4

Tamil 41 39.0

Sinhalese 9 8.6

Age

below 30 62 59.0

30-35 22 21.0

35-40 7 6.7

40-45 8 7.6

45-50 3 2.9

above 50 3 2.9

Marital Status

Single 47 44.8

Married 57 54.3

Years of experience

less than 5 60 57.1

5-10 25 23.8

10-15 7 6.7

15-20 8 7.6

above 20 5 4.8

Educational Qualification

O/L 5 4.8

A/L 44 41.9

Diploma 29 27.6

Degree 19 18.1

Post

Graduate

4 3.8

Other 4 3.8

Table 4.1: Frequency distribution table for demographic factors.


56

Reliability Test

More commonly used measure of reliability isinternal consistency, which applies to theconsistency among the variables in a summatedscale. The rationale for internal consistency isthat the individual items or indicators of the scaleshould all be measuring the same construct andthus be highly inter correlated. Internal reliabilityof the instrument was checked by usingCronbach’s alpha. The generally agreed uponlower limit for Cronbach’s alpha is 0.7, althoughit may decrease to 0.6 in exploratory research(Hair et al., 2008).

The ability utilization sub variables Cronbach’salpha value is shown in table 4.2. The resultindicates that the value is more than 0.7 and lessthan 0.9, so the sub variables are reliable tomeasure the ability utilization job satisfaction ofbank employees.

Table 4.2: Reliability Statistics

Principal Component Analysis

To reduce the respondents’ responses from 5items (5 sub factors) to a one important factorwas performed using principal componentanalysis. The proportion of variance explained is0.7 criterion is used to select the number ofprincipal components retained.

Table 4.3: Eigen analysis of the Covariance

Matrix

Table 4.4: Eigen vector

Eigen values and Eigen Vectors are shown intable 4.3 and table 4.4 respectively. According tothe eigen value proportion criterion, to explainthe ability utilization of job satisfaction only firsttwo principal components are sufficient, andthese two principal components extract 76.3% oforiginal information of ability utilization ofemployees job satisfaction.

The first two principal components are: , i = 1,2, where, - eigen vector, X-original likert scaledata matrix. Hence, the first two principalcomponents are:

These two principal components are reduced toone variable using the equation

where, λ= Eigen value, n= number of factorretain.

Variable Y describes the employees abilityutilization factor and it is used for statisticalanalysis.

t-Test Results and Discussion

The independent samples t-test procedurecompares means for two groups of cases. In thisresearch, t-test is used to check whether abilityutilization differed based on variables Gender(Male and Female), Types of Bank (Governmentand Private) and Civil Status (Single andMarried).

��

��

� ��

��

� ��

�

��

��

��

Cronbach’s Alpha N of Items

0.847 5


57

Gender Variable t-Test Results

Table 4.5: t-Test Results for Gender Variable

Types of Bank Variable t-Test Results

Table 4.6: t-Test Results for Types of Bank Variable

The t-test result is shown in table 4.5. From thistable the p value is p=0.303, it indicates thatGender variable is not statistically significant. So

it can be concluded that, there is no significantdifferent between male and female abilityutilization job satisfaction.

According to the t-test results in table 4.6 theprobability value is p=0.019, this value says thatstatistically significant at 2% level. Hence it canbe concluded that, there is a significant different

between private and government bank employeesability utilization factor. Further, private bankemployees ability utilization job satisfaction ifhigher than government bank employees.

Independent Samples Test



58

Civil Status Variable t-Test Results

Table 4.7: t-Test Results for Civil Status Variable

Civil status t-test result is shown in table 4.7.According to this table the p value is p=0.224,this is not significant. So it can be concludedthat, there is no significant different betweensingle and married employees ability utilizationjob satisfaction.

One Way ANOVA and Discussion

In this research one way ANOVA is used to testfor the differences among three or more means

of sub variables of main variables such as yearsof experience, educational qualification, ethnicityand age to check the ability utilization factor.

Years of Experience Variable ANOVA Results

Variable years of experience is categories intofive levels (see table 4.1) and one way ANOVAresults for this variable is given in table 4.8.

According to the p value in the above tablep=0.691, this is not significant. So it can beconcluded that, years of experience all categories are same with ability utilization jobsatisfaction.

Educational Qualification Variable ANOVA

Results

Variable educational qualification is categoriesinto six levels (see table 4.1) and one wayANOVA results for this variable is given in table 4.9.

Table 4.8: ANOVA Results for Years of Experience Variable


ANOVA Sources of Variance Sum of

Squares df Mean

Square F Sig.

Between Groups 4.221 4 1.055 .562 .691 Within Groups 187.694 100 1.877 Total 191.915 104


59

From the above table p value is p=0.066, this issignificant at 10% level. So it can be concludedthat, at 10% significant level at least oneeducational qualification categories aresignificantly different with ability utilization ofjob satisfaction. Hence mean seperation test iscarried out to check the different.

Mean Separation Analysis for Educational

Qualification Variable

In the above ANOVA table 4.9 no information isavailable to say which category is different fromothers. Therefore a mean separation is to befollowed to find out which educational

qualification category is different from others. Somean separation test is used to find whichcategory is different. According to the meanseparation test results (O/L and A/L), (O/L andDegree), (O/L and Post Graduate), (A/L and PostGraduate), (Diploma and Post Graduate),(Degree and Post Graduate), (Others and PostGraduate) are different ability utilization jobsatisfaction.

Age Variable ANOVA Results

Variable age is categories into six levels (seetable 4.1) and one way ANOVA results for thisvariable is given in table 4.10.

According to the p value p=0.959 in the abovetable, this is not significant. So it can beconcluded that, age all categories are same withability utilization job satisfaction.

4.7 Ethnicity ANOVA Results

Variable ethnicity is categories into three levels(see table 4.1) and one way ANOVA results forthis variable is given in table 4.11.

Table 4.10: ANOVA Results for Age Variable

Table 4.9: ANOVA Results for Educational Qualification Variable

ANOVA Sum of

Squares

df Mean

Square

F Sig.

Between Groups 18.806 5 3.761 2.151 .066

Within Groups 173.109 99 1.749 Total 191.915 104

ANOVA Sum of

Squares

df Mean Square F Sig.

Between Groups 1.972 5 .394 .206 .959

Within Groups 189.943 99 1.919

401 519.191 latoT


60

According to the p value p=0.277 in the abovetable, it can be concluded that this is notsignificant. So ethnicity all categories are samewith ability utilization job satisfaction.

Conclusions

The purpose of the study is to identify therelationship between ability utilization jobsatisfaction and demographic factors in Ampararegion government and private bank employees.The t-test result concludes that, there is nosignificant different between male and femaleemployees ability utilization, single and marriedemployees ability utilization, whereas, there issignificant different between private andgovernment bank employees’ ability utilizationjob satisfaction. Moreover, ability utilization jobsatisfaction of private bank employees is higherthan government bank employees in Ampararegion.

ANOVA results reveal and concluded that, (i)there is no significant different between years ofexperience of employees ability utilization jobsatisfaction (ii) there is no significant differentbetween all age level of employees abilityutilization job satisfaction and (iii) there is nosignificant different between all ethnicityemployees ability utilization job satisfaction. But,there is significant different between employeeseducational qualification level ability utilizationjob satisfaction. Hence, qualification wise theability utilization job satisfaction of employees issignificantly different in bank employees inAmpara region.

Table 4.11: ANOVA Results for Ethnicity Variable

ANOVA Sum of

Squares

df Mean Square F Sig.

Between Groups 4.783 2 2.391 1.301 .277

Within Groups 185.647 101 1.838

301 034.091 latoT


61

References

Eliyana, A., Yusuf, R.M. and Prabowo, K. (2012),The Influence of Employee’s Job SatisfactionFactors on Organizational Commitment.American Journal of Economics, Special Issue.pp. 141-144.

Hair, J.F., Anderson, R.E., Tatham, R.L. and Black,W.C. (2008), Multivariate Data Analysis, 6th

Edition, Low Price Edition, Pearson Education.

Malik, N. (2011), A Study on Job SatisfactionFactors of Faculty Members at the Universityof Balochistan, Journal of Research inEducation, 21.

Mallika N. and Ramesh M. (2010), Job Satisfactionin Banking: A Study of Private and PublicSector Banks. International Journal ofManagement (IJM). pp. 111-129.

Rahman, M.I., Gurung, H.B. and Saha, S. (2009),Job Satisfaction of Bank Employees inBangladesh: An Analysis of SatisfactionFactors; Daffodil International University,Journal of Business and Economics, Vol. 4,No.1 & 2).

Scott, M., Swortzel, K.A. and Taylor, W.N. (2005),The Relationships between SelectedDemographic Factors and the Level of JobSatisfaction of Extension Agents; Journal ofSouthern Agricultural Education Researc, 55.

Shafiq, A. and Ramzan, M. (2013), Determinantsof Job Satisfaction amongst Industrial Workers.The International Research Journal ofCommerce and Behavioral Science, Vol. 2, No. 10.

Thakur, M. (2007), Job satisfaction in banking: Astudy of private and public sector banks. TheICFAI Journal of Bank Management, Vol. 6,No. 4, pp. 60–68.

62

ABSTRACT

The experiences of a customer during thepurchase process definitely determine the level ofsatisfaction/dissatisfaction for the product orservice. The towering competition the marketplace persuaded the marketers for managing thecustomer interaction even before making theirdecisions. This view would be more significantin the case of high involvement products such asautomobiles. Automobile industry showsextensive competition and every player comeswith innovative strategies for acquiring as wellas retaining its customer base. Recent marketingthinkers suggested that customers are the mostinfluential advertising weapon in the modernworld because they have to negotiate the productwith new customers. They might feel that theyare not cheated by anyone in the sales channeland should get enough supports even after thepurchase of the product. This study becomesimportant as customers evaluate their postpurchase experiences as a criterion forrecommending an automobile to their near anddear. The data were collected among thecustomers of medium segment cars in Keralaand the results of the study states that customerspurchase experiences significantly influencespost purchase customer loyalty.

Key words: Purchase Experiences, CustomerLoyalty

Introduction

The experiences of customer during the purchaseprocess have a definite role in determining theirlevel of satisfaction. Earlier studies in consumerbehavior stated that customers’ experiences with

the product during the usage and the after salesservices extended by the service peopledetermine their level of satisfaction. The postpurchase experiences consists all possibleencounters between the customer and marketerafter the actual purchase of the car and they areconsidered as the antecedents and loyaltyconsidered as the consequences of customersatisfaction. The automobile consumers arehighly involved for their purchase and theirevaluation during the post purchase of the carbecomes extremely significant. . Through thisstudy, the researcher explored the influences ofpost purchase experiences towards customersatisfaction and its impact on customer loyalty.The results of the study states that certain itemsunder post purchase experiences significantlyinfluences customer satisfaction that determinescustomers loyalty.

Statement of the Problem

Marketing intermediaries generally perform asignificant role between the manufacturer andconsumers. It was identified that major chunk ofthe resources in any channel were projected themarketing of goods and services to the ultimateconsumer group. The customers’ experiences atthe suppliers’ premises has been given moreimportant and least attention was paid after thecustomer leaving from the shopping premises. Itcan be noted in the durable industry, especiallyin the automobile industry, where customersinvolvement is very high, sufficient attentionshould be given to the consumers’ experiencesafter the sale of their products. Severalautomobile companies are seldom given suchattention that leads to diminishing market sharesfor several players. The current study was

IMPACT OF POST PURCHASE EXPERIENCES ON

CUSTOMER LOYALTY: AN EMPIRICAL INVESTIGATION

V.K. Hamza

Assistant Professor, MES - Advanced Institute of Management and Technology, Marampally, Aluva, Ernakulam.E mail: [email protected]

Impact of Post Purchase Experiences on Customer Loyalty:An Empirical Investigation

63

addressed the low responsibility of dealerstowards customers during their post-purchaseexperiences stage.

Research Questions

It has been observed that the channel membersor dealers of automobile industry was leastconsidered the relationship with the consumersafter the purchase. They offer severalinfluencers for persuading the customers to buytheir product but not much of effective in theirrelation after the purchase of the automobile.The research question studied through this paperwas the identification of the importance of postpurchase experiences of customers towards thegeneration of loyalty.

Objectives of the Study

1. To identify the role of post purchaseexperiences for the generation of customerloyalty

2. To identify the major items during thepost purchase experience stage thatdetermine customer loyalty

3. To explore the intervening role ofcustomer satisfaction between the postpurchase experience and customer loyalty

Post Purchase Customer Loyalty

Post Purchase Customer Loyalty simply meansthat the bonding and attachment of a customertowards a specific product after experiencing thevalue generated from them. Numerous studieshave reported that customer loyalty is animportant outcome of customer satisfaction(Oliver, 1999, Lee et al., 2001, Fornell, 1992). Itis a desire to continue the relationship with anorganization that can be viewed through repeatpatronage of a customer (Neal 1999, Czepiel &Gilmore 1987). Earlier studies imply theimportance of loyalty for the survival of businessorganization that highlights companies’ should

work for beyond customer satisfaction (Reicheld& Schefter 2000, Anderson & Mittal 2000).There are several dimensions for customerloyalty and most of the researchers discussed thebehavioral as well as attitudinal evaluations(Jacoby & Kyner 1973). The behavioraldimension conceptualizes the repeat purchasingbehavior and the attitudinal dimension focusesthe future intention to purchase from the sameshop that reflects in the cognitive and emotionalattachment (Jacoby & Chestnut 1978, Dick &Basu 1994). Loyal customers are treated as thereal asset of any organization that should bemaintained like any other tangible asset becausesuch customers may have resistance to switching(Cronin & Taylor 1992) and ready to paypremium (Zeithaml et al 1996) and recommendthe brand to peers (Feick et al 2001) andminimize the evaluation of alternatives duringthe purchase (Reichheld 1996). For the currentstudy, the researcher used the standardized scaledeveloped by Harris and Goode (2004) for themeasurement of action based loyalty. This scaleconsists of four items based on seven-pointLikert type response intended to measure thedegree to which a person express his intention tocontinue with the same company/product (Oliver1997).

Customer Satisfaction

Customer satisfaction simply means anevaluation of the surprise inherent in a productacquisition and/or consumption experience. It isan evaluation rendered that the experience was atleast as good as it was supposed to be (Hunt1977, p. 459-460). The ultimate evaluation of thecustomer could include the performance of thereal product (Anderson & Mittal, 2000). It isclear that market and consumer segments are theimportant factors to consider while measuringcustomer satisfaction (Anderson & Mittal 2000,Mittal & Kamakura 2001) and valuable fromboth customer goodwill perspective andorganization’s financial perspective (Malthouseet al 2003). Consumer satisfaction is an ongoingprocess and it never ends (Peters & Waterman


64

1982) that leads positive consequences such asre-patronage intention and loyalty (Anderson etal 1997, 1994, Reichheld & Sasser 1990). Aconsumer can be both satisfied and dissatisfiedwith different aspects of the same product andthus can have complaints and appraisals for thesame (Mittal 1998, Chan et al 2003). If anorganization focuses on its most demandingcustomers and/or most demanding marketexpectations, it is likely to exceed all othercustomers and expectations (Sheth & Kellstadt1992). It is the marketers’ duty to recognize themost influential aspects determining customers’satisfaction and through this study, a list of postpurchase encounters and its impact on customersatisfaction and loyalty were hypothesized.

Purchase Experience

The expectancy confirmation paradigm becamean important criterion for customers purchaseexperiences evaluation (Khalifa & Liu 2003,LaTour & Peat, 1979). Customer approachesany product or organization with certainexpectations and they started to evaluate theirexpectation with actual performance or encounter(Cohen & Goldberg 1970, Olshavsky & Miller1972). On the basis of these comparisons, thecustomer decided himself whether he is satisfiedor not. So the post purchase experienceevaluation becomes an important determinant orantecedent of customer satisfaction thatultimately determine customer loyalty (Ulaga2001, Bower & Garda 1985, Jones & Sasser,1995).

The marketer’s job does not end with the sale ofthe product but continue even after and shouldmonitor purchase evaluation of their customers.Sometime customer may rethink about theforegone alternative considered beforepurchasing the existing product (Sugden 1985),or rethink about the other possibilities that couldhave been chosen (Zeelenberg & Pieters 2006).Customers’ affective mental status alsodetermines their satisfaction evaluations.Holbrook and Zirlin (1985) states that sensoryexperiences during the purchase process

influence the customer significantly and theyimpact directly to customer emotions (Pham2004). Likewise shopping enjoyment influencespositive behavior changes in customers(Jarvenppa and Todd 1997, Novak et al 2000,Koufaris 2002) that motivates them to purchases(Dennis, Newman & Marsland 2005, Batra &Ahtola 1991).

The determinants of customers’ evaluationdefinitely affect customer satisfaction andthereby customer loyalty, a clear understandingof such determinants becomes highly importantfor effective marketing of products. Theconsumers’ behavior is highly sensitive tocontext and cultures, so the researcher developedthe items under post purchase experiencesthrough in-depth interview.

Methodology

The population for the study consists of mediumsegment car users in the state of Kerala. As perthe wordings from the Transport Department, itis registered around 42000-48000 mediumsegment cars were registered in Kerala during2013. The medium segment car means any carwhose basic model’s cost lay between 4.5 lakhsto 6.5 lakhs Indian rupee. The researcherassumed that criteria under post purchaseexperience evaluations are context specific andvarying as per the culture and geographicalregion, a list of such items are developed throughin-depth interview with selected automobilescustomers and based on content analysis, a list of27 criteria was developed as the items under postpurchase experiences. An instrument wasdeveloped consisting of the items under postpurchase experiences, scale for measuringcustomer satisfaction and customer loyalty, andused the same for data collection. Initialreliability and validity of the instrument hasensured before starting the detailed datacollection. About 300 questionnaires weredistributed and among them, 228 questionnairesare filled completely and used for further dataanalysis.


65

Item Generation and Validation ofthe Instrument

The instrument consists of three scales such asPost Purchase Experiences (PPE), CustomerSatisfaction and Loyalty. The PPE scale wasdeveloped by the researcher and the remainingwas adapted from standardized scales. Thedetails of the instrument development werestated below:

The items under Post Purchase Experiences(PPE) were developed through in-depthinterview and a content analysis was followed.The same kind of items giving identical meaningwere removed and based on the ‘item to totalstatistics’, it has shortlisted 27 items under PPE.Three factors were generated throughconfirmatory factor analysis that represents theentire 27 item and named as Experiences withCar (7 items), Experiences with Service Station(12 items) and Experiences with Dealer (7items). The following figure shows thehypothesized confirmatory representation ofvarious items under each factor of PPE.

Figure : Confirmatory Factor Analysis

The items in the PPE scale has undergone forface and content validity and undergone essentialrevisions and rewordings before the final datacollection. The reliability coefficients of the PPEscale shows an acceptable statistical reliabilityfor the instrument as per the literature such asCronbach’s Alpha (0.715), Spearman-BrownCoefficient (0.765) and Guttman Split-HalfCoefficient (0.812).

Table 1 : Reliability Statistics (Scale-I)

Below mentioned are the list of items under PPEscale and their respect latent variables. There are27 measurement items and 3 latent variablesrepresenting each items was listed for the studyand details are given below.

Value .761N of Items 14a

Value .785N of Items 13b

27.715.799.765

Cronbach's Alpha .715.812

Spearman-Brown Coefficient

Equal LengthUnequal Length

Guttman Split-Half Coefficient

y ( )Cronbach's Alpha Part 1

Part 2

Total N of ItemsCorrelation Between Forms


66

�

Sl.No. Measurement Variable Latent Variable1 Reliability About the car2 Riding comfort About the car3 Safety features About the car4 Future resale value About the car5 Working condition of the car About the car6 Ease of handling About the car7 Overall opinion About the car8 Promptness in attention About the dealer9 Facilities available at the office About the dealer10 Explanation of car’s features About the dealer11 Feeling of pricing practices of seller About the dealer12 Vehicle display About the dealer13 Speedy arrangement of finance and other documents About the dealer14 Overall opinion About the dealer15 Explanation of problems and repairs About service station16 Cordiality, friendliness and trustworthiness About service station17 Quality of work done About service station18 Response to telephone calls About service station19 Facilities at the waiting lounge About service station20 Facilities to watch the maintenance work About service station21 Fairness of service charges About service station22 General handling of cars About service station23 Service employees caring during complains, if any About service station24 Service personnel’s responsiveness during complains, if any About service station25 Service personnel’s knowledge for servicing the product About service station26 Waiting time for service or appointments About service station27 Responses to unawareness of the customers by the sales people About service station

Items in PPE Scale

in

No..lS1 libaabilRe2 cg ndiiR3 fyteffeaS4 reurutF5 nkirrkoW6 foesaE7 llareOv8 ptmmporP9 eitilicFa

�

in

eliabVartnemeruaseMytil

troffomocseurtaateffe

uelavelaserracehtfonoitdinocg n

gnidlnahfnoinpio

noitnettaatnisseneciffifffoehttaatelbaabliavase

alcSEPPinsmetI

�

eliabVartneatL

racehtutobAracehtutobAracehtutobAracehtutobAracehtutobAracehtutobAracehtutobA

reladeehtutobAreladeehtutobA

eal

�

e

�

ir

910 anaplxE11 g nileeF12 elcihVe13 dyepeS14 llareOv15 anaplxE16 ladiroC17 ytilQua18 nposeR19 eitilicaF20 eitilicaF21 ssenriaF22 larenGe23 ecivreS24 ecivreS25 ecivreS

�

in

infr

seurtaateffes’raarcfonoitarellesfosecitcaprg niciprfo

yaayplsdiestneumcdorehtod naecnaniffifotnntemgenaanrraar

noinpiosripaerd naansmelboprfonoita

ssenihtrwotusrtd naanssenidlneirf,ytiitendok rwofo

sllacenopheletotesngeunolg nitiwaehttase

krwoecnaannetniamehthctwaotsesgerahcecivresfos

sracfog nidlnahlynaanfi,sniaplmmpocg nidurg niraarcseeyoplmmpe

sniaplmmpocg nidurssenevisnposers’lennosrpeducoprehtg nicivresroffodgeewlokns’lennosrpe

�

reladeehtutobAreladeehtutobAreladeehtutobAreladeehtutobAreladeehtutobA

itatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobAitatsecivrrvesutobA

ynafi,s itatsecivrrvesutobAtduc itatsecivrrvesutobA

�

noinoinoinoinoinoinoinoinoinoinoi

�

25 ecivreS26 g nitiaW27 nposeR �

inducoprehtg nicivresroffodgeewloknslennosrpe

stnemtnippoaroecivrrvesroffoemitg asehtybsremotuscehtfossenerwaaraunotsesn �

tduc itatsecivrrvesutobAitatsecivrrvesutobA

eplopesela itatsecivrrvesutobA �

noinoinoi �

�

�

�

�

�

�

�

�

�

�

�

�

�

�

Table 2 : Items under PPE

Table 3: Reliability Statistics (Scale II)

Value� .861�N of Items� 2�a�Value� .822�N of Items� 1�b�

3�.867�.869�.836�

Cronbach's Alpha� .845�.873�

Correlation Between Forms�Spearman-BrownCoefficient� Equal Length�

Unequal Length�Guttman Split-Half Coefficient�

Re liab ility Stat ist ics (Sca le II: Sat isfact ion) �Cronbach's Alpha� Part 1�

Part 2�Total N of Items�

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

� � � � ��

� ��

� �

� � �

� �

The items used for measuring customersatisfaction with the car (Scale II) were based onthe standardized scale developed by Gaski &Etgar (1986). The scale was adapted for thecurrent study, and the reliability and validitywere ensured before data analysis. Thereliability coefficients are greater than 0.7 andconsidered for further analysis (Gaur & Gaur,2008).

For the current study, the researcher used thestandardized scale developed by Harris andGoode (2004) for the measurement of actionbased loyalty (Scale III). This scale consists offour items based on seven-point Likert typeresponse intended to measure the degree towhich a person express his intention to continuewith the same company/product (Oliver 1997).The items were adapted for the current study and


67

the reliability and validity were ensured beforefurther analysis.

Table 4 : Reliability Statistics (Scale III)

The items in the scales were measured with 7point from highly disagree to highly agree. Allthe items in the scale show an acceptablereliability and validity for the entire measuringinstrument. So the researcher considered thedata for further analysis.

Customer Satisfaction as an

Intervening Variable

An intervening variable is a qualitative orquantitative variable that links the relationshipbetween an independent variable and dependentvariable (Barron and Kenney 1986, Mohr et al2005, Cooper et al 1990). Mediation means theeffect of an independent variable on a dependentvariable is transmitted through a third variable(Alwin & Hauser, 1975) called mediatingvariable, and for this study, customer satisfactionhas considered as the mediating variable (Barronand Kenney, 1986). Several studies havereported the customer satisfaction as a mediatorbetween the antecedents and consequences(Anderson and Sullivan 1993, Fornell 1992,Oliver 1980, Bolton and Drew 1991, Szymanskiand Henard 2001) and the researcher interestedto study its role in the context of the currentstudy.

Hypothesis of the Study

H1: There is a positive relationship betweenFavorable Experiences with the Car andCustomer Satisfaction.

H2: There is a positive relationship betweenFavorable Experiences with the Service Stationand Customer Satisfaction.

H3: There is a positive relationship betweenFavorable Experiences with the Dealer andCustomer Satisfaction.

H4: There is a positive relationship betweenCustomer Satisfaction and Customer Loyalty.

H5: Customer Satisfaction mediates PostPurchase Experiences (PPE) towards CustomerLoyalty.

H5a: Customer Satisfaction mediatesExperiences with the Car towards CustomerLoyalty.

H5b: Customer Satisfaction mediatesExperiences with the Service Station towardsCustomer Loyalty.

H5c: Customer Satisfaction mediates Experiencewith Dealer towards Customer Loyalty.

Value� .728�N of Items� 2�a�Value� .769�N of Items� 2�b�

4�.796�.745�.745�

Cronbach's Alpha .708�.728�Guttman Split-Half Coefficient�

Cronbach's Alpha Part 1�Part 2�Total N of Items�Correlation Between Forms�

Spearman-BrownCoefficient� Equal Length�

Unequal Length�

Re liab ility Stat ist ics (Sca le III : Loya lty) ��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

��

��

� � � ��

�

�


68

Data Analysis

The following table shows the results of theinitial 4 hypothesis. The first hypothesis explainsthat Experiences with the Car positivelyinfluences Customer Satisfaction (b=0. 281,t=3.06) and proved through the study. Theresults of the second hypothesis shows thatExperiences with the Service Station influencesCustomer Satisfaction (b=0.155, t=2.06) and thethird hypothesis stated to measure the influencesof Experiences with the Dealer towardsCustomer Satisfaction also significant (b=0.406,t=2.47). The influences of customer satisfactiontowards customer loyalty also proved (b=0.339,t=2.94) and accepted the alternative hypothesis.Altogether, the data analysis shows that PostPurchase Experiences significantly influencescustomer satisfaction that leads to customerloyalty.

The hypothesis 5 has been stated to test themediation role of customer satisfaction betweenthe Purchase Evaluation and Customer Loyalty.The PPE consists of three dimensions and themediation role of customer satisfaction betweenindividual dimensions towards customer loyaltyhas been tested. Sobel (1982) gives anapproximate test of significance for the indirecteffect of the independent variable on thedependent variable through the mediator. Thistest is based on ‘t’ statistic that determinewhether the changes in the effect of theindependent variable such as Experiences withthe car, Experiences with the service station andExperiences with the dealer, shows a significantchange and therefore whether the mediationeffect is statistically significant towards customerloyalty.

Table 5 : Results of the Analysis

Figure 2 : Structural Model

AVE1 AVE 2 r r2 AVE1>r2 AVE2>r2 Discr. T Stat.Exp. with the Car --> Cust.Sati. 0.69553 0.72605 0.6258 0.39163 sig. sig. yes ��Exp. with Ser. Stan. --> Cust.Sati. 0.70316 0.73368 0.5124 0.26255 sig. sig. yes ��Exp. with the Dealer --> Cust.Sati. 0.71079 0.74131 0.4869 0.23707 sig. sig. yes ��Cust.Sati.--> Cust.Loyalty. 0.71842 0.74894 0.6521 0.42523 sig. sig. yes ��

DISCRIMINANT VALIDITY & SIGNIFICANCE OF THE HYPOTHESIZED PATH


69

Baron and Kenney (1986) suggested that avariable is function as a mediator when it meetsthe conditions such as variation in levels of theindependent variable significantly account forvariation in the presumed mediator, variations inthe mediator significantly accounts for variationsin the dependent variable, and when themediation path is controlled, a previouslysignificant relation between the independent anddependent variable is no longer significant.

The following figure explains the results ofhypothesis 5. The direct influences of Exp. withthe Car towards loyalty show beta value of0.4215 with a significant t statistic of 3.258.After introducing the customer satisfaction as amediator, the direct influences between Exp. withthe Car towards loyalty become insignificantwith beta value of -0.105 and t statistic of 0.351.Likewise, the influences of Exp. with the Service

Station towards loyalty show beta coefficient of0.4072 with a significant t statistic of 1.991.After introducing the mediator variable ofcustomer satisfaction, the direct influences ofExp. with the Service Station towards loyalty

becomes insignificant with beta value of 0.2351and t statistic of 1.004. In the case ofExperiences with Dealer towards Loyalty, thecustomer satisfaction plays a significant role as amediator. The direct relation between Exp. WithDlr. towards loyalty is significant with a betacoefficient of 0.1083 with t statistic of 2.553. Ifthe customer satisfaction has introduced as amediator, the original relationship between Exp.with Dealer towards Loyalty becomesinsignificant. The result of the Sobel test alsoshows that the mediation path in thehypothesized influence is significant as thecoefficient is greater 1.96.

On the basis of the test result and analysis, it canbe concluded that PPE significantly influencescustomer satisfaction that leads to customerloyalty. The occurrence of loyalty is based onthe extent of customer satisfaction. It can also be

noted that purchase evaluation doesn’t influencesor generate loyalty but affects customersatisfaction. As per the literature, it can be seenthat satisfied customers becomes loyal to theorganization.

�

Items Coefficients t Stat. Items Coefficients t Stat. Items Coefficients t Stat.

Exp.Car-->CL -0.105 6.351 Exp.SS-->CL 0.4072 1.991 Exp.Dlr.-->CL 0.1083 2.553

Exp.Car-->CL(M) �� Exp.SS-->CL(M) �� 1.004 Exp.Dlr.-->CL(M) 0.0754 0.865

Exp.Car-->CS 0.3125 Exp.SS-->CS 0.5063 Exp.Dlr.-->CS 0.4061

CS-->CL 0.3391 CS-->CL 0.5377 CS-->CL 0.3394

Exp.Car-->CS (SE) 0.0341 Exp.SS-->CS (SE) 0.0341 Exp.Dlr.-->CS (SE 0.0468

CS-->CL (SE) 0.1547 CS-->CL (SE) 0.1834 CS-->CL (SE) 0.1458

Sobel statistic: Sobel statistic: Sobel statistic:

Probability Probability Probability

2.24834822

0.02455399

H5(b) H5 ( C)H5(a)

2.13184943

0.03301923

2.87630294

0.00402363

Figure 3 : Results of the Hypothesis (Mediation Effect)


70

Discussion and Conclusion

This study has discussed the influences ofcustomers purchase evaluations with respect tocar, to the dealer and to the service stationtowards customer satisfaction and its impact oncustomer loyalty. The result shows that purchaseevaluations significantly determine customers’satisfaction. A positive and favorable evaluationby the customer about their post purchaseexperiences with respect to their car purchasedefinitely ensures customer satisfaction. Thisstudy conceptualizes that purchase evaluationsdoesn’t generate loyal customers but onlysatisfied customers are more tend to becomeloyal to the firm. This theoretical observationsuggests that post purchase evaluation doesn’tinfluences customer loyalty but determinecustomer satisfaction that mediates the influencetowards loyalty.

The results of the study are highly significant tothe automobile industry as it forms the base ofcustomer retention. The responsibility of themarketers will not end by merely selling the carbut proper maintenance and post purchasemonitoring should be ensured. The extensivecompetition and entry of new players in theautomobile industry persuaded the marketers forinventing new strategies. This study upholds theimportance of post purchase experiences ofcustomer as it becomes the determinants of theirsatisfaction evaluation. As far as a marketer isconcerned, a clear knowledge about the variousencounters that each customers faces after thepurchase of their car help to make them satisfiedcustomers and by the way loyal to the productand firm. So the marketers of medium segmentautomobile cars should consider the above statedencounters under PPE as the means to ensurecustomer satisfaction and loyalty.

Further Research Scope

This study conceptualizes the intervening role ofcustomer satisfaction between post purchaseexperiences and customer loyalty. It states thatthe influences can be possible only throughcustomer satisfaction and not supported thedirect relation. The intervention of otherpsychological variables such as customer valueperception, expectation, etc. can also play asignificant role for setting satisfaction level.These influences can be studied in order to getmore insight to the theoretical relations. In sucha way, the research topic can be taken to a newheight with other industry of varying nature asthe customer evaluation process become differentthat depends up on the industry.


71

References

Alwin, D. F., & Hauser, R. M. (1975), TheDecomposition of Effects in Path Analysis.American Sociological Review, Vol. 40, pp. 37-47.

Anderson, E.W. & Sullivan, M.W. (1993), Theantecedents and consequences of customersatisfaction for the firms. Marketing Science,Vol. 12, pp. 125–143.

Anderson, E.W., & Mittal, V. (2000). Strengtheningthe satisfaction–profit chain. Journal of ServiceResearch, Vol. 3, pp. 107–120.

Anderson, E.W., Fornell, C. and Rust, R.T., (1997),Customer satisfaction, productivity andprofitability: differences between goods andservices. Marketing Science, Vol. 16, No. 2,pp. 129-45.

Anderson, E.W., Fornell, C., & Lehmann, D.R.,(1994), Customer Satisfaction, Market Share,and Profitability: Findings from Sweden.Journal of Marketing, Vol. 58 (7), pp. 53-66

Baron, R. M., & Kenny, D. A. (1986), Themoderator-mediator variable distinction insocial psychological research: Conceptual,strategic, and statistical considerations. Journalof Personality and Social Psychology, Vol. 51,pp. 1173-1182.

Batra, R., & Ahtola, O. (1991), Measuring thehedonic and utilitarian sources of consumerattitudes. Marketing Letters, Vol. 2 (4), pp. 159-170.

Bolton, R. N. & Drew, J. H. (1991), A MultistageModel of Customers ́ Assessment of ServiceQuality and Value. Journal of ConsumerResearch, Vol. 54 (4), pp. 69 - 82

Bower, M., & Garda, R.A. (1985), The role ofmarketing in management. The McKinseyQuarterly (3), pp. 34-46.

Chan, L.K., Hui, Y.V., Lo, H.P., Tse, S.K., Tso,G.K.F. & Wu, M.L. (2001), Consumersatisfaction index: new practice and findings.European Journal of Marketing, Vol. 37 (5),pp. 872-909.

Cohen. J. B., & Goldberg, M.E. (1970), TheDissonance Model in Post-Decision ProductEvaluation. Journal of Marketing Research,Vol. 1 (8), pp. 315-321.

Cooper, M.L., Russell, M. & Andfrone, M.R.(1990), Work stress and alcohol effects: A testof stress-induced drinking. Journal of Healthand Social Behavior, Vol. 31, pp. 260–276.

Cronin, J.J. & Taylor, S.A. (1992), Measuringservice quality: a reexamination and extension.Journal of Marketing. Vol. 56(7), pp. 55-68.

Czepiel, J. A. & Gilmore, R. (1987), Exploring theConcept of Loyalty in Services, in The ServicesMarketing Challenge: Integrating forCompetitive Advantage, J. A. Czepiel, C. A.Congram, and J. Shanahan, eds. Chicago, IL:AMA, pp. 91-94.

Dennis, C., Newman, A. & Marsland, D. (2005),Objects of Desire: Consumer Behaviour inShopping Centre Choices, Basingstoke:Palgrave MacMillan.

Dick, A. & Basu, K. (1994), Customer loyalty:towards an integrated framework. Journal ofthe Academy of Marketing Science, Vol. 22 (2),pp. 99-113.

Feick, L. & Lee, J. (2001), The impact of switchingcost on the customer satisfaction- loyalty link;mobile phone service in France. Journal ofService Marketing, Vol. 15(1), pp. 35-48.

Fornell, C. (1992), A national customer satisfactionbarometer: The Swedish experience. Journal ofMarketing, Vol. 56, pp. 6–21.

Gaski, J. F. & Etzel, M. J. (1986), The Index ofConsumer Sentiment Toward Marketing.Journal of Marketing, Vol. 50 (7), pp. 71-81.

Harris, L. C. & Goode, M.M. H. (2004), The fourlevels of loyalty and the pivotal role of trust: Astudy of online service dynamics. Journal ofRetailing, Vol. 80(2), pp. 139-58.

Holbrook, M. B. & Zirlin, R. B. (1985), Artisticcreation, artworks, and aesthetic appreciation.Advances in Non-Profit Marketing. Vol. 1, pp. 1-54.


72

Hunt, H.K. (1977), Consumer satisfaction anddissatisfaction: perspectives and overview. inHunt, H.K. (Ed.), Conceptualization andMeasurement of Consumer Satisfaction andDissatisfaction, Marketing Science Institute,Cambridge, pp. 455-88.

Jacoby, J. & Kyner, D. B. (1973), Brand LoyaltyVs. Repeat Purchasing Behavior. Journal ofMarketing Research, Vol. 10 (1), pp. 1-9.

Jacoby, J., & Chestnut, R. (1978). Brand loyaltymeasurement and management. New York:Wiley.

Jarvenpaa, S. L. & Todd, P. A. (1997), ConsumerReactions To Electronic Shopping on theWorld Wide Web. International Journal ofElectronic Commerce. Vol. 2, pp. 59-88.

Jones, T. O., & Sasser, W. E., Jr. (1995), Whysatisfied customers defect. Harvard BusinessReview, Vol. 73(11-12), pp. 88–99..

Khalifa, M., & Liu, V. (2003), Determinants ofSatisfaction at Different Adoption Stages ofInternet-Based Services. Journal of theAssociation for Information Systems, Vol. 4 (5),pp. 206- 232.

Koufaris, M. (2002), Applying the technologyacceptance model and flow theory to onlineconsumer behavior. Information SystemsResearch, Vol.13 (2), pp. 205-23.

LaTour, S. A. & Peat, N. C. (1979), Conceptualand Methodological Issues in ConsumerSatisfaction Research. in Advances inConsumer Research, Vol. 6, William L. Wilkie,ed. Ann Arbor, MI: Association for ConsumerResearch, pp. 431-37.

Lee, J., Lee, J., & Feick, L. (2001), The impact ofswitching costs on the customer satisfaction -loyalty link: Mobile phone service in France.Journal of Services Marketing, Vol. 15, pp. 35-48.

Malthouse, O., Oakley, J., & Calder, B. (2003),Modeling the Relationship Between CustomerSatisfaction and Loyalty. Marketing Science,College Park, Maryland, pp. 13-15.

Mittal, V., & Kamakura, W.A. (2001), Satisfaction,repurchases intent, and repurchases behavior:Investigating the moderating effect of customercharacteristics. Journal of Marketing Research,Vol. 38, pp.131–142.

Mittal, V., Ross, W.T. Jr. & Baldasare, P.M. (1998),The asymmetric impact of negative andpositive attribute-level performance on overallsatisfaction and repurchase intentions. Journalof Marketing, Vol. 62 (1), pp. 33-47.

Mohr, C. D., Armeli, S., Tennen, H., Temple, M.,Todd, M., Clark, J., & Carney, M. A. (2005),Moving beyond the keg party: A daily processstudy of college student drinking motivations.Psychology of Addictive Behaviors, Vol. 19, pp. 392-403.

Neal,W.D. (1999), Satisfaction is nice, but valuedrives loyalty. Marketing Research, pp. 21-23.

Novak, T. P., Hoffman, D. L. & Yung, Y. F. (2000),Measuring the Customer Experience in OnlineEnvironments: A Structural ModelingApproach. Marketing Science, Vol. 19 (1), pp. 22-42.

Oliver, R. L. (1977), Effect of Expectation andDisconfirmation on Post-exposure ProductEvaluations: An Alternative Interpretation.Journal of Applied Psychology, Vol. 62 (8), pp. 480-486.

Oliver, R. L. (1980), A Cognitive Model of theAntecedents and Consequences of SatisfactionDecisions. Journal of Marketing Research,Vol.17 (9), pp. 460-469.

Oliver, R. L. (1999), Whence Consumer Loyalty?Journal of Marketing, Vol. 63 (4), pp. 33-44.

Olshavsky, R.W. & Miller, J.A. (1972), Consumerexpectations, product performance, andperceived product quality. Journal ofMarketing Research, Vol. 9 (2), pp. 19-21.

Peters, T, & Waterman, R. (1982), In Search ofExcellence: Lessons from America’s Best RunCompanies. New York: Harper and Row.

Pham, M. T. (2004), The Logic of Feeling. Journalof Consumer Psychology, Vol. 14, pp. 360–369.


73

Reichheld, F. F. & Schefter, P. (2000), E-loyalty:Your secret weapon on the Web. HarvardBusiness Review, Vol. 78, pp. 105–113.

Reichheld, F. F. (1996), Learning from customerdefections. Harvard Business Review, Vol. 74,pp. 56-67.

Sheth, J.N. & Kellstadt, C.H. (1992), A NormativeModel of Retaining Customer Satisfaction.Paper Presented at P.D. Converse AwardsSymposium, University of Illinois, Urbans-Champaign.

Sobel, M. E. (1982), Asymptotic confidenceintervals for indirect effects in structuralequation models. Sociological Methodology,Vol. 13, pp. 290–313.

Sugden, R. (1985), “Regret, Recrimination, andRationality,” Theory and Decision, 19, pp. 77-99.

Szymanski, D.M., & Henard, D.H. (2001),Customer satisfaction: A meta-analysis of thenempirical evidence. Journal of the Academy ofMarketing Science, Vol. 29(1), pp. 16-35.

Ulaga, W. (2001), Customer Value in BusinessMarkets: An Agenda for Inquiry. IndustrialMarketing Management, Vol. 30(4), pp. 1-7.

Zeelenberg, M. & Pieters, R. (2006), “LookingBackward With an Eye on the Future:Propositions toward a Theory of RegretRegulation,” in Judgments over Time: TheInterplay of Thoughts, Feelings, and Behaviors,eds. L.J. Sanna and E.C. Chang, New York,NY: Oxford University Press, pp. 210-29.

Zeithaml, V.A., Berry, L.L. & Parasuraman, A.(1996), The behavioural consequences ofservice quality. Journal of Marketing. Vol. 60, pp. 31-46.

74

ABSTRACT

The consumer price index measures changes inthe price level of consumer goods and servicespurchased by households, which reflect inflation.Various studies have been conducted tomodelling and forecasting Consumer Price Index(CPI) by developed countries .However, suchstudies have not been reported in Sri Lanka. Thispaper is an attempt to modelling the ColomboConsumer Price Index (CCPI) by using monthlyCCPI data from January 2003 to May 2011. Forthis purpose, Stepwise Regression, PrincipalComponent Analysis and Vector Autoregressive(VAR) approach were used. The VAR model withthe first principal component of selected CCPIcomponents was identified the best fitted modelfor the CCPI series. The model was also testedto an independent data set using CCPI fromFebruary 2010 to May 2011.

Keywords: CCPI, Inflation, Modelling, VAR

Introduction

The Consumer Price Index (CPI) is an indicatorto measure the average change in the prices paidby consumers for a specific basket of goods andservices over time in a country. This “shoppingbasket” represents a different items consist ofcommon consumer goods and services which arepurchased by an average household. The weightsfor each item in the shopping basket aredetermined based by the amount spent on theseitems by households in a given country.

Some international standards for economicstatistics have evolved primarily in order toenable internationally comparable statistics to be

compiled. However, individual countries alsostand to benefit from international standards. TheCPI standards described in this manual drawupon the collective experience and expertiseaccumulated in many countries. All countries canbenefit by having easy access to this experienceand expertise. In many countries, the CPIs werefirst compiled mainly in order to adjust wages tocompensate for the loss of purchasing powerwhich is caused by inflation. (Sylvester Young,2008).

The CCPI was introduced by the Department ofCensus and Statistics (DCS), in 2007, which isbased on the Household Income and ExpenditureSurvey (HIES) conducted by the DCS in2002.The HIES data represents more up to dateconsumer patterns for a much larger sample size,as well as an increased coverage area withinColombo for price collection.

The CCPI is based on prices of food and non-alcoholic beverages, cloth and footwear, housing,water, electricity, gas and other fuels, furnishing,household equipment and routine maintenance ofthe house, health, transport, communication,recreation and culture, education, miscellaneousand goods and services that people buy for theirdaily living.

The modelling and forecasting is usually carriedout in order to provide an aid to decision makingand planning the future. Forecasting CCPI areimportantinputs for government, businessessector, policy makers, investors, workers andvarious individuals for various applications.

The objective of this study is to modelling CCPIby applying a vector auto regressive (VAR)model.The VAR model was introduced by Sims

MODELLING COLOMBO CONSUMER PRICE INDEX:

A VECTOR AUTOREGRESSIVE APPROACH

M.C. Alibuhtto

Department of Mathematical Sciences, Faculty of Applied Sciences, South Eastern University of Sri Lanka.

E-mail: [email protected]

Modelling Colombo Consumer Price Index: A Vector Autoregressive Approach

75

(1980) and this model is the mostsuccessful,flexible, and easy to use models forthe analysis of multivariate time series. It isanatural extension of the univariate autoregressivemodel to dynamic multivariate time series. TheVAR model has proven to be especially usefulfordescribing the dynamic behavior of economicandfinancial time series andfor forecasting.

Economic literature on issue of CCPImodellingand forecasting is concerned with thepositive relation of inflation level. Kenny et al.(1998) conducted a study to develop a multipletime series models for forecasting Irish Inflation.For this purpose, the Bayesian approaches to theestimation of VAR models were employed. Alarge selection of inflation indicators is assessedas potential candidates for inclusion in a VARmodel. The results of this study confirm that thesignificant improvement in forecastingperformance, which can be obtained by the useof Bayesian techniques.

Genberg and Chang (2007) conducted a study todevelop a multivariate time series model toforecast output growth and inflation in the HongKong economy. For this purpose, the three typesof VAR (unrestricted VAR, Bayesian VAR andconditional VAR) models were used. Based ontheir study, the results suggest that the BayesianVAR framework incorporating externalinfluences provide a useful tool to produce moreaccurate forecasts relative to the unrestrictedVARs and univariate time series models, andconditional forecasts have the potential to furtherimprove upon the Bayesian models.

Enders (2004), Fritzer et al. (2002), Lutkepol(2001) and many other authors suggest that forthe calculation of forecasts of economicindicators VAR models should be appliedbecause all variables in these models areendogenous, and,therefore, not a single variablemay be removed when explanations for thebehaviour of other variables areoffered. For theforecasting of economic indicators two types ofVAR models may be applied: simple, orunrestricted, VAR models and models with

certain restrictions on exogenous indicatorspresent in them, or restricted VAR models.

Materials and Methods

Data: The secondary data on monthly CCPI datafrom January 2003 to May 2011 were consideredfor the analysis and it was collected fromDepartment of Census and Statistics. Thecollected monthly CCPI data from January 2003to January 2010 (‘training set’) is used for modelfitting and data from February 2010 to May 2011(‘validation set’) is used for validation of themodel.

Vector Autoregressive Models (VAR)

VARis an econometric model has been usedprimarily in macroeconomics to capture therelationship and independencies betweenimportant economic variables. They do not relyheavily on economic theory except for selectingvariables to be included in the VARs. The VARcan be considered as a means of conductingcausality tests, or more specifically Grangercausality tests.

VAR can be used to test the Causality asGranger-Causality requires that lagged values ofvariable ‘X’ are related to subsequent values invariable ‘Y’, keeping constant the lagged valuesof variable ‘Y’ and any other explanatoryvariables. In connection with Granger causality,VAR model provides a natural framework to testthe Granger causality between each set ofvariables. VAR model estimates and describe therelationships and dynamics of a set ofendogenous variables. For a set of ‘n’ time seriesvariables , a VAR model withexogenous variables of order p (VARX (p)) canbe written as:

Where εt - error term

g( )'.,,........., 21 ntttt yyyy =

(1) tj

m

jjit

p

iit XByAAy �+++= ��

=�

= 110


76

Model Selection Criteria: The followingstatistical measures were used to find anappropriate model for CCPI.

Where;k = number of coefficient estimated, rss =residual sum of

square, n = number of observations

Results and Discussions

The objective of this study is modelling CCPIseries with VAR models.For this purpose twodifferent approach were used.

1. VAR model with some Components ofCCPI were selected by using stepwiseregression method.

2. VAR model with common value for allCCPI componets by using principalcomponent techniques.

VAR Model between CCPI versus Its

Selected Components

The stepwise regression method was used to findsuitable variables among ten components ofCCPI series taking one lag behind. Both theprobabillity of entry a varaible and to remove avariable was set as 0.05. The results of stepwiseregression are tabulated in table 1.

( )

)3(2log

)2(loglog

��

��

� �+��

��

�=

��

��

� �+��

��

�=

nk

nrssBIC

nkn

nrssAIC

Based on the above results, the selectedregression equation can be written as:

Table 1: Output ofthe stepwiseregression analysis on CCPIversus FD (-1),

CL (-1),…, MS (-1)Alpha-to-Enter: 0.05 Alpha-to-Removes: 0.05

Step 1 2 3 4 VIF Constant �0 23.064 17.406 -2.324 -7.647 - FD(-1)�1 t-Value p-Value

0.847 100.17 0.000

0.641 48.95 0.000

0.541 32.76 0.000

0.551 34.12 0.000

29.667

HO (-1) �2 t-Value p-Value

- 0.214 16.60 0.000

0.194 18.88 0.000

0.186 18.37 0.000

11.978

MS (-1) �3 t-Value p-Value

- - 0.284 7.63

0.000

0.554 5.64

0.000

203.717

FU (-1)�4 t-value p-value

- - - -0.242 -2.95 0.004

186.991

S 3.44 1.65 1.26 1.21 R-sq(adj) 99.18 99.81 99.89 99.90

Mallows Cp 629.2 83.1 17.2 10.0

1111 554.0242.0186.0551.0647.7 �� +�++�= ttttt MSFUHOFDCCPI (A)


77

Results in table1indicates that four indices canbe selected from ten indices and thesefourindices are significant at 5% significance level.The selected indices are Food (FD), Housing(HO), Furnishing (FU) and Miscellaneous (MS).It should be noted that VIF of each varaible isvery high confirming the excisatnce of highlysignificant multicollinearity among the fourexplantory varaibles and consequently the abovemodel is not recommended to forecast. It is alsofound that the constant term in the model issignificantly different from zero. The time seriesplot of the four selected indices and CCPI seriesare shown in figure 1. It indicates that the CCPIand selected four indices show upward trendsand these all series are non-stationary.

Figure 1: The time series plot of selected four

indices and CCPIseries.

Identification of Lag Order of VAR Model

Selection Criteria

Results in table 2 indicates that the minimumvalues of Schwarz Information Criteria (SIC) andHannan-Quinn (HQ) statistic were obtained atlag 1. Therefore, it can be concluded that theoptimal lag length of this model is one. Thus,Granger Causality test was carried out for CCPI,FD, HO, FU and MS and the results are shownin table 3.

Table 2: Values of SIC and HQ statistics atvarious lag orders

YearMonth

20102009200820072006200520042003JanJanJanJanJanJanJanJan

240

220

200

180

160

140

120

100

Dat

a

CCPINFDFUHOMS

Variable

Time Series Plot of CCPIN, FD, FU, HO, MS

3.2 Lag O

Lag SIC HQ 0 30.84 30.75 1 17.84* 17.30* 2 18.47 17.48 3 18.90 17.49 4 19.54 17.46 5 20.37 18.01


78

Table 3 indicates that thesevennull hypothesesare rejected at 5% significance level (p-value <0.05). The F-Statistic values are significant andprovide strong evidence for the argument thatthere is bi-directional linear granger causalitybetween CCPI and selected indices of CCPI(HO, FU and MS) but FD and CCPI has onlyunidirectional granger causality relationship.

VAR Model with FD, HO, FU and MS

The parameter estimates of VAR model for CCPI versus FD, HO, FU and MS are shown intable 4.

Table 4 indicates that the estimates of allparameters are significant at 5% significancelevel (p-value < 0.05). As the constant term wasnot significant in this model and model withoutconstant term was considered. The final VAR

model with components of CCPI can be wrritenas:

The suitable Principal Component (PC)

for all components of CCPI

It is obvious that the correlations between allcomponents of CCPI are interrelated and fairlylarge. Also, there is significant multicollinearityexists among the components of CCPI, the dataset can be used to reach the possibility of

dimensional reduction among indices. The resultsof communalities of CCPI components aretabulated in table5.

)(313.0183.0140.0391.0329.0

1

1111

BMSFUHOFDCCPICCPI

t

ttttt

�

��

+

�++=

3.2 Lag O

Null Hypothesis F-Statistics P-Value FD does not Granger Cause CCPI 3.90 0.041 CCPI does not Granger Cause FD 2.70 0.103 HO does not Granger Cause CCPI 14.08 0.000 CCPI does not Granger Cause HO 6.25 0.014 FU does not Granger Cause CCPI 12.40 0.000 CCPI does not Granger Cause FU 28.02 0.000 MS does not Granger Cause CCPI 4.31 0.032 CCPI does not Granger Cause MS 8.16 0.001

Table 3:

Results of Granger Causality test between CCPI and its components

Variables Coefficients Stand. Error t-value P-value

CCPI (-1) 0.329 0.115 2.859 0.005

FD (-1) 0.391 0.065 6.062 0.000

HO (-1) 0.140 0.021 6.692 0.000

FU (-1) -0.183 0.069 -2.649 0.008

MS (-1) 0.313 0.056 5.606 0.000

Table 4:

Results of parameter estimation of identified VAR model


79

Table 5: Results of communalities of CCPI

components

Table 5 indicates that the maximum of 99.5% thevariance in furnishing and the minimum of26.9% of the variance in communication indexare accounted by the extracted factors. Hence,the communication index can be dropped to findthe principal component because it has lesscommunality value. The results of communalitiesof nine components of CCPI are tabulated intable 6.

Table 6: Results of communalities of nine

components of CCPI.

Table 6 indicates that the maximum of 99.8% ofthe variance in miscellaneous and the minimumof 86.7% of the variance in recreation index areaccounted by the extracted factors.

Table 7: Eigen values of the correlation

matrix of CCPI components

Table 7 indicates that the first componentaccounts for 94.399 % of the variance. Allremaining components are not significant.Hence, the first component has been chosen.

Table 8: Eigen scores of the first Principal

Component (PC)

According to the table 8, the nine componentscan be reduced to single Principal Component(PC) and a new variable is denoted by PCand itcan be written as:

Variable Initial Extraction Food 1.000 0.974

Clothing 1.000 0.979 Housing 1.000 0.885 Furnishing 1.000 0.995

Health 1.000 0.887 Transport 1.000 0.942

Communication 1.000 0.269 Recreation 1.000 0.859 Education 1.000 0.960

Miscellaneous 1.000 0.992

consumer goods and services purchased by households, which reflect inflation. Various studies have been conducted to modelling and forecasting Consumer Price Index (CPI) by developed countries .However, such studies have not been reported in Sri Lanka. This paper is an attempt to modelling the Colombo Consumer Price Index (CCPI) by using monthly CCPI data from January 2003 to May 2011. For this purpose, Stepwise Regression, Principal Component Analysis and Vector Autoregressive (VAR) approach were used. The VAR model with the first principal component of selected CCPI components was identified the best fitted model for the CCPI series. The model was also tested to an independent data set using CCPI from February 2010 to May 2011.

Keywords: CCPI, Inflation, Modelling, VAR

Variable Eigen scores Food(FD) 0.338

Clothing(CL) 0.340 Housing(HO) 0.325 Furnishing(FU) 0.342

Health(HL) 0.321 Transport(TR) 0.333 Recreation(RE) 0.319 Education(ED) 0.339 Misc(MS) 0.343

)(343.0339.0319.0333.0321.0342.0325.0340.0338.0

CMSEDRETRHLFUHOCLFDPC

++++

++++=


80

VAR model between CCPI versus first PC

VAR Lag Order Selection Criteria

Table 9: Results of lag order selection

Results in table 9indicates that the minimumvalues of SIC and HQ statistic were obtained atlag 2. Therefore, it can be concluded that theoptimal lag length of this model is two. Thus,Granger Causality test was carried out for CCPIand PC and the results are shown in table 10.

Table 10: Results of Granger Causality test

between CCPI and PC

Table 10 indicates that both the null hypothesesare rejected at 5% significance level (p-value <0.05). The F-Statistic values are significant andprovide strong evidence for the argument thatthere is a bi-directional linear Granger causalitybetween CCPI and PC.

VAR Model with PC

The parameter estimates of VAR model for CCPIversus PC are shown in table 11.

Table 11: Results of parameter estimation

Table 11 indicates that the estimates of allparameters are significant at 5% significancelevel (p-value < 0.05). The fitted VAR model canbe wrriten as:

Comparisons between Fitted two VAR

Models

In the model estimation, the AIC and SIC valuesfrom each estimated models are computed. AICand SIC values will be used in order to estimatewhich model is a better model for CCPI. For thispurpose, the model with the lowest AIC and SICvalues are concluded to be a better model. Theresults are reported in table 12.

Table 12: Comparison of the fitted two VAR

models

The results indicate that the both AIC and SICvalues from modelD is the lowest compared withmodelB. Also, log likelihood value is high formodelD. Therefore, it shows that the modelD isthe best model for forecasting monthly CCPIseries.

Lag SIC HQ 0 17.38 17.34 1 8.45 8.34 2 8.44* 8.27* 3 8.54 8.28 4 8.68 8.35 5 8.73 8.34

Null Hypothesis F_Value P_Value

CCPI (N) does not Granger Cause PC 15.268 0.000

PC does not Granger Cause CCPI (N) 9.141 0.000

)(377.4201.0258.0623.0762.1 2121

DPCPCCCPICCPICCPI ttttt

+

+��= ��

Model Log likelihood AIC SIC B -144.67 3.51 3.60 D -136.44 3.41 3.55

Variables Coefficients Stand. Error t-value P-value CCPI (-1) 1.762 0.132 13.346 0.000 CCPI (-2) -0.623 0.146 -4.279 0.000 PC (-1) -0.258 0.063 -4.080 0.000 PC (-2) 0.201 0.059 3.426 0.001

C 4.377 1.712 2.556 0.012


81

3.5.4 Forecasting Performance of the

Selected VAR Models

Table 13: Results of forecast performance

statistics

Table 13 indicates that the mean of percentageerror (MAPE) for validation set of the modelDislower than the modelB. Therefore, it can beconcluded that the modelDis a better forecastmodel for monthly CCPI series.

Conclusions and

Recommendations

This study aimed to modelling CCPI series usingVAR approach. Two types of VAR models wereestimated and model was also tested to anindependent data set using CCPI from February2010 to May 2011. The comparativeperformance of these VAR models have checkedand verified by using the model selectionprocedure (AIC and SIC). The comparisonindicates that the VAR model with the firstprincipal component of selected CCPIcomponents was identified the best fitted modelto forecast the CCPI series. The error series ofthe fitted model was found to be a white noiseprocess.

References

Enders, W., (2004), Applied Econometric TimeSeries, New York: John Wiley and Sons.

Fritzer, F., Moser, G., and Scharler, J., (2002),Forecasting Austrian HICP and Its ComponentsUsing VAR and ARIMA Models,Oesterreichische National bank Working PaperNo. 73, pp. 1-45.

Genberg, H., and Chang, J., (2007), A VARFramework for Forecasting Hong Kong’sOutput and Inflation, Research Department,Hong Kong Monetary Authority, WorkingPaper 02.

Kenny, G., Meyler, A., and Quinn, T., (1998),Bayesian VAR Models for Forecasting IrishInflation, Central Bank of Ireland, TechnicalPaper, 4/RT/98.

Lutkepol, H. (2001). Vector Autoregressive andVectorError Correction Models. Institute ofStatistics and Econometrics.

Sylvester Young, D., (2008), Global Wage Report2008/09, International Labour Office (ILO),Geneva, ISBN 978-92-2-121499-1

Sims, (1980), Macroeconomics and Reality,Econometrica, 48, pp. 1-48.

Model Data set MAPE

B Training set 0.61

Validation set 0.81

D Training set 0.59

Validation set 0.62

82

ABSTRACT

Information and communication technologies(ICT) have become ordinary entities in allaspects of life. In Trincomalee zone, according tothe pass level rate below 50% of students arepassed in ICT education in 2011. Based on thatresearcher want to be identify factors influencingtowards ICT education or not. The objectives areto identify the factors affecting the ICTeducation, to identify which factor has highlyinfluenced towards ICT education, to identify theimpact of social group influence on ICTeducation and to provide the suggestion todevelop towards ICT Education. Conceptualvariables are resource availability, personalInterest, global Trend on ICT, outcomes andimpact of ICT education.200 Advance Levelstudents were taken as sample from all seven1AB schools in the Trincomalee zonaldivision.The findings of the research are all thevariables are moderately supportedin theresearch. Suggestion are proper trainings shouldbe provided to the existing teachers to get betteroutcomes and impact.

Key words: ICT education, Attitudes, Trend

Background

IT refers to an entire industry that usescomputers, networking, software and otherequipment to manage information. Generally, ITdepartments are responsible for storing,processing, retrieving and protecting digitalinformation of the company. For achieving thesetasks, they are equipped with computers, DBMS,servers and security mechanisms, etc.Professionals working in IT departments rangefrom system administrators, databaseadministrators to programmers, networkengineers and IT managers. When executing abusiness, IT facilitates the business by providingfour sets of core services. These core services areproviding information, providing tools toimprove productivity, business processautomation and providing means to connect withcustomers. Currently, IT has become an essentialpart in business operations and has provided lotof job opportunities worldwide. Knowledge in IThas become essential to succeed in theworkplace. Typically, IT professionals areresponsible for a range of duties including simpletasks such as installing software to complex taskssuch as designing and building networks andmanaging databases. ICT is a term widely usedin the context of education. Even though there isno universally accepted definition for ICT, itmainly refers to utilizing digital technologiessuch as computers, television, email, etc to helpindividuals or organizations to work with digital

STUDENTS’ ATTITUDES TOWARDS INFORMATION AND

COMMUNICATION TECHNOLOGY (ICT) EDUCATION

(SPECIAL REFERENCE TO ADVANCED LEVEL STUDENTS

OF 1AB SCHOOLS IN TRINCOMALEE DIVISION.)

Subathini Priyadharsan

Senior Lecturer, Department of Business and Management Studies, Faculty of Communication and Business Studies, Trincomalee Campus, Konesapuri, Nilaveli.

S. Sivasenthuran

Student, External Degree Program, Dept. of BMS, Faculty of Communication and Business Studies, Trincomalee Campus, Konesapuri, Nilaveli

Students’ Attitudes Towards Information and Communication Technology (ICT) Education

83

information. ICT can be seen as an extendedsynonym for IT. Therefore, ICT can be seen asan integration of IT with media broadcastingtechnologies, audio/ video processing andtransmission and telephony.

Introduction of Information Technology into thesecondary school curriculum in Sri Lanka is avery recent development. The subject GeneralInformation Technology (GIT) was included forG.C.E. Advanced Level Examination in 2005while plans are afoot to introduce InformationTechnology as a subject for G.C.E. OrdinaryLevel from 2008. The results of the first GITexamination held in 2005 clearly show poorperformance by students. Considering the largeamount of money spent for InformationTechnology education in schools and theaspirations of students in acquiring a soundknowledge in Information Technology while inschool, it is essential that an analysis is done toevaluate the merits and demerits of InformationTechnology education in schools.

In Sri Lanka Schools are being divided as 1AB,1C, Type I, and Type II schools. According toNational Standards they can be categorized asNational Schools and Provincial Schools.In thisstudy 1AB Schools are being selected.

Literature Review

Information and Communication Technology

ICT has been used in educational settings sinceits inception, but recent empirical research hasaffirmed that it plays a vital role in high-quality

learning and teaching. Such research insightshave shown that advances in technology haveopened up new possibilities for the way in whichteachers educate their classes, giving potentialfor innovative ways to encourage students tobecome more engaged in their schooling. Toenable the best possible outcomes for theirstudents it is vital that schools are able to keepup with this progress.The 2008 MelbourneDeclaration on Education Goals for YoungAustralians (MCEETYA, 2008: 6) affirmed theimportance of ICT literacy within the classroom,stating that: rapid and continuing advances ininformation and communication technologies(ICT) are changing the ways people share, use,develop and process information and technology.In this digital age, young people need to behighly skilled in the use of ICT. While schoolsalready employ these technologies in learning,there is a need to increase their effectivenesssignificantly over the next decade.

Currently Sri Lanka has a concept of being "ICTliterate". However, although this is important inits own right, it usually refers to a facility withhardware and software rather than to anythingwe would more normally identify as literacy. YetICT has becomean element of functional literacy.This is not a futuristic vision. It is becoming an

everyday reality. People read text on screenrather than on paper, for example, when wereceive emails and access the Internet. (Report ofhousehold Computer Literacy Survey of SriLanka – 2004).In certain offices and homes, acomputer with a word-processing package is

• NR – Result still not release Source – Zonal Education- Trincomalee/2012


84

frequently used as the principal writing tool. ICTas a means for both producing and accessingtexts, and indeed as the medium in which someof those texts will exist, is already commonplace.ICT is not something we need to add on toliteracy. It must be an integral part of what wemean by being literate.According to theHousehold Survey on Computer Literacy of SriLanka – 2004, the target population was all thosein the age group of 5 – 69 years. Of thispopulation only 9.7 percent was found to becomputer literate (Household Computer LiteracySurvey of Sri Lanka – 2004).

ICT Education in South Asia

The World Bank supported InfoDev iscoordinating a survey of the use of informationand communication technology for education inIndia and South Asia. The purpose of this surveyis to gather together in a single resource the mostrelevant and useful information on ICT ineducation activities in India and South Asia.It isenvisioned that data collected as part of this surveyprocess can help to form a set of baseline data,against which future survey work and researchcould be compared. This data can be combinedwith data from other regions already surveyed orto be surveyed to help form a global database ofinformation related to ICTs in education indeveloping regions. Information andcommunication technologies (ICTs) are widelybelieved to be important potential levers tointroduce and sustain educational reform efforts, aswell as useful aids to both teaching and learning.However, despite evidence of increasinglywidespread use of ICTs in education initiativesaround the world, there is little guidance availablefor policy-makers and donor staff specificallytargeted at developing countries contemplating theincreased use of ICTs in education.

ICT Education in Sri Lanka

G.C.E O/L Syllabus

Information & Communication Technology (ICT)has become the state of the art technology of thecontemporary world. Every sector of the

economy is forced to use this technology to maketheir work effective and efficient and therebymaintain a competitive edge. This has changedthe types of skills and knowledge needed in theworld of work.Since Sri Lanka is in the earlystages of introducing ICT to lower grades, thepresent syllabus does not demand any ICTknowledge as an entry requirement. Therefore,this syllabus is intended to introduce ICT/computer science as a technical subject to beoffered at the G.C.E (O/L). The objectives of thissyllabus are to develop the competencies toutilize the ICT tools and to build a basictheoretical base for the student to act as afoundation to pursue higher studies in ICT.Thepresent infrastructure and human resourcedevelopment (teacher training) facilities are notsufficient to introduce this ICT syllabus.However, it is supposed that the required facilitiesfor schools will be fulfilled in the near future byvarious donor agencies like Asian DevelopmentBank, World Bank, etc. Sustainability of thissyllabus depends on the continuous workingconditions of ICT laboratories and the frequentprovision of updated knowledge of the ICTteachers. The content of this syllabus has to beupdated regularly to match the requirements ofthe rapidly changing technological world.Theteaching/learning of this syllabus could partly becarried out in the English medium. However, abilingual method should be used so that the termsof new ICT concepts are given in the media, themother tongue and English.The minimumproficiency expected could be assessed andevaluated throughout the course of study. Thepractical components of this ICT syllabus providea good opportunity for school based assessment.At the same time a fair percentage of contentcould be incorporated at national level evaluation.

ICT in G.C.E A/L

Information Communication Technology (ICT)will be included as a subject in the AdvancedLevel examination curriculum from 2009, saidSecretary to the Ministry of EducationAriyarathneHewage. He said approval for theICT policy for education will be obtained soon


85

and it will help develop capacity, train teachers,set up ICT centers and build private publicpartnerships. Ariyarathne was addressing theacademia career guidance pre meet organized bythe Information Communication TechnologyAgency (ICTA) which will host the National ICTCapacity Summit and National ICT Career Fair(NICS 2007) from September 1-2 inColombo.Ariyarathne said informationtechnology and computer science has grown at aphenomenal rate to a US$ 1 trillion industry, butthe country has not been able to keep pace withdevelopments due to the lack of skilled ITpersonnel. The challenge today is to get ITprofessionals in the market. The IT workforcegrew by 10,000 during 2004-2006.Over 14,000IT workers are required to meet the industryneeds in the next two years. This year 5,755graduates are needed for the industry, hesaid.The attrition rate of the IT workforce in thecountry has doubled from 6.6 percent in 2004 to13 percent in 2006. Fixation is an obstacle to thegrowth of the IT industry. Parents and teacherswant their students to become doctors andengineers and not IT professionals. Students aredirected to sit for examinations to enteruniversities without providing them careerguidance. Microsoft Chairman Bill Gates hasvolunteered to support Sri Lanka to developcapacity building in Information CommunicationTechnology (ICT) by setting up a Microsoftcenter for excellence in the country.The Ministryof Education has introduced several programs todevelop IT in the country and the e-villageconcept is one such program to bridge the digitaldivide. The village program was launched inMahawilachchiya, Nihiluwa, Mahalakotuwa,Pitakumbura, Nikawewa and Dhamana.GeneralManager Virtusa, MaduRatnayake said with theshift from an industry based economy to aknowledge based economy world dependence onIT has grown tremendously.The BPO industry inAsia is growing at a rapid pace. According to asurvey by AT Kearney Global Service Ltd. SriLanka is ranked 29 among the top 50 countriesin outsourcing. India, which is ranked first has avibrant IT industry, which contributes 5-6percent of the GDP. ICT will be the next growth

wave for jobs and wealth creation in the country,he said. Source: Sunday observer on 2007/06/17/

Improve teaching and learning quality

As Lowther et al. (2008) have stated that thereare three important characteristics are needed todevelop good quality teaching and learning withICT: autonomy, capability, and creativity.Autonomy means that students take control oftheir learning through their use of ICT. In thisway, they become more capable of working bythemselves and with others. Teachers can alsoauthorize students to complete certain tasks withpeers or in groups. Through collaborativelearning with ICT, the students have moreopportunity to build the new knowledge ontotheir background knowledge, and become moreconfident to take risks and learn from theirmistakes. Further, Serhan (2009) concluded thatICT fosters autonomy by allowing educators tocreate their own material, thus providing morecontrol over course content than is possible in atraditional classroom setting. With regard tocapability, once students are more confident inlearning processes, they can develop thecapability to apply and transfer knowledge whileusing new technology with efficiency andeffectiveness. By using ICT, students’ creativitycan be optimized. They may discover newmultimedia tools and create materials in thestyles readily available to them through games(Gee 2007, 2011), CDs, and television. With acombination of students’ autonomy, capability,and creativity, the use of ICT can improve bothteaching and learning quality.

Student Perspective

Frederick, Schweizer and Lowe (2006) showedthat student mobility, special needs, and anxietyover standardized test results are the mainchallenges associated with ICT use. Thesechallenges can be solved by providing moreauthentic group- and problem-based learningactivities, and adequate learning support (Whelan2008). Whelan (2008) also identified morebarriers from the student perspective, including:subpar technical skills that reduce access to ICT


86

in classroom; an insufficient number of academicadvisors and lack of timely feedback frominstructors; and reduced interaction with peersand instructors. Therefore, the authorrecommends the following strategies to facilitatethe learning process: more induction, orientation,and training for students; an increased emphasison the importance of instructor access andeffective administration; and the expansion ofpodcasting and online conferencing tools. Ingeneral, capacity building, curriculumdevelopment, infrastructure, policy, andgovernment support are required in order tolower student barriers and improve theeffectiveness of ICT use in the classroom. Inaddition, Castro Sánchez and Alemán (2011)encourage students to acquire specific technicalskills to facilitate learning in ICT environments.Provide adequate technical support (Liu andSzabo 2009; Tezci 2011a; Yildirim 2007).Technology should be used for more than justsupport of traditional teaching methods (Tezci,2011a). According to Tezci (2011a), teachersshould learn not only how to use technology toenhance traditional teaching or increaseproductivity, but also should learn from a studentcentered perspective how ICT can be integratedinto classroom activities in order to promotestudent learning. This means that teachers needto use ICT in more creative and productive waysin order to create more engaging and rewardingactivities and more effective lessons (Birch andIrvine 2009; Honan 2008). Hence, CastroSánchez and Alemán (2011) suggested thatteachers keep an open mind about ICTintegration in classroom. It is imperative thatteachers learn new teaching strategies to adapt tothe new instruments when teaching withtechnology. However, Yildirim (2007) found thatteachers use ICT more frequently for thepreparation of handouts and tests than to promotecritical thinking. Similarly, Palak and Walls(2009) found that teachers mainly use technologyto support their existing teaching approaches andrarely to foster student-centered learning.According to the authors, one possibleexplanation is a lack of models for how to use

technology to facilitate learning, and limitationsrelated to contextual factors such as class sizeand student ability. Further, Brush, Glazewskiand Hew (2008) found that pre service teacherpreparation does not provide sufficient ICTknowledge to support technology basedinstruction, nor does it successfully demonstrateappropriate methods for integrating technologywithin a curriculum. More training should beprovided in pre-service teachers’ curricula, andICT skills must be applied in the classroom inorder to integrate effective technology strategies(Supon and Ruffini 2009). To help teachers copewith these difficulties, Chen (2008) suggestedthat rather than only providing educationtheories, ICT researchers should also documentexamples of how teachers accomplishmeaningful and effective technology integrationto meet their pedagogical goals and needs.

Problem Statement

Nowadays ICT education is the most essentialone and there are plenty of opportunities for thestudents who have completed their schooleducation up to Advanced level, to get a highereducation qualification through different ICTeducational programs. In Trincomalee zone,according to the pass level rate below 50% ofstudents are passed in ICT education in 2009.Based on that researcher want to beidentifiedstudents' attitudes influence on ICTeducation or not?

Research Questions

The main research questions are identified asmentioned below

Ø What are the attitudes of students towardsICT education after school education?

Ø What are the factors affecting the students’attitudes on ICT education?

Ø What are the impacts of Social groupinfluence?


87

Objectives

Ø To identify the factors affecting the students’attitudes towards ICT education.

Ø To identify which factor highly influences onattitudes towards ICT education.

Ø To identify the impact of social groupinfluence on attitudes of student towards ICTeducation.

Ø To provide the suggestion to reduce thenegative attitudes of ICT Education.

Scope of the Study

This research mainly focuses on Advance Levelstudents’ attitudes towards ICT education. Evenif they get through their A/L what is their attitudetowards ICT education, the students of 2012batch will be selected for this study.

Significance of the study

ICT education has been taken for this researchbecause this has been one of the major tools inone person’s skills and knowledge development.The main purpose of this research is to find outthe major influence for the attitude of thestudents of the Trincomalee Zonal area. Educatethe student to use the opportunities with cheapand secure manner to get an ICT educationalqualification.

Limitations

Ø Data collection and research are restricted toonly 2012 Advance Level students.

Ø The survey restricted to only 1AB Schoolsfor Tamil medium in the Trincomalee ZonalDivision.

Assumption of the Research

Ø All the questions and statements are clearlyunderstood by the respondents.

Ø The collected data are true and fair andunbiased.

Conceptualization

As illustrated below (in figure 1), the conceptualframework is developed using four groups ofvariable which have mainly influenced the attitudesof Advance Level students towards ICT Education.

Figure 1: Conceptual framework

Source : Develop for research purpose

Conceptual framework is formulated based onthe following theories.Factors influencing the useof ICT can be divided into external factors andinternal factors. The two types of factors arerelated to each other and to ICT usage level(Tezci 2011a). A variety of external factors havebeen identified that influence the progression oreffectiveness of technology integration inschools. These include technology availability,accessibility of ICT equipment, time to plan forinstruction, technical and administrative support,school curriculum, school climate and culture,faculty teaching load and management routine,and pressure to prepare students for nationalentrance exams (Al-Ruz and Khasawneh 2011;Lin, Wang and Lin 2012; Tezci 2011a). Amongthese external factors, the most common are lackof access to computers and software, insufficienttime for course planning, and inadequatetechnical and administrative support (Chen,2008). Several internal factors also influencetechnology integration outcomes (Sang et al.2011). Internal factors related to teachers include:understanding of ICT use; beliefs, which mayconflict with the application of ICT; attitudestoward technology integration; perceptions,including intention or motivation to use ICT;self-confidence and knowledge; technologyskills; readiness to use ICT; and technology self-efficacy (Al-Ruz and Khasawneh 2011; Chen2008; Lin, Wang and Lin 2012; Sang et al. 2011;Tezci 2011a).

��

��

��

��

� ��

��

��

��

��

�

��


88

Operationalization

Details of the operationalization are given below.

Table 1: Operationalization

Variables Indicator Measurement

Resource Availability Financial support

Questionnaire Recognized Teachers & Trainings

Recognized Subjects

Personal Interests Perception on ICT courses.

Questionnaire

Family influences / Social Group

Knowledge/ Skills

Global Trend on ICT Awareness on importance of ICT

Questionnaire

Extent of ICT usage.

Job Market needs

ICT Education Outcomes and Impact

Performance

Questionnaire

Expectation

Satisfaction

Attitudes

Affective Questionnaire

Cognitive

Behavioral

Name of the School Total Population Percentage Selected Sample

T/Konneswara. Hindu college 178 36.70 37 T/St Joseph College 141 29.07 29 T/St Mary’s College 169 34.84 35 T/Orr’s Hill Vivekananda 152 31.34 31 T/Sri Shanmuga Hindu ladies 102 21.03 21 T/St Methodist College 116 23.91 24 T/Vipulananda college 112 23.09 23 TOTAL 970 200

Source: Developed for the research purpose

Sampling

A sampling is the collection of samples from thepopulation in the area where the research hasbeen done. From the researcher’s point of view,

the population includes 2012 batch A/L studentsfrom 1AB schools in the Trincomalee ZonalDivision. Identified the systematic randomsampling technique for research purpose.

Source: Zonal Education Department Trincomalee/2012

Table 2: Sample selection detail


89

Data analysis

The collected data have been analyzed by Mean,Correlation method and regression analysis.

Methods of Measurement

In this study, measurement was carried out byfive point Likert scale. The collected data fromthe questionnaires were analyzed based on thepositive effect on the students’ attitudes on ICTeducation by analyzing variables as mentionedbelow.

Strongly disagree-1/ disagree-2/ neutral-3/ agree-4/ strongly agree-5

Data evaluation

Each variable is given a scale from 1-5 to showthe extent of importance based onresponseunivariate measures are calculated foreach of the variables. For calculating theunivariate measures, the Microsoft Excel 2007and SPSS 17 windows have been used.

Table 3: Decision criteria and rule

Discussion of the Results

Table-4 Univariate analysis

Resource Availability:

Resource availability of students and schools arein moderately level that is the mean value is3.06. Financial Support: In the research areamajority students are coming under middle class.Therefore, most of the families are unable tospend money to buy computers and accessories.This may be the main reason for the above resultof the research on financial support.

Recognized Teachers & Training: According tothe result of the research, the indicator ofrecognized teachers and training are not insatisfactory level. The main reason could beinsufficient number of well trained, specializedteachers in IT in the school system. On the otherhand, the extent of providing training to AdvanceLevel students is not at the requiredlevel.Recognized Subject: Recognized subject isnot coming under satisfactory level. It reasoncould be according to the ICT syllabus, someuseful subjects don’t include in the ICT syllabussuch as computer programs, Web designing,CAD, etc.

Decision Criteria Decision Rule

1 xi 2.5 Low level of impact on attitudes on ICT Education

2.5 < xi 3.5 Moderate level of impact on attitudes on ICT Education

3.5 < xi 5 High level of impact on attitudes on ICT Education

Variables Mean Standard Deviation Resource Availability 3.06 0.588 Personal Interest 3.49 0.712 Global Trend on ICT 3.39 0.715 Outcomes and impact of ICT education

3.36 0.658

Source: survey data


90

Personal Interest

The second factor ‘Personal interest’ alsomoderately affects the students’ attitudes on ICTeducation. Perception on ICT courses: Here thegood sign is that the mean value of perception onICT courses is in higher level. That means thestudents in the research area is well aware aboutthe benefits and importance of ICT courses andvery interested to follow such courses.

Family Influence/Social groups: According to theAdvance level education system the three mainsubjects are only consideredin the calculation ofZ score other than ICT Education. Therefore,parents and social groups are not showing muchinterest to support on ICT education.

Knowledge & Skills: the last indicator of thepersonal interest, the medium of ICTexamination was English, while the majority ofschools in research area still uses nativelanguages, i.e. Sinhalese or Tamil, for teaching.

Global Trend on ICT

Global Trend on ICT is another variable whichmoderately influences the students’ attitudes onICT Education. Awareness on importance of ICT:According to the above research results thestudents almost have consciousness andunderstanding about the awareness on theimportance of ICT.

Extent of ICT usage: according to the result theindicator of Extent of ICT usage is moderatelevel, the reason for teachers and social groupsare couldn’t be instructed of usage of ICTeducation for every day to day student’sactivities.

Job market needs: according to the result of theresearch, the mean value of that indicator is 2.99.The main reason for Students’ could beunderstood get the job from his university study.

Outcomes and impact of ICT education

Outcomes and impact of ICT education alsomoderately affect the students’ attitudes.Performance: The indicator of performance levelis moderate. The main reason could be someschools are not provided special ICT trainingprograms and inter school ICT educationcompetitions for students.

Expectation: Expectation is the second indicatorof outcome & impact of ICT education.According to the research, analysis the meanvalue is 3.14 and It’s also moderately level, thereason for the ICT education is not compulsoryfor university entrances, further students'expectation of ICT education in Advance level isnot better than other subjects.

Satisfaction: Satisfaction is the last indicator ofthe outcome & impact of ICT education.According to the ICT examination system,examination result will be released about aftertwo years from the examination date. In thatsystem could be the impact of students’satisfaction about ICT education.

Correlation analysis

The table 5 gives details of the correlationbetween each pair if variables. The values hereare acceptable.


91

Data Reliability

Reliability of the variables, Cronbach’s coefficientalpha was calculated to evaluate the reliability of themeasures. An alpha level of 0.70 or above isgenerally considered to be accepted (Cronbach,

1951). All the measures in survey exceed thisthreshold. Resource availability (alpha= 0.841),personal interest (alpha=0.861), global trend (alpha= 0.803 ) and outcome and impact (alpha= 0.854).

Resource Availability

Personal Interest

Global Trend

Outcome & Impact

Attitudes


Pearson Correlation

1 .739** .731** .681** .702**

Sig. (2-tailed) .000 .000 .000 .000

N 200 200 200 200 200

Personal Interest

Pearson Correlation

.739** 1 .745** .818** .749**

Sig. (2-tailed) .000 .000 .000 .000 N 200 200 200 200 200

Global Trend

Pearson Correlation

.731** .745** 1 .821** .806**

Sig. (2-tailed) .000 .000 .000 .000

N 200 200 200 200 200

Outcome & Impact

Pearson Correlation

.681** .818** .821** 1 .800**

Sig. (2-tailed) .000 .000 .000 .000

N 200 200 200 200 200

Attitudes

Pearson Correlation

.702** .749** .806** .800** 1

Sig. (2-tailed) .000 .000 .000 .000 N 200 200 200 200 200

**Correlation is significant at the 0.01 level (2-tailed) There is a positive relationship between the conceptual variables and attitudes.

Coefficientsa

Model

Unstandardized Coefficients

Standardized Coefficients

t Sig.

B Std. Error Beta

(Constant) 0.15 0.155 0.965 0.036 Resource Availability

0.164 0.075 0.133 2.195 0.029

Personal Interest

0.138 0.073 0.136 1.884 0.051

Global Trend

0.361 0.073 0.356 4.935 0.043

Outcome & Impact

0.336 0.086 0.305 3.902 0.042

Table 5 Correlations

Table 6: Coefficients

Source: survey data


92

Regression Analysis

R, R Square and Adjusted R Square Analysis

Table 7:

R, R Square and Adjusted R Square Analysis

Adjusted R Square value is calculated whichtakes into account the number of variable in themodel and the number of respondents our modelis based on. This R Square value gives the mostuseful measure of the success of our model. Wehave adjusted R Square value of 0.721, thereforewe can say that our model has accounted for72% of the variance in the dependent variable.

Conclusions

The results of the survey indicate that thestudents’ attitudes towards ICT education is inthe moderate level. The result of the researchindicates that ICT related resources are fairlyavailable in the research area. But the availabilityof recognized / qualified teachers and thetraining method are not in the positive level. AsPersonal interest of students towards ICTeducation is comparable in higher level, there aresome opportunities to develop of ICT educationin the area.

The students are aware of Global trend on ICTand its importance. Even though it has beenidentified from the result of the research thatmost of the students are not properly trained orguided to meet the job market requirement.Therefore the outputs and impact of ICTeducation is not in a good position in theresearch area.

Recommendation


u Department of education has to take action toappoint new qualified IT teachers to theschools to provide better trainings toAdvance Level students. Further, it isrecommended to the Zonal educationdepartment has to provide necessarytrainings to existing IT teachers to developtheir skills.

u Schools have to provide useful special ICTtraining program for students in every ICTsubject.

u Ministry of education has to focus to re-design the IT syllabus according to theglobal trend and to meet the job marketrequirements. Some subjects are included inIT syllabus such as Program languages, webdesigning, AutoCAD, etc.

Personal Interestu Parents and social groups’ area should be so

much interesting to support on ICT educationfor the students. Further, they shouldencourage the ICT subject likely other threecompulsory subjects.

u The Department of Education should beannouncing an ICT subject is compulsory foruniversity entrance.

u Further Department of Education should bechange the medium of ICT examination isthe native language. i.e. Sinhalese or Tamil.

u The recognized ICT institutes are should beprovide special offer for school students,further should be given an internationalrecognized certificate.

Global Trend on ICT educationu Proper guidance should be given to the

students to follow the right course at theright time and right way by social groups.

u Trincomalee Zonal school computer labs areshould be connect under the one networksystem and conduct the inter schools ICTcompetition via online.

Model Summary

Model R R Square

Adjusted R Square

Std. Error of the Estimate

1 .852a .726 .721 .383


93

u The department of education should provideavalid ICT certificate for students get the greatjob.

u Proper guidance should be given by teachersfor the students to need of ICT when facingthe any job interview.

Outcome & impact of ICT educationu The department of examination should

release the ICT examination result withinthree months from the examination date.

u The schools should introduce a day for ICT(like English day, Science day etc.)

u The zonal Education department also has tointroduce inter school competition in ICTeducation give a valuable prize to the winner.

References

Ajzen, I., & Fishbein, M. (2000), Attitudes andattitude behavior: Reasoned and automaticprocesses. In European review of socialpsychology. W.Stroebe, M. Hewstone, (eds).Wiley, Chichester, England.

Cronbach, L.J. (1951), Coefficient alpha and theinternal structure of tests. Psychometrika, Vol. 16(3), pp. 297-334.

Chandima H. de Silva,University of Kelaniya, SriLanka, September 26-28, (2007), Villach,Austria.

Dagiene, Valentina., Teaching InformationTechnology in General Education: Challengesand Perspectives, R.T. Mittermeir(ed.), ISSEP2005, Lecture Notes In Computer Science3422, Springer-Verlag Berlin, pp. 53-64.

Evans, B. (2007), Student attitudes, conceptions,and achievement in introductory undergraduatecollege statistics. The Mathematics Educator,Vol. 17(2), pp. 24-30.

General Information Technology (GIT), Grade 12,Syllabus, National Institute of Education, SriLanka. (www.moe.gov.lk GIT resources).

Gurbiel, Ewaet al, Informatics and ICT in PolishEducation System, R.T. Mittermeir(ed.), ISSEP

2005, Lecture Notes In Computer Science3422, Springer-Verlag Berlin, pp. 46-52.

Jo Shan Fu (2013), ICT in Education: A CriticalLiterature Review and ItsImplications,International Journal of Educationand Development using Information andCommunication Technology (IJEDICT), 2013,Vol. 9, Issue 1, pp. 112-125

LiisaIlomäki (2008), The effects of ICT on school:teachers’ and students’ perspectives,Department of Teacher Education, Universityof Turku.

Photos Papaioannou, Kyriacos Charalambous(2011), Principals’ Attitudes towards ICT andTheir Perceptions about the Factors ThatFacilitate or Inhibit ICT Integration inPrimary Schools of Cyprus,Journal ofInformation Technology Education Vol. 10,2011

Ruth Geer, Trudy Sweeney (2010), Studentcapabilities and attitudes towards ICT in theearly years, ACEC 2010 conferenceproceedings.

Triandis, H.C. (1971), Attitude and attitude change.Wiley & Sons, New York, USA.

Zonal Education Department, Trincomalee.

journal of management journal/uose v_11_book_final.pdf · issn 1391-8230 volume xi no. 1 october...

Documents