chapter 3 data and methodology 3.1 statement of the...

51

Chapter 3

Data and Methodology

3.1 Statement of the Problem

The growth of higher education system in India has been muddled and

unplanned. The drive to make higher education socially inclusive has also led

to a swift and remarkable increase in the number of higher education

institutions without a fair increase in infrastructure and intellectual resources.

Consequently, academic standards have been put at risk (Béteille, André,

2005). Research in higher education institutions is at its lowest ebb. There is

an inadequate and diminishing financial support for higher education from the

government and from society.

According to a NASSCOM estimate, the supply of IT engineering

graduates in India was 95,000 in 2003-04, in 2004-05 it was 100,000, and in

2005-06 it was 111,000. Each year, India produces almost twice the number

of engineers produced by the US and a little less than twice of all that Europe

produces. A look into the impetus given to education in our country indicates

that the spending on education has grown five times in the last 50 years. The

challenge is not so much from other parts of the globe as it is from China.

China is today the largest producer of engineering graduates in the world.6

6 http:// www.mission10x.com

52

Now the number of institutes set up for various disciplines year on year

indicates an engineering education boom in India. Engineering colleges in the

country have been growing at 20 per cent a year. Further, the quality of

engineering education also needed a boost.

The paradox is that, despite the increase in the number of higher

education institutions, the competition for acquiring fresh talent every year is

so heated that it gives an impression that resources are really scarce. In

reality, there is a plethora of career options for graduates of current years.

The challenge is not the supply of talent but that of talent that meets the

needs of the corporate world. In other words, the challenge is that of

employability. Only 25 percent of technical graduates and 10-15 percent of

other graduates are considered employable by the growing IT and ITES

sectors. Even after employing these graduates, most companies have to

spend considerable amount of time and resources on their training so as to

develop the skills required by the industry. If the students augment their skills

in a few specific areas desired by the industry, employability in the country

can be significantly enhanced. While India currently boasts of one of the

world's largest qualified pools of scientific and engineering manpower, the

growing global demand for quality graduates is gradually widening demand-

supply gap.

Given these issues, the purpose of this investigation is to examine the

perceptions of quality criteria, which are identified by a broad range of

stakeholders in a national study of quality in higher education (Donald, J. G.,

53

& Denison, D. B. 1993, June).The stakeholders are leaders of university and

college boards, administrators, faculty, students, alumni and members of the

larger community. Quality of higher education cannot be achieved without

knowing the perceptions of stakeholders.

The following needs are addressed by this research.

3.2 Objectives Primary Objectives 1. Need for a unique quality measurement tool to measure the

perceptions of various stakeholders on the quality of students

and faculty of higher education in India.

2. Need to determine the Perception Gap between the various

stakeholders (faculty, students, alumni and industries) on the

quality of higher education in India.

3. Need to determine the effect of demographic variables on the

perceptions of faculty, students, alumni and industries

respectively.

Secondary Objectives

The perception gap between the various stakeholders (Primary

Objective No.2) is identified through the following secondary

objectives:

54

1. To determine the Perception Gap between the industries and

students on the quality of students and faculty.


faculty on the quality of students and faculty.


alumni on the quality of students and faculty.

4. To determine the Perception Gap between the faculty and students on

the quality of students and faculty.

5. To determine the Perception Gap between the faculty and alumni on


6. To determine the Perception Gap between the alumni and students on


3.3 Conceptual Framework

Based on the clearly defined problem, it is paramount to measure the

perceptions of various stakeholders such as faculty, students, alumni and

industries on the quality of students and faculty of higher education in India. It

is also very important to determine the Perception Gap between the

stakeholders on the quality of higher education in India. As the existing

literature supports the need to determine the perceptions of the stakeholders

of higher education, it is very significant to determine the Perception Gap

between the Stakeholders. The whole study is conceptually derived as

shown in Figure 3.1. The measurement of perceptions of the stakeholders on

55

quality of faculty and students is based on the instruments used by Janet G.

Donald and D. Brian Denison, 2001.

Perception Gap

Perception

Instrument

Figure 3.1 Conceptual Framework

3.4 Significance of the Study

This Study makes a vital contribution in the form of a research tool,

further additions to the existing literature and valuable suggestions to the

administrators of higher education in India.

• A standardised measurement tool: The contribution of this research will

be a unique questionnaire to measure the perceptions of various

stakeholders on the quality of students and faculty of higher education

in India. There is no published work citing the availability of such

contextual instrument. Such instrument will give a direction for other

Industries

Faculty

Students

Alumni

Survey

Industries vs Faculty

Industries vs Students

Industries vs Alumni

Faculty vs Students

Faculty vs Alumni

Students vs Alumni

56

contextual academic researchers (India based) to further work on the

instrument and make it valid by testing it in different parts of India.

• Addition to quality literature: This research attempts to study the

perceptions gap between various stakeholders of higher education on

the quality of students and faculty. Until now the literature addresses

the perceptions of academic leaders on quality of higher education.

This research will give further additions to the literature on the

perceptions of students, alumni, faculty and industries on quality.

• Benefits to college administrators: The management of higher

education institutions can think in a new perspective and can face the

challenge of deterioration of quality beyond the perceptions of

academic leaders. The management of higher education institutions

will be given with the choice of looking at the perceptions of teaching

faculty, students and alumni to understand the methods to enhance the

quality of higher education institutions

• Benefits for teaching faculty: The teaching faculty shall understand the

perception gap between them, the students, alumni and industries and

improve their competence in delivery and content.

• The Indian educational system: To further the cause of Indian higher

educational system this research will attempt to study the relationship

57

between the stakeholders namely faculty, students, alumni and

industries.

Given the Higher Education System undergoing a colossal change with

privatisation and globalisation of education, this study will aid the development

of the system by bringing in a socially relevant tool and suggestions which will

enhance the quality of higher education institutions.

This study is focussed on “Quality Assessment in Higher Education” to

determine the perceptions gap between the stakeholders of higher education,

namely students, faculty, alumni and industries.

3.5 Defining the Hypotheses

Formal Directional Hypothesis: As the major question of this

research is to identify the perception of the stakeholders on the quality of

higher education in India and to measure the perception gap between the

stakeholders, the major directional hypotheses of the research are as follows:

58

H1. There is a significant difference between the perceptions of industries

and faculty on the quality of higher education.


and students on the quality of higher education.


and alumni on the quality of faculty of higher education.

H4. There is a significant difference between the perceptions of faculty and

students on the quality of higher education.

H5. There is a significant difference between the perceptions of faculty and

alumni on the quality of higher education.

H6. There is a significant difference between the perceptions of students

and alumni on the quality of higher education.

3.5.1 Sub-Hypothesis The sub-hypothesis is made based on the main directional

hypothesis. Sub-hypothesis details the relationship between the variables

such as age, gender, educational qualifications, year of passing and industrial

experience and the perceptions of stakeholders on the quality of higher

education.

59

H7: There is a significant relationship between age, educational

qualification and industrial experience of the Industry personnel and

their perceptions on the quality (academic performance) of higher

education in India.

H8: There is a significant relationship between age, gender and industrial

experience of the Industry personnel and their perceptions on the

quality (academic performance) of higher education in India.

H9: There is a significant relationship between age, year of passing and

industrial experience of the industry personnel and their perceptions on

the quality (Intelligence) of higher education in India.

H10: There is a significant relationship between educational qualifications of

the Industry personnel and their perceptions on the quality

(Learning Skills) of higher education in India.

3.6 Research Design

The design of this study is explained with the detailed instrumentation

process where the operationalisation of variables is described. The various

60

validities required for research work and the reliability of the study are

explained. The validities include the external validity which stresses the

generalizability of samples, construct validity which validates the

measurement model that it measures what it intends to measure, face and

content validity to ensure that the instrument covers all the variables which

could exhaustively measure the intended concept.

Reliability of the study explains the dependability of the measuring

instrument. Beyond these, the design defines the nature of the research, the

sampling frame, the population base from which samples are derived, the

sampling procedure, method of selecting the samples, the data collection

protocol which mentions about the method of collecting the filled up

questionnaires and finally defines the independent, dependent and control

variables used in the study. This part of the research gets the crucial focus, as

it gives the skeletal structure to the whole research work.

3.6.1 Instrumentation -- (Validation of Quality Criteria) This part of the design gives the whole explanation about the

development of the research instrument. The perception of stakeholders on

the quality criteria will be measured using the questionnaire as developed by

Donald, J. G., & Denison, D. B. 2001. This questionnaire is an ideal

instrument to measure the perceptions of stakeholders on the quality of higher

education because these criteria had been identified by a broad range of

61

stakeholders in a national study of criteria and indicators of quality in higher

education (Donald & Denison, 1993; Nadeau et al., 1992).

The stakeholders were governors of university and college boards,

administrators, faculty, students, and members of the larger community. They

independently identified criteria of quality in universities and colleges that they

considered to be important, and then verified the criteria over three rounds of

a Delphi procedure (Linstone & Turoff, 1975). Moreover, as the Indian context

differs from the western context, it needs a customised instrument taking into

count the local value systems. So, it demands a separate validation process

to develop a unique instrument to cater to the Indian higher educational

institutions’ requirement and to measure their quality. The 33 criteria for the

quality of students and the 20 criteria for the quality of faculty that emerged

from this study reflected the perspective of the broad higher education

community pertaining to Indian higher education, but not specifically students'

own views.

3.6.2 Operationalisation of the Quality Variables Operationalisation means successfully translating the concepts that

we have in mind into simple measurable terms. It is defined as a method of

portraying a concept in terms of the procedure used to measure it. It is also

referred to as the process in which the variables are identified as specific data

and facts, suitable for recordable/repeatable acts (operations & combinations)

to be performed on them (e.g. Defining, locating, measuring etc.).

62

Le Compete & Schensul (1999) define the operationalisation as

defining a concept concretely in such a way that it can be understood,

observed or categorized accurately by any researcher reviewing the same

data or observing the same event. So, the instrumentation process requires a

thorough operationalisation of the variables and the subjects of the study. The

operationalisation process starts with the translation validity and ends with

construct validity of the variables.

3.6.3 Method of Measurement

The identified variables have to be measured using an appropriate

method. Sackmann (1991) argues that there are different methods to

measure the quality in an organisation. The methods can be arranged in a

continuum. One end of the continuum comprises inductive modes of enquiry

like ethnographic studies through participant observation, in-depth interviews

and so on, the other end of the continuum consists of structured methods like

Questionnaires, Structured Interviews, Check list and so on (Sackmann,

1991). The ethnographic perspective often cited as the insider’s perspective is

considered to give complete face validity, richness of data, contextualization

of data and thorough insight through empathic listening. But, it faces

difficulties in terms of cost, researcher imposing his/her own views, analysis of

data, impossibility in meeting every stakeholder and difficulties in comparison

between organisations (Sackmann, 1991).

63

The questionnaire and other methods often cited as outsider’s/external

perspective allows for statistical treatment of data, complete enumeration from

all employees in an organisation cheap and faster and gives scope for

comparison between one organisation to another organisation. But, it also

carries disadvantages in the form of lack of contextuality, impersonal

approach, and therefore, fewer chances for sensitive questions being

answered (Sackmann, 1991).

As the existing design is meant for eliciting the perception of

stakeholders on the quality of higher education and to find the perception gap

between the stakeholders, a questionnaire method of measurement is

sufficient. Further, in many of the studies, quantitative questionnaires alone

are used (Hofstede, Neuijen, Ohayav & Sanders, 1990; House et al., 1999;

Smith, Misumi, Tayeb, Peterson & Bond, 1989; Smith, Dungan &

Trompenaars, 1996).

The measurement scale for the questionnaire is on a normative scale

basis as the ipsative formats (Paired Comparison & Q-Sort techniques) are

found problematic (Cable & Judge, 1997; Tepeci, 2001). A simple Likert type

scaling is found to be less time-consuming and easily interpretable (Tepeci,

2001). So, all the 33 items for Students and 20 items for Faculty are

developed into value statements to reduce socially desirable responses. Such

questions are randomized and simplified for clarity and simplicity of the

questionnaire. Each question is measured using a Likert scale that consists of

five points. Nunnally (1967) maintains that the reliability of a tool increases

64

with the number of scale points.Further, Nunnally (1967) opines that a graphic

scale with numbers will be preferable. The questionnaire is designed in two

parts, the first part measuring the perceptions of stakeholders on the criteria

for the quality of students and the second measuring the criteria for the quality

of faculty. The first part has 33 items and the second part has 20 items to

verify the correlation between the items. The scale measure for the first part is

a 5-point Likert Scale which measures each item from Not at all Important (NI)

to Extremely Important (EI). Similarly, the second part also measures

perceptions on the criteria for the quality of faculty in a 5-point scale ranging

from Not at all Important (NI) to Extremely Important (EI).

Not at all (NI) Somewhat (SI) (I) Quite (QI) Extremely (EI) Important Important Important Important Important

Figure 3.2 Measurement Scale

3.6.4 Pilot Study

After validating the questionnaire items for face and content

adequacies, it necessitates to check the criterion related validity of the

variables concerned, in the research design process. This part carries

importance as the construct validity and predictive validity of the model

involved in the study are verified.

1 2 3 4 5

65

The pilot study is conducted to finalise the list of relevant items which

can be further factor analysed. This will identify the hidden factor structure

involved with the items. Also, the reliability of the scale and relevant subscales

can be verified. The independent variable -- the demographic data of various

stakeholders (Independent) and its predictive power on any of the criteria for

the quality of Higher Education (dependent variables) -- can be identified. This

will prove the predictive validity of the construct involved.

3.6.4 a Construct Validity

Construct validity refers to the degree to which the inferences can be

legitimately made from the operationalisation in the study to the theoretical

constructs on which the operationalisation is based. In a simple sense, it will

check whether the operationalisation measures what it intends to measure

(Trochim, 2000). It will assess how well the ideas of the researcher are

translated into actual programmes or measure. It comprises two parts, namely

convergent validity and discriminant validity.

Convergent validity is the degree to which the operationalisation is

similar to (converges on) the other operationalisation that it theoretically

should be similar to. The contrary is true for proving discriminant validity

where the operationalisation is dissimilar from other operationalisation that it

theoretically should not be similar to (Trochim, 2000). Guion (1965) defines

construct validity as the degree to which the variance in a given set of

measures is due to the variance in the underlying construct. The factors

66

derived from factor analysis are constructs and the operational definition of

construct validity is the factor loading. This permits a specific numerical

statement of construct validity that is important for both criterion and predictor

measurement.

So, by means of factor analysis it is possible to construct a test giving a

relatively pure measurement of specific theoretical construct. This is achieved

by a factor analysis of the items in the test that individually are considered as

variables. Smit (1991) specifies that the analysis of internal factor structure of

the variables culminates in a factor loading, leading to a measure of a specific

construct.

3.6.4 b Predictive Validity

It is the process of assessing the operationalisation ability to predict

something it theoretically is able to predict (Trochim, 2000). Here, the

construct’s predictive ability in determining the effect size of the dependent

variable is counted as the predictive power/validity. Existence of predictive

ability will show that the measure can correctly predict something, which when

theoretically assumed, should be able to predict.

The pilot study is made with the following objectives:

• To identify the relevant quality variables which are highly correlated to

each other and to remove the distant items which are not closely

correlated to other items

67

• To figure out the initial factor structure of the variables so that the

identified factors will show Convergent and Discriminant validity

through their factor loadings showing construct validity

• Identifying the subscale reliability, mean and variance of the

construct .Measuring the predictive power of the construct by studying

the relationship between the quality criteria and perceptions of various

stakeholders

• To arrive at a preliminary questionnaire to measure the perceptions of

stakeholders on the quality of higher education in India.

With the above objectives, the pilot study is carried out to measure the

the perceptions of various stakeholders on the criteria for quality of higher

education.

3.6.4 c Methodology of Pilot Study

The questionnaire deployed for the pilot study is divided into two parts.

The first part measures the perception of stakeholders on the quality of

students with the face validated 33 items and the second part measures the

perceptions on the quality of faculty with 20 items. The Coronach Alpha

measure for the instrument are α =0.837 for Students and 0.8 for faculty

respectively which are high in reliability (Tepeci, 2001).

For the pilot study three, Engineering colleges offering engineering and

management courses affiliated to Anna University, Chennai Zone X have

68

been selected in random. By covering the colleges offering professional

courses of both engineering and management in the preliminary study, the

representativeness of the instrument is assured across a broad set of

samples. Hundred students and twenty five faculties are taken for the study to

ensure sampling adequacy which will enable further statistical enquiry.

3.6.4 d Outcomes of Pilot Study

For the preliminary analysis, the factor structure for the perceptions of

stakeholders on quality items alone have been found out, as the objective of

the study primarily is to filter the more relevant items for further studies.

Deletion of irrelevant items based on KMO Statistic

Using SPSS 13.0 software, the values of all the 33 items of the first

part of the instrument and 20 items of the second part of the instrument are

tested for Keyser-Meyer-Olkin (KMO) statistic and the Anti Image Correlation

(AIC) Matrix respectively. In the AIC matrix, the individual MSA (Measures of

Sample Adequacy) items whose values lie above 0.6 are filtered out. The

Initial KMO Measure is 0.642 for all 33 items of first Part and 0.715 for all 20

items of the second part. As a result of MSA value filter, 23 items of Part I and

18 items of Part II emerged with high KMO measure which are used in

subsequent analyses for factor structure. The 23 items of PartI and 18 items

of Part II give the overall KMO measure of 0.842 and 0.852 respectively.

69

Table 3.1

Data Reduction for Questionnaire Construction

Instrument α KMO No. of Variables

Before Filtering the Data Students 0.837 0.642 33

Faculty 0.8 0.715 20

After Filtering the Data Students 0.832 0.842 23

Faculty 0.831 0.858 18

3.7 Population, Sample and Data Collection

The higher education in India is undergoing a massive change with the

mushrooming of private colleges. Rapid industrialisation and globalisation

have resulted in a sudden rise in the demand for high quality higher education

in India. It is one of the major priorities in the field of education now

(Bhattacharya, 2004).

2.5 million Graduates and 3, 50,000 Engineers are spewing out every

year in India from 300 Universities and 15,600 colleges. Of the 1,300

engineering colleges in India, Tamilnadu itself boasts of more than 300

colleges. There are 20 Engineering Colleges in the Zone X of Anna

University, Tamilnadu. Around 10,000 students and 1,000 faculties belong to

70

the University and Industries of Chennai and Bangalore and alumni of those

colleges form the population of the study.

Initially, the sample size required for the study is measured. Sample

size of a study is a function of proposed analyses, namely factor and

regression analysis Tepeci (2001). Hair, Anderson, Tatham, and Black (1998)

argue that factor analysis requires 5 to 10 subjects per questionnaire item.

Guadagnoli, Edward and Wayne F. Velicer (1988) recommend a minimum

sample size of 100-200 for conducting factor analysis based on the results of

numerous simulation studies.

x = Z2 ( c/100) x r(100-r)

n = Nx / ( N-1) E2 + x)

E = [(N-n) x / n (N- 1)] 1/2

n = Sample Size E= Margin of Error N= Population Size

r = Fraction of responses the researcher is interested in

Z2 ( c/100) = Critical value for a confidence interval ‘c’

Figure 3.3

Sample Size Estimator Formula (Based on Normal Distribution)

As per Comrey (1988), a sample size of 200 is adequate for factor

analysis involving not more than 40 items. Moreover, to get a sample size

71

which is good enough to represent a normal distribution, Raosoft calculator7 is

used. It is based on the measurement model which uses normal distribution

as shown in Figure 3.3.

The margin of error is assumed as 5 %. It represents the amount of

tolerable error. The confidence level is fixed at 95 % which indicates the

amount of tolerable uncertainty. As per this method, the minimum sample size

recommended by the calculator for the study is 377 for students and 278 for

faculties. This also more than satisfies the requirements for factor analysis

and further regression analysis, as the items of the questionnaire are 23 and

18 only.

The sampling procedure is made on a multi-stage random method

where samples are drawn in two stages. Initially, from the 20

engineering colleges affiliated to Anna University, Chennai Zone X, six

colleges are selected on a random basis. Such selected colleges are

approached for permission from the Principal/Head of the Institution and

based on that status these colleges are included in the sampling basket. In

the whole process, all colleges gave permission for data collection. In the

second stage, the selected colleges’ faculty and students are selected in a

random way without any priory control. At least, 25 % of each college

population is covered to ensure aggregate level/cross level analysis, as a

reasonable number of respondents are required per college to get the

aggregate score for each college. The number of respondents per college

7 http://www.raosoft.com/samplesize.html

72

varied as the number of faculty and students also vary for each college. The

sampling procedure is followed and total sample size for faculty and students

reaches 201 (Response rate is 60%) and 487 (Response rate is 80%)

respectively.

The data collection protocols are through direct approach and

enumerator methods. In the direct approach method, the researcher

personally went to the selected colleges and distributed the questionnaires.

Although personal contact method is followed in certain cases, predominantly

data collection is through enumerators who are the college faculty members

(known to the researcher), working in the engineering colleges coming under

the sample basket. They helped in collecting data from their respective

colleges. Here, the questionnaires were mailed to the enumerators and they,

in turn, distributed the questionnaires in random, based on the college pay roll

name list.

The instructions given to the enumerators are that they have to cover

at least 25 % of the faculty in their institution, excluding them. Once they

collected the sample responses, they mailed it back to the researcher. The

enumerators carried out the task, out of the courtesy they showed to the

researcher as a friend. The resultant sample responses collected from the

faculty and students of institutions thus arrived at 201 and 487 respectively

and each college’s sample coverage percentage varied from 25 to 30.

Questionnaires were also mailed to alumni of randomly selected colleges and

73

Industry personnel in Chennai and Bangalore through email and responses

were received from 160 alumni and 100 Industry personnel.

In one section of the questionnaire, faculty, students, alumni and

industries were presented with the set of 23 criteria for quality of students and

in another section with the set of 18 criteria for the quality of faculty. faculty,

students, alumni and Industries were asked to use a 5-point response scale

(1 = not at all important, 2 = somewhat important, 3 = important, 4 = quite

important, 5 = extremely important) to indicate how important they felt each

criterion was for evaluating the quality of a student and faculty. Factor

analysis and the reliability test were conducted on the collected data and the

tables were formulated (Table 3.2 – 3.9).

Table 3.2 depicts the data collected on the perceptions of faculty on the

criteria for the quality of students. Principal components analysis of the

composite ratings on data collected was done utilizing a varimax rotation

extracted seven factors. The first factor encompassed 4 of the 23 criteria.

Loading most strongly on this factor, branded Academic Performance, were

the following criteria: (a) Personal Student Development, (b) Completion of

Program requirements, and (c) Expertise at the end of the program. The

second factor, labelled Social Responsibility, included (a) Commitment to Life

Long Learning (b) Commitment to Physical Fitness, (c) Leadership Skills and

(d) Commitment to Social Concerns.

74

Table 3.2

Criteria for Quality of Students according to Faculty Perception Reliability Test: Alpha = 0.873 KMO = 0.848

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance ACADEMIC PERFORMANCE

5 Commitment to Learning I 0.500 R

0.744 3.766 0.964

11 Personal Student Development

I 0.666 R

12 Completion of Program requirements

I 0.763 R

13 Expertise at the end of the program

I 0.673 R

SOCIAL RESPONSIBILITY 16 Commitment to Life Long

Learning II 0.654 R

0.622

3.641

1.115

17 Commitment to Physical Fitness

II 0.581 R

18 Leadership Skills II 0.508 R 23 Commitment to Social

Concerns II 0.657 R

COMMUNICATION SKILLS 20 Written communication skills III 0.799 R

0.704 4.0

0.890 21 Presentation skills III 0.653 R

22 Oral Communication skills III 0.752 R LEARNING SKILLS

4 Intelligence IV 0.394 R 0.7

3.69

0.986

7 Openness and Flexibility IV 0.561 R 8 Ability to interact with others IV 0.740 R 9 Effective study skills & habits IV 0.560 R 10 Moral & Ethical Reasoning IV 0.529 R

ACADEMIC PREPAREDNESS 2 Preparedness for a specific

Program V 0.717 R

0.577

3.862

0.832

6 Sense of Responsibility V 0.718 R 19 Ability to apply knowledge V 0.504 R

EMPLOYMENT COMPETENCE 14 Ability to get a job VI 0.761 R

0.590 4.152 0.995 15 Performance on the job VI 0.632 R

GENERIC SKILLS 1 Secondary School

Preparation VII 0.757 R

0.475 3.652

1.152

3 Basic Mathematical Competency

VII 0.689 R

75

A third factor expressed as Communication Skills consisted of

(a) Written communication skills, (b) Presentation skills, and (c) Oral

Communication skills. Openness and Flexibility, Ability to interact with others,

Effective study skills and habits and Moral and Ethical Reasoning were loaded

together to form a fourth factor, Learning skills. The fifth factor, academic

preparedness, consisted of (a) secondary school preparation, (b) general

academic preparedness, and (c) preparedness for a specific program. The

sixth factor named Employment Competence included (a) Ability to get a job

and (b) Performance on the job. Secondary School Preparation and Basic

Mathematical Competency form the seventh factor named Generic Skills.

Academic Competence, Team bonding skills, Interpersonal Skills and

Presentation Skills were found as the factors for the criteria for the quality of

faculty according to the perceptions of faculty as shown in Table 3.3. The first

factor consisted of seven of eighteen criteria. Ability to Explain Clearly, Depth

of Knowledge, Commitment to knowledge updation and Commitment to

research are the strongly loaded criteria on the factor. The second factor

named Team Bonding Skills consisted of five criteria, namely (a) Ability to

work as a Team Member (b) Problem Solving Skills (c) Leadership Skills (d)

Commitment to Ethical Values and (e) Commitment to Social Concerns.

Ability to encourage the students, Approachability by Students and

Enthusiasm were loaded together to form the third factor labelled as

Interpersonal Skills. The fourth factor branded as Presentation Skills include

(a) friendly with students, (b) temperament and (c) sense of humour.

76

Table 3.3

Criteria for Quality of Faculty according to Faculty Perception

Reliability Test: Alpha = 0.882 KMO = 0.866

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance ACADEMIC COMPETENCE

1 Ability to Explain Clearly I 0.717 R 0.796

4.151

0.830

3 Depth of Knowledge I 0.662 R 4 Presentation Skills I 0.698 R 6 Ability to use Computer &

Technology I 0.512 R

10 Commitment to Knowledge updation

I 0.574 R

14 Confidence I 0.548 R 18 Commitment to research I 0.575 R

TEAM BONDING SKILLS 7 Ability to work as a Team

Member II 0.614 R

0.795

3.79

1.015

8 Problem Solving Skills II 0.541 R 9 Leadership Skills

II 0.415 R

11 Commitment to Ethical Values

II 0.672 R

12 Commitment to Social Concerns

II 0.769 R

INTERPERSONAL SKILLS 2 Ability to Encourage the

students III 0.521 R

0.678

3.9

0.8 5 Approachable by

Students III 0.775 R

13 Enthusiasm III 0.620 R PRESENTATION SKILLS

15 Friendliness with Students

IV 0.759 R

0.649 3.244 1.174 16 Temperament IV 0.609 R 17 Sense of Humour IV 0.739 R

77

Table 3.4 illustrates the data collected on the perceptions of students

on the criteria for the quality of students. Principal components analysis of the

composite ratings on the data were collected utilizing a varimax rotation

extracted seven factors. The first factor covered 5 of the 23 criteria. Loading

most strongly on this factor, branded communication skills were (a) written

communication skills (b) presentation skills, and (c) oral communication skills.

The second factor, labelled Learning Skills, included four criteria among them

(a) Commitment to Life Long Learning and (b) Commitment to Physical

Fitness were strongly loaded.

A third factor expressed as Generic Skills consisted of (a) Sense of

responsibility (b) Openness and Flexibility, and (c) Ability to interact with

others. Ability to get a job and performance on the job were loaded together to

form a fourth factor, Employment competence. The fifth factor, academic

Performance, consisted of (a) Moral and ethical reasoning, (b) Personal

Student Development ,(c) Completion of program requirements and (d)

Expertise at the end of the program. The sixth factor named academic

preparedness consisted of three criteria and among them (a) Secondary

school preparation and (b) Preparedness for a specific program were strongly

loaded. The seventh factor named Intelligence included (a) Intelligence and

(b) Effective study skills and competence.

78

Table 3.4

Criteria for Quality of Students according to Students’ Perception


Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance COMMUNICATION SKILLS

18 Leadership Skills I 0.513 R

0.686 3.933 1.006 19 Ability to apply knowledge I 0.561 R 20 Written communication skills I 0.624 R 21 Presentation skills I 0.785 R 22 Oral Communication skills I 0.611 R

LEARNING SKILLS 5 Commitment to Learning II 0.488 R

0.554

3.51

1.321

16 Commitment to Life Long Learning

II 0.703 R


II 0.620 R


II 0.422 R

GENERIC SKILLS 6 Sense of Responsibility III 0.581 R

0.613 3.91

1.027 7 Openness and Flexibility III 0.724 R

8 Ability to interact with others III 0.738 R EMPLOYMENT COMPETENCE

14 Ability to get a job IV 0.765 R 0.618 4.256 0.888

15 Performance on the job IV 0.672 R ACADEMIC PERFORMANCE

10 Moral & Ethical Reasoning V 0.577 R 0.575

3.556

1.152

11 Personal Student Development

V 0.672 R


V 0.621 R


V 0.397 R

ACADEMIC PREPAREDNESS

1 Secondary School Preparation VI 0.763 R

0.454 3.7 1.046 2 Preparedness for a specific program

VI 0.587 R


VI 0.487 R

INTELLIGENCE 4 Intelligence VII 0.584 R

0.381 3.807 0.964 9 Effective study skills & habits VII 0.626 R

79

The results of factor analysis carried out on the perceptions of students

on criteria for the quality of faculty are shown in Table 3.5. Social

Responsibility, Interpersonal Skills, Presentation Skills, Academic

Competence, and Research Skills were found as the five factors.

The first factor consisted of four of eighteen criteria. Commitment to

social concerns and Commitment to ethical values were the strongly loaded

criteria on the factor. The second factor branded as Interpersonal Skills

consisted of seven criteria and (a) Friendliness with the students (b)

Enthusiasm and (c) Sense of humour were the robustly loaded on the factor.

Ability to use computer and technology and ability to work as a team member

loaded strongly on the factor labelled as Presentation Skills. The fourth factor

branded as Academic Competence includes (a) ability to explain clearly and

(b) depth of knowledge. Commitment to research is the only criteria loaded

adequately for the factor named research skills.

Table 3.6 demonstrates the data collected on the perceptions of alumni

on the criteria for the quality of students. Principal components analysis of the

composite ratings on data collected utilizing a varimax rotation extracted eight

factors. The first factor covered 6 of the 23 criteria and is called academic

performance. Completion of program requirements and Expertise at the end

of the program were loaded significantly on the factor. The second factor

named as Communication skills includes the criteria (a) written

communication skills (b) presentation skills, and (c) oral communication skills

80

with strong loading. The third factor, labelled Intelligence consisted of three

criteria.

Table 3.5

Criteria for Quality of Faculty according to Students’ Perception


Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance SOCIAL RESPONSIBILITY

9 Leadership Skills I 0.565 R 0.707

3.678

1.192

10 Commitment to Knowledge updation

I 0.553 R


I 0.776 R


I 0.746 R

INTERPERSONAL SKILLS

2 Ability to Encourage the students

II 0.478 R 0.705

3.954

0.973

5 Approachable by Students

II 0.366 R

13 Enthusiasm II 0.617 R

14 Confidence II 0.433 R 15 Friendliness with

Students II 0.719 R

16 Temperament II 0.484 R 17 Sense of Humour II 0.676 R

PRESENTATION SKILLS

4 Presentation Skills III 0.361 R 0.609

3.813

1.047

6 Ability to use Computer & Technology

III 0.682 R

7 Ability to work as a Team Member

III 0.727 R

8 Problem Solving Skills III 0.520 R ACADEMIC COMPETENCE

1 Ability to Explain Clearly

IV 0.775 R 0.490

4.336

0.803

3 Depth of Knowledge IV 0.693 R REASERCH SKILLS

18 Commitment to Research V 0.690 R

81

The fourth factor named academic preparedness, consisted of two

criteria (a) secondary school preparation and (b) preparedness for a specific

program. Employment competence, social responsibility, learning skills and

generic skills were the other four factors structured by the factor analysis on

the perceptions of alumni on the criteria for the quality of students.

Commitment to social concerns is the significantly loaded criteria on the factor

called social responsibility. The seventh factor named as learning skills was

strongly loaded with the criteria of ability to apply knowledge. The eighth

factor labelled as generic skills was very strongly loaded with the criteria of

openness and flexibility.

The factor analysis carried out on the perceptions of alumni on the

criteria for the quality of Faculty and six factors were structured and shown in

Table 3.7. Enthusiasm, Presentation Skills, Team Bonding Skills, Social

Responsibility, Academic Competence and Interpersonal Skills were found as

the six factors. The first factor consisted of four of eighteen criteria.

Enthusiasm was the strongly loaded criteria on the factor. The second factor

branded as presentation Skills consisted of three criteria and (a) ability to

explain clearly and (b) depth of knowledge skills were the robustly loaded on

the factor. Ability to work as a team member and ability to use computer and

technology loaded strongly on the factor labelled as team bonding skills. The

fourth factor branded as social responsibility includes (a) commitment to

ethical values and (b) commitment to social concerns with strong loading. The

sixth factor named as Interpersonal skills was very strongly loaded with the

criteria, ability to encourage the students and approachability by students.

82

Table 3.6

Criteria for Quality of Students according to Alumni Perception Reliability Test : Alpha = 0.859 KMO = 0.795

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance ACADEMIC PERFORMANCE

8 Ability to interact with others I 0.463 R 0.720

3.716

1.137

10 Moral & Ethical Reasoning I 0.414 R 11 Personal Student

Development I 0.571 R


I 0.719 R


I 0.609 R

14 Ability to get a job I 0.635 R COMMUNICATION SKILLS

20 Written communication skills II 0.716 R 0.655

3.89

1.187 21 Presentation skills II 0.624 R

22 Oral Communication skills II 0.671 R INTELLIGENCE


III 0.554 R 0.587

3.785

1.238

4 Intelligence III 0.697 R 5 Commitment to Learning III 0.631 R

ACADEMIC PREPAREDNESS 1 Secondary School

Preparation IV 0.672 R

0.55 3.341

1.36

2 Preparedness for a specific program

IV 0.677 R

EMPLOYMENT COMPETENCE 6 Sense of Responsibility V 0.638 R

0.576 3.908

1.025 9 Effective study skills & habits V 0.674 R

15 Performance on the job V 0.506 R SOCIAL RESPONSIBILITY


VI 0.635 R 0.398 3.447 1.328


VI 0.790 R

LEARNING SKILLS 16 Commitment to Life Long

Learning VII 0.525 R

0.522 3.821

1.156

18 Leadership Skills VII 0.578 R 19 Ability to apply knowledge VII 0.683 R

GENERIC SKILLS 7 Openness & Flexibility VIII 0.826 R

83

Table 3.7

Criteria for Quality of Faculty according to Alumni Perception

Reliability Test : Alpha = 0.850 KMO = 0.805

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance ENTHUSIASM

13 Enthusiasm I 0.772 R

0.736 3.944 1.229 14 Confidence I 0.465 R 15 Friendliness with

Students I 0.743 R

17 Sense of Humor I 0.650 R PRESNTATION SKILLS

1 Ability to Explain Clearly

II 0.688 R

0.710 4.138 0.966 3 Depth of Knowledge II 0.645 R 4 Presentation Skills II 0.488 R

TEAM BONDING SKILLS 6 Ability to use Computer

& Technology III 0.741 R

0.699

3.720

1.25


III 0.794 R

8 Problem Solving Skills III 0.494 R

9 Leadership Skills III 0.518 R SOCIAL RESPONSIBILITY


IV 0.810 R 0.713

3.519

1.347


IV 0.771 R

ACADEMIC COMPETENCE 10 Commitment to

Knowledge updation V 0.471 R

0.488 3.546 1.201 16 Temperament V 0.687 R 18 Commitment to

Research V 0.630 R

INTERPERSONAL SKILLS

2 Ability to Encourage the students

VI 0.752 R 0.645 4.10 1.035

5 Approachable by Students

VI 0.772 R

84

Commitment to knowledge updation, temperament and Commitment to

research were the only criteria loaded heavily for the fifth factor named

academic competence.

Table 3.8 shows the data collected on the perceptions of Industries on

the criteria for the quality of students. Principal components analysis of the

composite ratings on data collected utilizing a varimax rotation extracted eight

factors. The first factor encompassed 3 of the 23 criteria. Loading most

strongly on this factor, branded communication skills were (a) Written

communication skills, (b) Presentation skills, and (c) Oral Communication

skills.. The second factor, Academic performance, included openness and

flexibility, (b) Completion of Program requirements, and (c) Expertise at the

end of the program. A third factor expressed as Generic Skills and consisted

of (a) Basic Mathematical Competency, (b) Sense of responsibility and (c)

ability to interact with others. Effective study skills and habits, Moral and

Ethical Reasoning, Commitment to Life Long Learning and ability to apply

knowledge form the fourth factor named as learning skills.

Commitment to social concern was the criteria strongly loaded for the

fifth factor called social responsibility. The sixth factor labelled as employment

competence includes the criteria, ability to get a job and the performance on

the job with strong loading. The seventh factor, academic preparedness, was

strongly loaded with criteria (a) secondary school preparation and (b)

preparedness for a specific program.

85

Table 3.8

Criteria for Quality of Students according to Industry Perception Reliability Test: Alpha = 0.832 KMO = 0.732

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance COMMUNICATION SKILLS

20 Written Communication Skills I 0.789 R 0.781

4.13

0.723 21 Presentation skills I 0.742 R

22 Oral Communication skills I 0.818 R ACADEMIC PERFORMANCE

7 Openness and Flexibility II 0.573 R

0.606 0.37 0.85 12 Completion of Program

requirements II 0.712 R


II 0.714 R

GENERIC SKILLS


III 0.684 R

0.594 4.017 0.778 6 Sense of Responsibility III 0.757 R 8 Ability to interact with

others III 0.635 R

LEARNING SKILLS 9 Effective study skills & habits IV 0.487 R

0.62 3.86 0.881 10 Moral & Ethical Reasoning IV 0.555 R 16 Commitment to lifelong

learning IV 0.686 R

19 Ability to apply knowledge IV 0.795 R SOCIAL RESPONSIBILITY

17 Commitment to Physical V 0.618 R 0.515 3.41 0.980 23 Commitment to Social

Concerns V 0.763 R

EMPLOYMENT COMPETENCE 11 Personal Student

Development VI 0.411 R

0.687

4.00 0.725 14 Ability to get a job VI 0.846 R

15 Performance on the job VI 0.722 ACADEMIC PREPAREDNESS

01 Secondary School Preparation

VII 0.806 R

0.313 3.44 1.159 02 Preparedness for a specific program

VII 0.675 R

18 Leadership Skills VII 0.407

INTELLIGENCE 04 Intelligence VIII 0.576 R

0.542 4.01 0.80 05 Commitment to Learning VIII 0.447 R

86

The factor analysis carried out on the perceptions of Industries on

criteria for the quality of Faculty and five factors were structured and shown in

Table 3.9. Presentation Skills, Social Responsibility, Academic Competence,

Interpersonal Skills and Team Bonding Skills were found as the five factors.

The first factor consisted of four of eighteen criteria. Friendliness with

the students was the strongly loaded criteria on the factor. The second factor

branded as social responsibility consisted of four criteria with strong loading

for commitment to social concerns. (a) Ability to explain clearly and (b) depth

of knowledge skills (c) Problem solving skills and (d) commitment to

knowledge updation were robustly loaded on the factor branded as academic

competence. Ability to encourage the students was the criteria strongly loaded

for the fourth factor called interpersonal skills. Ability to work as a team

member and ability to use computer and technology were loaded strongly on

the factor labelled as team bonding skills.

87

Table 3.9

Criteria for Quality of Faculty according to Industry Perception

Reliability Test : Alpha = 0.866 KMO = 0.808

Factor Analysis

S.No Questionnaire Factor Loading Alpha Mean Variance PRESENTATION SKILLS

4 Presentation Skills I 0.559

0.676 4.128 0.703 5 Approachable by

Students I 0.663

9 Leadership Skills I 0.572 R 15 Friendliness with

Students I 0.847 R

SOCIAL RESPONSIBILITY 12 Commitment to Social

Concerns II 0.688 R

0.732 3.5 1.066 16 Temperament II 0.645 R

17 Sense of Humour II 0.488 R 18 Commitment to

Research II 0.745 R

ACADEMIC COMPETENCE 1 Ability to Explain

Clearly III 0.643 R

0.714 4.308 0.542 3 Depth of Knowledge III 0.754 R 8 Problem Solving Skills III 0.549 R 10 Commitment to

Knowledge updation

III 0.641 R

INTERPERSONAL SKILLS 2 Ability to Encourage

the students IV 0.757 R

0.666 4.193 0.67 11 Commitment to Ethical

Values IV 0.477 R

14 Confidence IV 0.593 R 13 Enthusiasm IV 0.536 R

TEAM BONDING SKILLS

06 Ability to use Computer & Technology

V 0.758 R 0.594 3.81 0.815


V 0.735 R

88

3.8 Variables Involved in the Study

The operationalisation of the control, independent variables and

dependent variables are explained in this section.

3.8 a Control Variables

A control variable is a variable that affects the dependent variable. By

"controlling a variable", it is possible to balance its effect across subjects and

groups so that one can ignore it, and just study the relationship between the

independent and the dependent variables.

3.8 b Independent Variables

The independent variable is considered to be the programme or the

cause or the treatment (Trochim, 2000). If it is manipulated for the study, the

independent variable is considered to be ‘active’ and if not, and studied as

such for its effect on another variable, then it is considered to be an ‘attribute’.

In this study, the demographic data, namely gender, age, educational

qualification, marital status and experience of the respondents are the

attribute independent variables. The demographic data of different

stakeholders, namely faculty, alumni and industries are collected.

i. Gender: Male / Female

ii Age:

89

iii Educational Qualification:

iv Marital Status:

v Designation:

vi Year of passing

vii Experience: In terms of years

From the category of students, demographic data such as gender, age,

degree, branch and year are collected. After the final phase of data collection,

the effects of these variables on the perception of stakeholders on the quality

of higher education are studied by using regression analysis. All the variables

are subject to factor analysis and the resultant factors and individual value

factors are calculated by adding the individual item variables under each

factor.

3.8 c Dependent Variables

The dependent variable is what is affected by the independent variable

(Trochim, 2000). It is called the criterion measure, the independent variable

being the experimental measure. The dependent variables are quality criteria.

Academic Performance, Social Responsibility, Communication Skills,

Learning Skills, Academic Preparedness, Employment Competence and

Generic Skills are the factors for measuring the quality of students. Academic

Competence, Team Bonding Skills, Interpersonal Skills, and Presentation

Skills are the factors for measuring the quality of faculty.

90

3.9 Data Analysis

This section of research design deals with the explanation of various

analyses carried out in this research work. The data after collection is entered

in the MS-Excel spread sheet and later the spread sheet is converted to a

‘.sav’ file to operate in the SPSS version 13.0. Initially, the data is checked for

omitted data, if any, and replaced with group means. Further, the data is

checked for the fulfilment of the basic assumptions of multivariate analysis,

namely linearity, normality and homoscedasticity.

3.9 a Exploratory Factor Analysis

The data collected in the final phase of the study is initially subjected to

an exploratory factor analysis using SPSS, to facilitate the process of eliciting

a factor structure. Here, Principal Components Analysis (PCA) with ‘Varimax’

rotation is used to identify the hidden factors/underlying factors. Here, the

factor structure obtained from the pilot study is not considered because the

sample size variation in the final study will change the factor structure

completely and the initial factor analytic study on the pilot data is only to know

the consistency of the data and to establish the validity of the construct

(Kerlinger, 1986; Guion, 1965).

The factor analysis is considered to be the queen of all multivariate

analyses, as it will help in establishing the construct validity of any tool. When

factor analysis is able to generate similar factors, when subject to new

91

samples comparing to the factors previously generated from an old sample, it

is said to be useful in hypothesis testing (Kerlinger, 1986). Moreover, factor

analysis will be useful to create summated scales which are useful in additive

analysis of the variables under each factor grouping and such additive

values/composite scores can be taken for interpretation on any of the factor

(Hair et al., 1998). So, the exploratory factor analysis is applied for the final

set of data to know the new factor structure emanating from the new set of

sample collected in the final phase of the study.

Before subjecting the data to exploratory factor analysis, the data is

tested for Kaiser-Meyer-Olkin (KMO) statistic which will analyse the sampling

adequacy of the data for factor analysis. The KMO statistic should not be less

than 0.6.

3.9 b Regression Analysis

Regression analysis is used to determine the functional relationship

between a dependent variable and a host of predictors. Multiple Linear

Regression is a logical extension of the simple linear regression analysis and

involves two or more independent variables forming the basis for estimating

the values of a dependent variable. The relationship between each

independent variable and the dependent variable is a linear one. The principal

advantage of multiple regression analysis is that it allows to use more of

available information to estimate the dependent variable (Dipak Kumar

Bhattacharya, 2006). The investigator assembles data on the underlying

92

variables of interest and employs regression to estimate the quantitative effect

of the causal variables upon the variable that they influence. The investigator

also typically assesses the “statistical significance” of the estimated

relationships, that is, the degree of confidence that the true relationship is

close to the estimated relationship

3.9 c Hierarchical Regression Analysis Hierarchical regression is used to evaluate the relationship between a

set of independent variables and the dependent variable, taking into count the

impact of different set of independent variables on the dependent variable.

Here, the independent variables are entered into the analysis in a sequence

of blocks/ groups that may contain one or more variables (Hair et al., 1998). A

hierarchical regression can have as many blocks as possible. A common

hierarchical regression analysis involves two blocks, one being control

variables and other a set of independent variables. Support for a hypothesis

would be expected to require statistical significance for the addition of each

block of variables and many times, the effect of blocks of variables previously

entered into the analysis will be excluded, whether or not a previous block is

statistically significant.

The analysis is aimed at obtaining the best indicator of the effect of the

predictor variables. The statistical significance of previously entered variables

is not interpreted (Hair et al., 1998). In the results analysis, the ‘R2 change’ is

used to confirm the hypothesis that the additional block of independent

93

variables is able to explain variance over and above the control/other

independent variable blocks.

3.9 d T-Test

The t-test is probably the most commonly used Statistical Data

Analysis procedure for hypothesis testing. Actually, there are several kinds of

t-tests, but the most common is the "two-sample t-test" also known as the

"Student's t-test" or the "independent samples t-test". The two sample t-test

simply tests whether or not two independent populations have different mean

values on some measure. From the data collected from the stakeholders, it

was observed that the two groups have different average scores. But does

this represent a real difference between the two populations, or just a chance

difference in samples? The statistics t-test allows to answer this question by

using the t-test statistic to determine a p-value that indicates how likely these

results could have been gotten by chance. By convention, if there is a less

than 5% chance of getting the observed differences by chance, the null

hypothesis is rejected and found a statistically significant difference between

the two groups is found.

3.9 e Correlation

In statistics, correlation (often measured as a correlation coefficient, r)

indicates the strength and direction of a linear relationship between

94

two random variables. That is in contrast with the usage of the term in

colloquial speech, which denotes any relationship, not necessarily linear. In

general statistical usage, correlation or co-relation refers to the departure of

two random variables from independence. In this broad sense, there are

several coefficients, measuring the degree of correlation, adapted to the

nature of the data. A number of different coefficients are used for different

situations. The best known is the Pearson product-moment correlation

coefficient, which is obtained by dividing the covariance of the two variables

by the product of their standard deviations

The following tools are used for analysis and hypothesis testing.

• To measure the gap between any two criteria, deviation (Difference in

individual factors sum) and correlation measure are used. This will

ensure that the results are methodologically triangulated.

• Moreover, the gap between any two groups is depicted through a

pictorial radar diagram to give a visual glimpse of the gap.

• Differentiating the differences in perception between any two groups in

the study is made through simple percentage analysis and later the

difference is established through student’s t-test.

• To identify the relationship between the demographic data of the

respondents and the quality perception factors, hierarchical

regressions is used.

chapter 3 data and methodology 3.1 statement of the...

Documents