int. j. production economics - biu · maintain competitive advantage high tech ﬁrms must...

Int. J. Production Economics 128 (2010) 413–426

Contents lists available at ScienceDirect

Int. J. Production Economics

0925-52

doi:10.1

n Corr

E-m

(C.-H. W

journal homepage: www.elsevier.com/locate/ijpe

Integrating hierarchical balanced scorecard with non-additive fuzzy integralfor evaluating high technology firm performance

Chun-Hsien Wang a,n, Iuan-Yuan Lu b, Chin-Bein Chen c

a Department of Bio-industry and Agribusiness Administration, National Chiayi University, Chiayi, Taiwan, No. 580, Sinmin Road, Chiayi City 600, Taiwan, ROCb Department of Business Administration, National Sun Yat-sen University, Taiwan, No. 70, Lienhai Road, Kaohsiung 80424, Taiwan, ROCc Department of Business Administration, National Sun Yat-sen University, No. 1, Section 2, Da Hsueh Road, Shoufeng, Hualien 97401, Taiwan, ROC

a r t i c l e i n f o

Article history:

Received 6 March 2006

Accepted 22 July 2010Available online 4 August 2010

Keywords:

Performance measurement

Performance measurement system

balanced scorecard

Hierarchical balanced scorecard

Fuzzy measure

Non-additive fuzzy integral

73/$ - see front matter Crown Copyright & 2

016/j.ijpe.2010.07.042

esponding author. Tel.: +886 5 273 2880; fax

ail addresses: [email protected], abs

ang).

a b s t r a c t

Efficient and accurate performance measurement systems serve as a useful tool enabling managers to

control, monitor and improve high technology firm processes, productivity and performance. A model is

developed for measuring the acceptable performance of high tech firms based on the interaction

financial, customers, internal business process and learning and growth perspective. The HBSC structure

integrated with non-additive fuzzy integral for designing, developing and implementing high

technology firms relevant to performance measurement was employed to overcome interaction among

the various perspectives. Sixteen samples from eight high tech firms are used throughout the study to

explain how the execution of the model works. Utilizing the proposed model, the fuzzy assessment of

the decision-maker and the interaction among various evaluation criteria can be a focus of the

evaluation of the aggregation performance, thus ensuring more effective and accurate performance

evaluation and decision-making. In the light of this empirical evidence, the results provide guidance to

high tech firms performance measurement in both identification appropriate metrics and overcoming

key implementation obstacles for improving firm-operating efficiency and hence assistance for future

strategic adjustment.

Crown Copyright & 2010 Published by Elsevier B.V. All rights reserved.

1. Introduction

In the current extremely competitive global environment, hightechnology firms have been becoming increasingly reliant on coreresources for maintaining long-term competitive advantage. Tomaintain competitive advantage high tech firms must recognizeand emphasize relevant, integrated, strategic, improvement orientedand whole performance measurement systems, doing so by adoptingvarious management philosophies and tools such as benchmarking,total quality management (TQM), and business process redesign(BPR) to help to define goals and performance expectations. Hightechnology firms must integrate and develop appropriate perfor-mance metrics to explain and quantitatively analyze the criteria usedto measure the effectiveness of the operational system and itsnumerous interrelated components. Balance scorecard (BSC) (Kaplanand Norton, 1992, 1996) has been developed to integrate perfor-mance measurement system with organizational goal, and alignsproduction, marketing, organization process, non-financial and tradi-tional functions with firm strategies using performance driver(leading indicators) and outcome measures (lagging indicators).

010 Published by Elsevier B.V. All

: +886 5 273 2874.

[email protected]

Consequently, the performance measurement system is entire andadopts a multidimensional structure perspective. Performance mea-surement is a multidimensional structure involving the variouscomponents which contribute differently to overall high technologyfirm performance. A systematic and efficient approach towardsperformance measurement is based on constructing a system modelwhich in turn relies on corporate cross-function evaluation ofperformance. Evaluation methods thus must be applied togetherwith numerous approaches to improve the accuracy of corporateperformance measurement. However, performance measurement isdifficult and complex and evaluators lack widely recognizedperformance measurement tools and well-defined criteria for makingaccurate measurements. Constructing and possessing availableperformance measurement tools not only increases evaluationefficiency but also saves costs. Traditional corporations generallyuse financial aspects to measure business performance, for examplereturn on assets (ROA), return on investment (ROI), return on sales(ROS), etc. However, those traditional performance measurementssuffer various limitations (Fisher, 1992; Ecceles, 1991). The mostsignificant limitation of traditional performance measurements is thatthey are based on financial perspectives, which emphasize theoperational results, but not the internal process would result inignoring forecasting function and which lacks a long-term orientation.

The financial aspect comprises only part of the firm perfor-mance measurement system. Particularly, the new operational

rights reserved.

www.elsevier.com/locate/ijpe

dx.doi.org/10.1016/j.ijpe.2010.07.042

mailto:[email protected]



dx.doi.org/10.1016/j.ijpe.2010.07.042

C.-H. Wang et al. / Int. J. Production Economics 128 (2010) 413–426414

manufacturing environment, global competition, ease of imita-tion, and the information revolution represent the main reasonsfor firm managers not only needing to focus on financialperformance but also changing their method of assessing overallperformance, for example by using measurement of internaloperational processes to monitor continuous quality promotion,processes improvement and innovation capabilities. In thisenvironment firms must integrate information on financial andnon-financial performance measurement systems to facilitatestrategy implementation and further predict long-term perfor-mance. To measure corporate financial and non-financial perfor-mance simultaneously, the BSC proves the ability of visibleperformance measurement approaches in strategy implementa-tion and management control (Kaplan and Norton, 1992). The BSCscheme integrates the interests of the key stakeholders, custo-mers and employees on a scoreboard (Kaplan and Norton, 1996).The essence of BSC lies in seeking a balance between financial andnon-financial measures.

Traditional financial indicators that use performance measureshave been criticized as inadequate for the present rapidlychanging business environment, especially when intangible assetsrather than tangible assets are the main sources of competitiveadvantage for high technology firms. These intangible assetsinclude customer satisfaction, process innovation capability, totalcost reduction/control capability, etc. To overcome the limitationsof financial-based measures, non-financial measures have beenrecommended owing to them being believed to be leadingindicators of financial performance (Kaplan and Norton, 1992,1996). Notably, practitioners and researchers have recommendedincreasing non-financial measures that reflect key value-creatingactivities, namely non-financial value drives (Kaplan and Norton,1992, 1996; Ecceles, 1991). Non-financial information is crucial inthe high technology industry, including telecommunications,biotechnology and software development (Amir and Lev, 1996).On the other hand, because of numerous non-financial indicatorsare difficult to quantify, including customer satisfaction, total costcontrol/management capabilities and employee productivities,yet they can significantly impact overall firm performancemeasurement. To overcome the limitations of traditional methodsof measuring performance and solve problems involving difficultto quantify non-financial indicators lack sufficient information,and difficult to measure accurately. However, the previousliterature seems to lack an objective definition and a consensusof the precise nature of performance measurement systems forhigh technology firms. This study introduces the subject byarguing about why firms need to assess performance, why theyneed to link performance measures to strategies or evenemphasize financial and non-financial performance of firmswithout clearly defining the nature of a performance measure-ment system, and where its virtue resides in terms of manage-ment control system. Therefore, the first objective of this study isto devise a framework for developing the hierarchical balancedscorecard (HBSC) performance measurement metrics in complexand competitive operational high tech environment.

Kaplan and Norton (1996) further contended that in the BSCprogram a cause-and-effect relationship exists between thefinancial and non-financial perspectives. BSC combines importantpractices and concepts from various disciplines and theories intoa single performance measurement system to improve financialperformance. Essentially, non-financial perspectives are leadingindicators, derived from establishing a causal link betweenimproved performance in terms of non-financial and financialmeasures. Based on BSC, four different perspectives would beaccurate the causal linkages between non-financial measures andfinancial measures, and focusing on improving leading indicatormeasures should improve the performance in terms of financial

measures. Employing the HBSC method of measuring hightechnology firm performance should consider the interactiverelationship between different perspectives. Thus, the secondobjective of this study is to solve the interactive impact throughwhich the non-additive fuzzy integral provides an appropriateapproach and process for handling interaction problems; thispaper utilizes a properly designed and implemented HBSCstructure, which should yield better results than alternativestrategies. The proposed framework designed by incorporatingthe HBSC structure with non-addition fuzzy integral to provide anoverview of high tech firm performance and prevent localoptimization. The contribution of this study lies in demonstratingthe limitations of a ‘green field’ approach in the development ofHBSC to help research and practice performance measurementand enhanced management effectiveness and efficiency. Thepresent HBSC performance measurement system is applied toovercome difficulties in performance measurement and focus onaligning the system of measuring high tech firm performancewith existing performance measures and parallel initiatives cross-functional performance measures for the particular high techfirm.

Recently, many researchers have been developed and modifiedfuzzy analytic hierarchical process (FAHP) and analytic networkprocess (ANP) approach in order to apply in diverse domains. Leeet al. (2008) integrate the fuzzy AHP and BSC to evaluateperformance of an information technology department in themanufacturing industry. Ravi et al. (2005) analyzed alternatives inreverse logistics for end-of-life computers by combining of BSCand ANP approaches. The ANP is used to structure the hierarchyand relative weightings of interdependent performance perspec-tives and indicators. Chan (2006) used the AHP and BSC inimplementing healthcare organizations performance assessment.The AHP is applied to calculate the relative weights for eachperformance assess. However, those approaches still cannotreflect the degree of interaction among performance evaluativeperspectives and indicators. Therefore, non-additive fuzzyintegral method was designed to solve the degree of interactionbetween performance perspectives and their correspondingperformance indicators within HBSC. The most importantcomponent of non-additive fuzzy integral is providinginformation integration capability within interdependency orinteractive characteristics without loss information. The proposednon-additive fuzzy integral incorporate HBSC performancemeasurement system could specifically design to mathematicallyrepresent uncertainty and vagueness and provide formalized toolsfor dealing with the imprecision intrinsic performance criteriaand the degree of interaction. Since fuzziness, vagueness andinteraction are coexisting characteristics existing in realitydecision-making processes. These coexisting characteristics madethe particular way of decision-making approach, for example,non-additive fuzzy integral is need rather general approach applyin the circumstance. Banker et al. (2004) combined a dataenvelopment analysis (DEA) based method with BSC analysis toassess interrelationships and tradeoff exist among alternativeperformance dimensions in the US telecommunications. Fourquantitative performance indicators are used to adapt the frame-work of four perspectives of BSC. However, in their studies do notconsider qualitative indicators in the model. According to Young-blood and Collins (2003) argued that although the BSC providesvaluable feedback on a variety of performance metrics, but thosemetrics did not consider the relative importance weigh and theissue of interaction and trade-offs between metrics. Unfortu-nately, those approaches might hold some drawback and pitfallsin the application. First of all, these technique did not considerhow to resolve the interdependent existing the quantitativeand qualitative indicators. Secondly, with an interactive BSC

C.-H. Wang et al. / Int. J. Production Economics 128 (2010) 413–426 415

framework, they could not determine simultaneously the degreeof interaction between the various perspectives and indicatorsthrough strict and objective methodology. Thirdly, under the BSCframework, it did not provide performance values to theircorresponding quantitative indicators, either for the individualperspectives or for their consolidations (Abran and Buglione,2003). The BSC framework also does not provide the quantitativeand qualitative indicators how much each perspectivecontributes, even on the relative importance weight for eachperspective and its corresponding indicators. This study adoptnon-additive fuzzy integral embedded in the HBSC frameworkthat provides the elements need eliminate or decrease theweaknesses of the traditional approaches and further relaxrestrictions for monitoring the deployment of firm strategy. Theremainder of this paper is organized as follows. Section 2 reviewsthe literature on BSC performance. Next, Section 3 outlines theHBSC structure with the fuzzy approach for measuring firmperformance. Section 4 then describes the method and algorithmfor evaluating high tech firm performance, after which theperformance-grade and importance weighting for each criterionwere determined. Subsequently, Section 5 utilizes the fuzzymeasures and non-additive fuzzy integral to measure firmsynthetic performance. Section 6 then assesses high tech firmperformance via the proposed model. Finally, the model resultsand conclusions are discussed.

2. Balanced scorecard and performance measurement system

In early studies, the measurement of firm performance focusesonly on organization financial performance. Watson (1975)proposed a threefold classification of performance measures,namely internal measures, interdependency measures and envir-onmental measures. Similarly, Cameron and Whetten (1983)stated that organizational performance can be assessed in termsof process, response or impact. Moreover, Chakravarthy (1986)studied computer firm operations and found that financialperformance measurement is an inadequate indicator of a broaderconstruct. Chakravarthy argued that future-oriented indicatorssuch as investment in R&D, product R&D, process R&D andproduct quality relative to competitors should also be incorpo-rated into the performance measurement indicators. Moreover,Fisher (1992) studied several high-tech manufacturing plants anddemonstrated the non-financial performance measure indicatesan attempt to reassert the primacy of operations over financialmeasures. Fisher found that by using non-financial measures andcontrol systems, managers attempt to track progress on theactionable steps that help firms succeed in the market. Similarly,Amir and Lev (1996) examined the value-relevance problem oftechnology-based industries, and pointed out that non-financialindicators such as growth proxy and market penetration ameasure of operating performance significantly influence value.Amir and Lev also highlighted the complementarities betweenfinancial and non-financial data are highlighted. More recentlynumerous studies have adopted wide aspects for assessing overallfirm performance. Atkinson et al. (1997) proposed that perfor-mance assessment should be considered by two different sources;internal and external, two different sources of information toassess organizational performance as stakeholders and classifythem into two groups the environmental stakeholders (custo-mers, owners and the community) and the process stakeholders(employees and suppliers). Similarly, Ecceles (1991) posits high-tech manufacturer recently assumed direct responsibility forincluding customer satisfaction, quality, market share, informa-tion technology and human resources in their formal performancemeasurement system. However, a broader approach is necessary

for measuring firm performance. For example, a broaderconceptualization of firm performance measurement wouldemphasize indicators of operational performance that is,non-financial information as well as indicators of financialperformance measures. The BSC measure framework fulfills thisbroader view.

Original balanced score was developed for performance measure-ment by Kaplan and Norton (1992), and employs performancemetrics from financial, customer and internal business processes, andlearning and growth perspectives; BSC provides a bridge linking thefinancial and non-financial perspectives into an integrated perfor-mance measurement system that aligns organization goals and othertraditional functional areas with corporate strategy using bothleading indicators (performance driver-oriented indicators) andlagging indicators (outcome-based measures indicators) to monitorstrategy implementation. The original BSC is an integrated perfor-mance measurement framework that helps firms articulate,communicate and translate strategy into action. The BSC frameworkclassifies performance measurement into four perspectives (Kaplanand Norton, 1992): financial, customers, internal business processes,learning and growth performance. The financial perspective involvesthe question, ‘To succeed financially, how we should appear to ourshareholders?’ The customer perspective concerns the question, ‘Toachieve our vision, how should we appear to our customers?’ Theinternal business processes perspective involves the question, ‘Tosatisfy our shareholders and customers, at what business processesmust we excel?’ Finally, the learning and growth perspectiveconcerns the question, ‘To achieve our vision, how should we sustainour ability to change and improve?’ BSC provides key performancemeasures that go beyond financial matters and consider customers,employees and internal operational processes. By simultaneouslyaggregating and coordinating information from several variousperspectives, managers can acquire the data required to increasedecision making efficiency and decision quality. Performancemeasurement thus is an interactive and continuous process.

Kaplan and Norton (1996) developed original BSC turnedbusiness strategies into measurable indicators. Kaplan and Nortonargued that the BSC program involves a cause-and-effect relation-ship among the different measurement in the selected perspec-tive, as illustrated in Fig. 1. Similarly, numerous different scholarshave provided empirical evidence supporting the existence of acause-and-effect relationship among BSC various perspectives(Olve et al., 1997; de Haas and Kleingeld, 1999; Nørreklit, 2003;Niven, 2002; Sandstrom and Toivanen, 2002; Quezad et al., 2009;Schmidberger et al., 2009). The relationship reflects the interplayand interdependencies among financial and non-financialmeasures. While specific high tech firms employed the learningand growth perspective to develop new processes andtechnologies to reduce costs and increase efficiencies in theinternal business processes perspective. Eventually, the result ofutilizing the learning and growth perspective provides increasedcustomer value and satisfies customer requirements, and finallysignificantly enhances financial perspective performance.Consequently, a well-constructed BSC method must consider theinteractive relationship among various selected perspectives andtheir measurement criteria.

The hierarchical balanced scorecard is a tool for translatingstrategy into action through the performance measures thatprovides the point of reference and focus for a high tech firm formdiagnosing a series of performance criteria to strategic imple-mentation. In line with Niven (2002) argument that the balancedscorecard deploys the organizational strategy, breaking it downinto its component parts through objectives, measures, targetsand initiatives in each of the four balanced scorecard perspectives.The use of the indicators relates to the balanced scorecardapproach, widely used in business strategy (Kaplan and Norton,

FinancialPerspective

Internal Business Processes Persp.

CustomerPerspective

Goal

Learning and Growth Persp.

Can we continue to improve and create value?

What must we excel at?

How do we look to shareholders?

How do customers see us?

Measure

Goal Measure

Goal Measure

Goal Measure

Fig. 1. Cause-and-effect for the balanced scorecard adopted from Kaplan and Norton (1992).


2001). The balanced scorecard approach advocates the develop-ment of supplementary indicators to measure the performance ofstrategies, with non-financial and even intangible indicators. Inthis study, performance indicators were proposed that can help toassess the performance of the implemented strategy for high techfirm. Adopting the proposed hierarchical balanced scorecard as aperformance measurement framework for translating the strategyinto different set of activities, high tech firm not only diagnoseperformance and strategy direction but also serves to guide allemployees’ activities toward the achievement of the acceptableperformance level and strategic implementation roadmap. Un-fortunately, BSC lacks a method of effectively measuring thoseimpacts. In the decision-making domain the non-additive fuzzyintegral method is used to assess interaction between alternativesand their effects on objectives. Applying the non-additive fuzzyintegral to the HBSC enables decision-makers to examine thecomparison of performance measure with established evaluationstandards and better determine the interaction between variousperformance measures perspectives and criteria. The process ofselecting the criteria metrics and HBSC structure is furtherexplained in Section 3.

3. HBSC with fuzzy approach for measuring firm performance

The balanced scorecard is designed to help the seniormanagers/managers to identify the issues of the business thatthey must be addressed in order to successfully achieve theorganizational strategy (Kaplan and Norton, 2001). It is also ahighly formalized performance measurement system to integrateorganizational strategic indicators and further turns actionsrelated to strategies into tangible and intangible indicators. Ourproposed measurement system extends and modifies from Kaplanand Norton (2001) in which dealt with indicators is thehierarchical balanced scorecard that is considered to be anappropriate tool for communicating bridge in a simple andparallel high tech firm strategy. The HBSC is explained as a seriesof measurements that give senior managers a comprehensive andstrategic view of the organization including the effects ofoperational and performance measures rely on four perspectives.The HBSC turns actions related to strategies into tangible andintangible indicators. Through comprehensive HBSC framework,strategy implementation (or vision) is easily configurable by wayof analysis a series of performance indicators and the contributionof tangible and intangible indicators can be made more explicit

and thus more controllable. Performance metrics are designed toconduct measurements throughout the four perspectives of BSCto determine whether the high tech firm requirements reach anexpectable performance level. The performance metrics used inthis study reduce subjectivity in measuring the performance ofhigh tech firms by providing a quantitative basis in a complexdecision-making process in which the overall performance of theavailable high technology firms needs to be measured in terms ofmultiple selection criteria. Complex decision-making problemscan divide objectives into hierarchical structures for furtherevaluation. Based on the four different perspectives of BSCpresented by Kaplan and Norton (1996); this study has madesignificant modifications and extended traditional BSC into aHBSC structure to suit the performance evaluation of high techfirms as shown in Fig. 2. Furthermore, the primary benefit of HBSClies in being able to provide a mechanism for managingperformance measurement system design process complexity.The HBSC provides a framework which is generally used to theassess performance of high tech firms and to help firms turnbusiness strategies into measurable indicators. The four dimen-sions of criteria for evaluating and selecting high tech firms arederived via literature review and in-depth interviews withscholars focused on high tech management and technologyinnovation, and with high tech industry experts who are high-level managers with an average of over 7 years of experience inthe high technology industry. Based on the HBSC structure, fourselection dimensions were identified, including the financialperspective (FP), customer perspective (CP), internal businessprocesses perspective (IB) and learning and growth perspective(LG). Four dimensions are used throughout this study. Eachdimension consists of a set of performance indicators/criteria thatquantitatively express the effectiveness or efficiency or both, of apart of or a whole process, or systems, against a given norm ortarget (Fortuin, 1988). Performance measurement thus involvesmeasuring activities using a set of performance indicators/criteriawhere performance indicators/criteria are also known as perfor-mance metric. These selection dimensions and their performancecriteria are discussed individually below.

Regarding the financial perspective (FP), financial performancedirectly reflects firm structuring profit and efficiency thus mostfirms use a financial index such as ROA or ROI to represent theirperformance (Chen, 1996). Six key financial measures wereconsidered important indicators of performance measurementfor high tech firms. This study used traditional finance perfor-mance indicators developed by Kaplan and Norton (1996).

Objective

Market shareCustomer retention andloyaltyCustomer satisfactionNew customer acquisition Customer profitability

Financial perspective (FP)

Customer perspective (CP)

Internal business processes perspective (IB)

Learning and growth perspective (LG)

Aspect

Firm

Per

form

ance

Eva

luat

ion

Criteria

Revenue growthReturn on capital employedEconomic value added Cash flowReduce cost rate

Return on investment

Regulation and management capabilityIntegrate R&D terms capabilityAlignment with customer/supplier Respond time to customer requests

Process continuous improvement capability

Process innovation capabilityConforming rate Improvement manufacture cycle capabilityTotal cost reduce/control capability

Manufacturing processes

Internalorganization processes

Information management capabilityInformation acquisition capabilityInformation maintain capabilityIT infrastructure

Motivation mechanismEmployee alignment with organizationDegree of empowerment

Information systemcapability

Employeescapability

Motivation,empowermentalignment

Employee satisfactionEmployee retentionEmployee productivities

Sub-criteria

Fig. 2. Performance measurement system of HBSC.


Financial performance indicators used for these six measures(shown in Fig. 2) determine specific high tech firm financialperformance. Senior managers and senior auditor managersassign various weights to the measures for each high tech firm.The performance of each high tech firm in terms of these sixindicators of financial performance is combined into a compositefinancial measure to determine the overall financial performancelevel of high tech firms. From the customer perspective (CP),numerous firms now focus on continuously improving productsor services to meet changing customer needs, particularly in theextremely competitive high tech industry. To determine theeffectiveness of their efforts, firms must determine objectives fortargeted customers that are linked to the overall strategy.Therefore, customer perspective measurements are laggingindicators that indicate to management how well customerobjectives are being achieved. The five core measurementsinclude market share, customer retention and loyalty, customersatisfaction, new customer acquisition and customer profitability.In the internal business processes perspective (IB), the traditionalperformance measurement system provides no insight into theproblems of internal business processes, resulting in firms beingunable to determine what is causing and driving their perfor-mance. The BSC identifies the interrelated nature of businessfunctional areas and processes. By recognizing the performancemeasurement of internal business processes, firms can under-stand the effects of their driver indicators in fulfilling targetcustomer objectives and ultimately firm financial objectives.Thus, firms must develop performance measures to track andmonitor the pivotal internal process and activities for supportingand achieving customer and firm objectives. Numerous empiricalstudies have argued that internal activities influence firmperformance, ranging from manufacturing processes (Bozarth,1993; Zirger and Hartley, 1996; de Ron, 1998; Ward and Duray,

2000; Evans, 2004) to internal organization management activ-ities (Zirger and Hartley, 1996; Sattler and Sohoni, 1999; Evans,2004; Schmidberger et al., 2009). The above activities are involvedon the part of firm performance measurement. Particularly, thehigh tech industries face a reduced product life cycle, and adynamic and complex competitive environment resulting in hightech firms needing to continuously improve and monitor internalorganization activities to achieve long-term survival and compe-titiveness. Furthermore, the evaluation of internal businessprocesses depended not only on the internal organizationprocesses but also on firm relative manufacturing process. Manyprocesses cut across organization function, or divisions, andrequire collaboration between individuals in different depart-ment. For example, the product development processes needsrequire employees from R&D, manufacturing and marketing tocollaborate to ensure the design, development and marketing ofnew products which can be manufactured cheaply and efficiently.Consequently, to measure the processes performance and refer toexpert opinions and previous studies, the factor of internalbusiness processes of high tech firm are broadly classified intotwo groups as follows: (1) Manufacturing processes, includingprocess continuous improvement capability, process innovationcapability, conforming rate, improvement in manufacturing cyclecapability and total cost reduction/control capability. (2) Internalorganization processes, including regulation and managementcapability, integrated R&D capability, alignment with customers/suppliers expectations, response time to customer requests. In thelearning and growth perspective (LG), Kaplan and Norton arguedthat learning and growth are the most difficult to measure. Indevising the high tech firm learning and growth perspective, theteam considered the following three measurements: informationsystem capability, employee capability and motivation andempowerment alignment. Each of these measurements further


contains several sub-criteria. Information system capabilityindicates information management capability, informationacquisition capability, information maintenance capability andinformation technology (IT) infrastructure. If interdependenciesare present within the same level of analysis in a dotted-linerectangle, looped arcs were used to represent such interdepen-dencies (see Fig. 2). The HBSC exist the interactive effect (l-value)within the same level criteria in which cannot captureinterdependencies through conventional fuzzy AHP method.Fig. 2 shows the various interdependencies among the sameaspects, criteria and sub-criteria (by means of the arcs andarrows). Therefore, the aspects and criteria are subjective due towhich a synthetic score through simple weight method such asfuzzy AHP is difficult to determine the interactive effect (l-value).These sub-criteria are presented in Fig. 2.

The role of performance measurement under the HBSC frame-work is to evaluate and monitor processes and outcomes ofindividual high tech firm. The original of HBSC is designedthrough a shared understanding and translation of the organiza-tional strategy into goals and measurable performance indicatorsin each of the four perspectives. In addition, performancemeasurement has been recognized that can be used to influencebehavior and decision-making, thus, affect the strategy (Neelyet al., 1994; de Lima et al., 2009; Barclay and Osei-Bryson, 2010).Further, Kaplan and Norton (1996) proposed that strategy may berevised and adjusted accordingly before the next cycle ofperformance measurement take place, in a process of incrementalstrategy adjustment by through feedback loop mechanism withinBSC. In this study, we argued that such strategic controls could beinduced using feedback loop and performance measurement forthe same purpose. In other words, strategies are realized throughthe HBSC performance measurement system. The HBSC frame-work provides one means of inducing strategies realization via aseries of performance indicators. More specifically, in this studywe set out to explore the extent to which senior managers of hightech firm seek to influence the realization of their strategiesthrough HBSC performance measurement system, since theperformance measures are most important as the primarymechanism to control and adjust strategy implementation.According to expert consensus, previous studies, including Kaplanand Norton (1996), encouraged the inclusion of 4–7 measures ineach evaluation category. Furthermore, owing to the limitationsof human short-term memory, this study argues that there are nomore than seven indicators of each aspect. Meanwhile, Keeneyand Raiffa (1976) propose that five properties must be followedduring criteria formulation, namely: completeness (the criteriamust cover all important aspects of the decision-making pro-blems), operational (the criteria must be meaningful for decision-making analysis), decomposable (the criteria can be decomposedfrom a higher to a lower hierarchy to simplify the evaluation),nonredundant (there must be no double counting of criteria) andminimum size (the number of criteria should be minimized).Based on these principles, this study developed the HBSCperformance measurement system, which can provide a measure-ment mechanism and appropriate measurement criteria andeliminate conflicts in the performance measurement system.

4. Method and algorithm for evaluating high tech firmperformance

To apply the performance measurement system of HBSC to realhigh tech firms, it is necessary to further identify the measure-ment indicators used for performance grading and importanceweighting. These high tech firms comprised a sample randomlyselected from among the Taiwanese high-tech industry

specifically from among those firms with one thousand or moreregular employees. Individual high tech firms are taken here asthe unit of analysis since a single firm can provide a set ofperformance measurement indicators developed using the check-list of performance measures based on the performance measure-ment system used by HBSC (see Fig. 2). Researchers distributedthe performance measures checklist via e-mail to senior managersand senior auditors of eight high tech firms to measure theiroverall performance, since the BSC is mainly designed to providesenior managers with an overview for evaluating organizationalperformance (Ghalayini et al., 1997). Managers in charge of firmstrategy have a good understanding of the internal/externaloperating environment of the business.

As previously mentioned, the entire high tech firm perfor-mance evacuation process is complex and imprecise, and relies onsubjective evaluator judgments. Managers making a performanceevaluation must consider various aspects and criteria, which mayinclude interactive, interdependent, and nonnumerical values.The basic assumptions were significant different between fuzzyAHP and fuzzy measure. The fuzzy AHP methods are directlyextension traditional Saaty AHP techniques with fuzzy numbersto the alternative selection and justification problem by using theconcepts of fuzzy set theory and hierarchical structure analysis.The main objective of fuzzy AHP is used to accurately capture theimportance weights of decision-maker judgment under vaguenessand uncertainty of human cognitive processes (Ayag, 2005).A simple additive weighting method was addressed in fuzzy AHPmethod to aggregating fuzzy arithmetic (Ayag, 2005) and, further,in real MCDM that the criteria are not necessarily mutuallyindependent when we employ the simple additive aggregatemethod (Chiou and Tzeng, 2002). More importantly, our proposedHBSC framework has an interaction effect among evaluationaspect and criteria based on the initial BSC assumptionsintroduced by Kaplan and Norton (1992). These assumptionsshould be carefully made so as to only obtain reliable data thatdirectly affect the attributes of the hierarchical framework. Inaddition, the non-additive fuzzy integral assumes monotonicityand boundaries property are more general than the conventionaladditive measures. This is why we used l-fuzzy measure and theChoquet integral to replace the fuzzy AHP approach.

Traditionally, the probabilities need to assign certain prob-abilities value to the elements. Unfortunately, the probabilitiesmethod is severely limited as it cannot capture any inherentrelation through expert’s knowledge and professional backgroundin term of performance scores. This limitation can be overcome byusing the fuzzy numbers and non-additive fuzzy integral withrespect to a fuzzy measure to aggregate multiple informationinput without loss any useful information. In fact, there are yetfew tools available for aggregation that is well rooted in thetheory. The most common aggregation tool which is used todaystill the weighted arithmetic mean, with all its well-knowndrawbacks (Grabisch, 1996). However, the weighted and averagemethods are based on the assumption that the attributes involvedare noninteractive and, hence, their weighted effects are viewedas additive (Wang et al., 1998). Wang et al. (1998) also furtherpointed out that the interactions among the attributes can beincorporated by expressing the importance of the various sets ofattributes in terms of an appropriate non-additive measure.Therefore, the fuzzy numbers and probability measures aredistinct aggregation tools and, thus, to handle the interactionsamong the criteria we employ the non-additive set functions as atool instead of traditional weighted and average methods in thesynthetic evaluation process. However, traditional decision mak-ing methods such as the multi-criteria decision making (MCDM),and the analytic hierarchy process (AHP) cannot adequatelyhandle the ambiguity and multiplicity of criteria involved the

Table 1Performance grade membership function.

Assessment Membership function

Performance-grade

Very poor (VP) (0.00, 0.30, 0.60)

Poor (P) (0.50, 0.60, 0.70)

Fair (F) (0.60, 0.70, 0.80)

Good (G) (0.70, 0.80, 0.90)

Very good (VG) (0.90, 1.00, 1.00)

~ P (p

)µ

1.0

1.0

0.2 0.4 0.60 0.8

VP G VG

p

P F

Fig. 3. Membership function for five performance-grades.


overall performance assessment. Rephrase Zadeh (1975) intro-duced the fuzzy set theory to enable uncertain and imprecise realworld systems to be captured via linguistics variables. Fuzzy logicthus is a useful tool for dealing with decisions involving complex,ambiguous, and vague phenomena based on the meanings of thelinguistic variables. Particularly, human decision making analysisinvolves obscure, uncertain and imprecise events and factors thathave difficulty in assigning a crisp value to subjective judgments(Chen and Hwang, 1992). Linguistics expression provides a usefulapproach for interpreting the semantics of vague based on thesubjective judgments of evaluators. High tech firm performancemeasurement and evaluation decisions are made mainly on thebasis of the opinions of senior auditor managers/senior managers.Linguistic expression can handle ambiguities in assessing dataand the vagueness of linguistic expression, in which evaluatorscan easily describe their measurement and subjective judgmentsusing the evaluation criteria within the framework of fuzzy settheory.

However, the use of performance metrics does not avoid oreliminate the need for human judgment in performance measure-ment. Subjective evaluator judgments are frequently involvedwith regard to the importance weighting of criteria, resulting infuzzy and imprecise data being used, and requiring the applica-tion of a fuzzy approach to effectively tackle such decisionproblems. Senior managers and senior auditor managers conductsubjective judgment and evaluate overall performance usingHBSC (see Fig. 2), and the evaluators assess performance-gradeand rate importance for each criterion and sub-criterion. Thesecriteria variables of each aspect of performance evaluation,performance-grade and their importance rating can be assessedusing fuzzy numbers, which can express linguistic variables.Restated, the triangular fuzzy numbers are used to expresslinguistic variables and appraise the performance-grade andimportance rating of the evaluation criteria.

4.1. Determination of performance-grade

To determine the performance-grade of each criterion, thisstudy invited three scholars and four experts from the high techindustry to assign their own performance-grades. The scholarscame from the department of technology management of theuniversity, and the high tech industry experts averaged over7 years experience in the high tech industry. All participantsreceived a set of performance measurement criteria based on theHBSC structure (see Fig. 2), and then described how theperformance measure was calculated and how to divide intodifferent performance-grades.

To objectively grade performance according to consensusamong scholars and high tech industry experts and useful dataprovided by real cases for high tech firm performance measure-ment. According to expert consensus, the five rating levels foreach performance-grade at the lowest level of the HBSC structureare appropriate. The intervals of performance-grade are organizedinto five rating levels depending on expert consensus. The fivenumerical intervals, [0%, 60%], [50%, 70%], [60%, 80%], [70%, 90%]and [90%, 100%] are used to measure high tech firm performanceachievement percentage, and are reflected in the triangular fuzzynumbers VP to VG level, respectively, where VP indicates theworst and VG indicates the best performance-grade of eachperformance evaluation criterion listed in Table 1. The definitionsof the triangular fuzzy numbers of levels VP to VG are intrinsicallysimilar to the intervals. For example, consider triangular fuzzynumber ~Ak, k¼VP,. . .,VG for a particular criterion A. Let~Ak ¼ ðak, bk, ckÞ, k¼VP, P, F, G where a, b and c represent threepoints at the left, medium and right of a particular triangular

fuzzy number and the corresponding interval is [ak, ck]. The mostprobably value bk then can be obtained as (ak+ck)/2. Fig. 3 shows amembership function diagram for the performance-grade isexemplified. Consequently, the fuzzy numbers for all criteria canbe defined and used to obtain the performance-grade of eachevaluation criteria at the lowest level of the HBSC structure formeasuring firm performance.

To achieve a consistent measurement of performance-grade,five performance-grades, VP to VG, are used to measure thecriteria depending on the evaluation criteria of the HBSCstructure. The performance-grade ~p can be estimated using thefollowing rating {very poor (VP), poor (P), fair (F), food (G) andvery good (VG)} and its associate membership function isillustrated in Fig.3, and was used to measure the rating effect ofthe different evaluation criteria used to assess firm performance.For each high tech firm requesting that senior managers andsenior auditors measure the performance achievement of eachcriterion using a triangular fuzzy number in the interval [0, 100%]according to the performance-grade. The 0 represents theminimum and 100 denote the maximum performance achieved.Therefore, the evaluators used the linguistic variables to directlyassess the impact of various criteria on the firm performanceevaluation. On the other hand, the importance of each evaluationcriterion should be determined based on the performanceevaluation criteria of the HBSC performance measurementsystem.

4.2. Determining the degree of importance

Numerous techniques exist, including AHP, TOPSIS and SAW,which can be applied to obtain importance weights for variousevaluation criteria. However, the use of these techniques forweighting performance has many drawbacks since these techni-ques do not express the uncertainty associated with presentingevaluator judgments in numerical form and evaluator subjectivejudgments and background knowledge significantly influencethose methods. Restated, importance weights depend on evalua-tor subjective judgment and cognition. Thus, importance weightsand the effect ratings of various criteria assigned by evaluators are


expressed in linguistic variables. Linguistic variables are proposedas a means of helping evaluators improve their decisions;moreover, linguistic variables an efficient approach for handlingevaluator subjective judgments and situational uncertainty.Additionally, appropriate fuzzy numbers provide an easy meansof presenting linguistic variables (Petrovic and Petrovic, 1996;Karsak and Kuzgunkaya, 2002). That is the importance weightingsof various criteria can be determined by direct assignment into afuzzy number by evaluators.

1.0

1.0

0.2 0.4 0.60 0.8

L H VH

w

~ W (w

)µ

VL M

Fig. 4. Linguistic model LMw1 for importance weighting.

1.0

1.0

0.2 0.4 0.60 0.8

L H VH

w

~ W (w

)µ

VL M


1.0

1.0

0.2 0.4 0.60 0.8

VL H VH

w

~ W (w

)µ

L M


Table 2Linguistic values of performance-grade and grade of importance.

Grade of importance (LMw1) Grade of importance (LMw

Assessment Membership function Assessment

Very low (VL) (0.00, 0.00, 0.25) Very low (VL)

Low (L) (0.20, 0.30, 0.40) Low (L)

Medium (M) (0.35, 0.50, 0.65) Medium (M)

High (H) (0.60, 0.70, 0.80) High (H)

Very high (VH) (0.50, 1.00, 1.00) Very high (VH)

The rating of importance ~w can be estimated using thefollowing ratings {very low (VL), low (L), medium (M), high (H)and very high (VH)} and its associate membership function.Figs. 4–6 were used to measure the importance of the variouscriteria. However, because of possible differences in experiencesand knowledge background among evaluators, those evaluatorsprovide the degree of various linguistic weightings for particularperformance evaluation criterion. To clarify these differences,different types of linguistic models can exhibit evaluatorinclinations regarding each performance evaluation.

The triangular membership functions are overlaps whichrepresent the dependence of the different linguistic models on theevaluator being more professional than other evaluators. Threelinguistic models, LMw1 to LMw3, represent the degree of impor-tance for particular criteria metrics (see Figs. 4–6 and Table 2). Forexample, LMw3 for explaining the grade of importance criteriametrics are the fuzziest among LMw1 and LMw1 and are used by theevaluators. The usefulness of the varied linguistic models dependson the ability of the evaluator to clearly distinguish the differencesof performance measurement criteria than other.

To determine the overall performance of particular high techfirms, multiple evaluation criteria are used, and are frequentlystructured into a multi-level HBSC system (Fig. 2) whichaccommodates fuzzy set theory to provide evaluator blueprintsduring the performance evaluation.

4.3. Algorithm and procedures of hi-tech firms performance

evaluation

During overall fuzzy evaluation, evaluator perceptions of criteriavary, and all have varying valuations of linguistic variables ofimportance weights. Given m evaluators for example, scholars andexpert senior managers, the criteria importance weight,~wi, i¼ 1, 2, . . ., m, and the evaluated values (or ratings) of perfor-

mance-grade metrics, ~pi, i¼ 1, 2, . . ., m, for a particular high techfirm, can be aggregated, based on the fuzzy arithmetic of Dubois andPrade (1980) to three vertices of triangular fuzzy number andcalculated aggregation determined via m evaluators using

~W ij ¼ ðLwij, Mw, RwijÞ ¼Xm

p ¼ 1

wpij

!=m,

Xm

p ¼ 1

wpij

!=m,

Xm

p ¼ 1

wpij

!=m

!,

ð1Þ

~Pij ¼ ðLpij, Mpij, RpijÞ ¼Xmp ¼ 1

ppij

!=m,

Xmp ¼ 1

ppij

!=m,

Xm

p ¼ 1

ppij

!=m

!,

ð2Þ

where ~W ij ¼ ðLwij, Mwij, RwijÞ and ~Pij ¼ ðLpij, Mpij, RpijÞ are triangular

fuzzy numbers, and their points on the left, middle and rightpositions, Lpij, Mpij and Rpij, represent the overall average ratings of

aspect i, criteria j over m evaluators, while both ~Wh

ij and ~Ph

ij ,

h¼1,2,y,m are fuzzy numbers for each evaluator.

2) Grade of importance (LMw3)

Membership function Assessment Membership function

(0.00, 0.00, 0.30) Very low (VL) (0.00, 0.00, 0.35)

(0.15, 0.30, 0.45) Low (L) (0.15, 0.30, 0.45)

(0.30, 0.50, 0.70) Medium (M) (0.25, 0.50, 0.75)

(0.55, 0.70, 0.85) High (H) (0.55, 0.70, 0.85)

(0.70, 1.00, 1.00) Very high (VH) (0.65, 1.00, 1.00)


Meanwhile, the fuzzy aggregation evaluation by each high techfirm yields is a fuzzy number. Therefore, these three fuzzy modelsare transformed into crisp numbers. Many defuzzification meth-ods have been developed, including center of sum, center ofgravity, mean of maxima, and the a-cut method, for achieving thispurpose. The defuzzying method established by Chen and Klein(1997) is a highly sensitive and effective approach for discrimina-tion during fuzzy ranking by performing many simulatedexperiments involving the application of various linear ornonlinear fuzzy numbers and various degrees of overlap of fuzzynumbers. Chen and Klein employed a method utilizing fuzzysubtraction of a referential rectangle, ~R from a fuzzy number, ~X ;the rectangle is derived by multiplying the height of themembership function of ~X by the distance between the two crispmaximizing and minimizing barriers. Here, ~R can be considered afuzzy number. Fuzzy subtraction of the referential rectangle, ~Rfrom the fuzzy number, ~X can be performed at level ai as follows:

~Xaio�4 ~R ¼ ½li, ri�½��½c, d� ¼ ½li�d, ri�c�, i¼ 0, 1, 2,. . .,1, ð3Þ

where o�4 and [�] represent fuzzy subtraction and intervalsubtraction operators, respectively; li and ri denote the left andright loci of ~R, and c and d are the left and right barriers,respectively. The defuzzification rating of the fuzzy number isthen

Dð ~X Þ ¼

Pni ¼ 0 ri�cPn

i ¼ 1ðri�cÞ�Pn

i ¼ 0ðli�dÞand n-1, ð4Þ

where n denotes the number of a-cuts; as n approaches N, thesum is the measured area.

In Eq. (4),Pn

i ¼ 1ðri�cÞ is positive,Pn

i ¼ 1ðli�dÞ is negative and0rDð ~X Þr1 if 0rxr1. Chen and Klein (1997) demonstrated thattheir defuzzification method are more efficient and accurate thanother methods such as weighted center, area center, totalintegral value, left and right assigned scores, Huang’s rankingand Choobineh and Li’s ranking methods. As shown in theexperimental results on Chen and Klein (1997), theirdefuzzification method is a well-defined approach and herecan be used to implement the defuzzification processes inthis study.

5. Fuzzy measures and non-additive fuzzy integral formeasuring firm synthetic performance

To solve the assumption of hierarchical BSC system that not allcriteria are completely independent, this study employed fuzzyintegrals that were monotonic and non-additive to determine theimportance weights for performance evaluation criteria. Theoriginal fuzzy integrals were introduced by Sugeno (1974). Thisstudy also used ‘importance weight’ to model evaluator subjectivejudgments and their preference structure. Consequently, a fuzzymeasure used in this study can be explained as the subjectiveimportance of evaluator criterion during the evaluation process.Sugeno and Terano (1977) incorporated the l-additive axiom tosimplify information accumulation. In the fuzzy measure space(X, b, g), let lA(�1, N). If AAb, BAb A\B¼f, and

glðA [ BÞ ¼ glðAÞþglðBÞþlglðAÞglðBÞ ð5Þ

hold, then the fuzzy measure g is l-additive. This particular fuzzymeasure is termed l-fuzzy measure because it must satisfyl-additive, and is also known as the Sugeno measure (Sugeno,1974). To differentiate this measure from other fuzzy measures,l-fuzzy measure is denoted by gl. When l¼0, the measure isadditive. To date, the l-fuzzy measure has been widely used as afuzzy measure. Additionally, the l-fuzzy measure of the finiteset can be derived from fuzzy densities, as indicated in the

following equation. Based on Eq. (5), the fuzzy measureg(X)¼gl({x1, x2,y,xn}) can be formulated as follows (Keeney andRaiffa, 1976; Leszczynski et al., 1985):

glð x1,x2,. . .,xnf gÞ ¼Xn

i ¼ 1

giþlXn�1

i1 ¼ 1

Xn

i2 ¼ i1þ1

gi1 gi2þ � � � þln�1g1g2 � � � gn

¼1

l

Yn

i ¼ 1

ð1þlgiÞ�1

�� for �1rlo1: ð6Þ

Based on the boundary conditions in Eq. (6), gl(X)¼1, l can beuniquely determined via the following equation:

lþ1¼Yn

i ¼ 1

ð1þlgiÞ: ð7Þ

In the case of l-fuzzy measure identification, fuzzy density gi,i¼1, 2,y,k and parameter l must be determined. Since l-fuzzymeasures values, and AAbðXÞ for a set X¼{x1,x2,y,xk} aresubjectively determined, it is difficult to obtain consistentmeasures values that satisfy fuzzy measurement properties fromhuman experts.

In a fuzzy measure space (X, b, g), let h denote a measurablefunction from X to [0, 1]. The fuzzy integral of h over A withrespect to g is then defined asZ

AhðxÞdg ¼ sup

aA ½0,1�½a4gðA \ FaÞ�, ð8Þ

where Fa¼{x9h(x)Za}(Wang et al., 1998) and A represents thedomain of a fuzzy integral. When A¼X, the fuzzy integral can alsobe denoted as

Rh dg. For simplicity, consider a fuzzy measure g of

(X, P(X)) where X is a finite set. Let h:X-[0, 1] and assume,without loss of generality, that the function h(xi) is monotonicallydecreasing in i, such that hðx1ÞZhðx2ÞZ � � �ZhðxnÞ. Elements in Xthen can be renumbered to ensure that Eq. (9) has the followingequilibrium:Z

hðxÞ3g ¼ 3n

i ¼ 1½hðxiÞ4gðHiÞ�, ð9Þ

where Hi ¼ fx1,x2,. . .,xig, i¼ 1, 2,. . .,n:In this study, h(U) can be considered as the performance of a

particular criterion of a particular aspect and g(U) represents thesubjective importance weight of each criterion. The fuzzy integralof h(U) with respect to g(U) provide an overall assessment of thecriterion. For simplicity, the same fuzzy measure of the Choquetintegral, rather than the fuzzy integral in Eq. (9), is applied, asfollows:

ðcÞ

Zhdg ¼ hðxnÞgðHnÞþ½hðxn�1Þ�hðxnÞ�gðHn�1Þ

þ � � � þ½hðx1Þ�hðx2Þ�gðH1Þ

¼ hðxnÞ½gðHnÞ�gðHn�1Þ�þhðxn�1Þ½gðHn�1Þ�gðHn�2Þ�

þ � � � þhðxnÞgðH1Þ, ð10Þ

where H1 ¼ fx1g, Hfx1, x2g,. . .,Hn ¼ fx1,x2,. . .,xng ¼X. In the litera-ture, the fuzzy integral defined by (c)

Rhdg is termed a

non-additive fuzzy integral. The proposed model using thenon-additive fuzzy integral does not require the assumption ofthe mutual independence of criteria. The primary advantage ofnon-additive fuzzy numbers compared to other integrationmethods is that it is able to represent certain kind of interactionbetween criteria, ranging from redundancy (negative interaction)to synergy (positive interaction) (Grabisch, 1996). Based on theargument of Grabisch (1996), there is almost no well establishedmethod to deal with interacting criteria except non-additive fuzzymeasure, and usually people tend to avoid the problem byconstructing independent criteria. The non-additive fuzzy has also


the ability to represent the human preferences in a formal andsystematic way (Grabisch, 1995; Marichal and Roubens, 2000).The model can thus be used in nonlinear situations. Even criteriathat are objectively and mutually independent are not consideredindependent of subjective evaluators.

The role of Choquet integral in the area of synthetic evaluationand information fusion have been examined by Onisawa et al.(1986) and Tahani and Keller (1990). In addition, the Choquetintegral was used as an important aggregation tool (Tahani andKeller, 1990; Grabisch, 1995, 1996) which is used for aggregatinginformation from multiple information sources with respect to thefuzzy measure (Tahani and Keller, 1990). Information from multipleinputs or sources may agree or conflict with each other. Therefore,the task of information fusion is to search for a maximum degree ofagreement between the conflicting supports of an object. In thisstudy, the performance measurement system of HBSC exist inherentinteractive among all of strategic aspects and multiple informationsources, requiring appropriate tools to aggregate multipleinformation sources and to handle interactive relationship is needs.The Choquet integral has the ability to handle conflicts in evidenceand to make aggregations in the multiple information sourceswithout loss any information. This is why that the Choquet integralemployed as our tool to integrate various information sources inthis study.

6. Empirical analysis for Taiwan high tech firm performanceevaluation

During recent decades, the high tech industry has played acentral role in the economic development of Taiwan. The hightech industry has enjoyed average annual growth rate of 18.5%since 1991, and maintained 21.4% annual growth during 2004. In1991, the contribution of the high tech industry to GDP stood at11.76%, which increased to 14.07% by 2003 (MOEA, 2004).Restated, the contribution of high tech industry production toGDP has continuously increased and the high tech industryappears to be steadily growing. Moreover, the Taiwanese govern-ment also provides favorable conditions such as researchsubsidies, tax reductions and funding to promote high techindustry competitiveness. As a result, the Taiwanese high techindustry has been becoming increasingly competitive. However,many high tech firms still lack efficient and accurate evaluationmethods for measuring their operation performance. To overcomethis dilemma, this study designed a HBSC system and fuzzymeasure based on non-additive fuzzy integral methods tomeasure high tech firm operating performance. These methodsof fuzzy measurement and non-additive fuzzy integral appearmore appropriate for measuring the dependent criteria in a fuzzyenvironment. Firms can use the proposed method for self-assessment or independent assessment by third parties to identifyand audit problems in their practices. This study designed aperformance measurement checklist using the hierarchical BSCsystem. The date acquired for this study was obtained by16 participants and the performance checklist was sent to seniormanagers and internal auditors within eight high tech firms, allwith more than 6 years of work experience and all specialists inhigh tech management. The high tech firms are located in theHsinchu and Tainan science parks in Taiwan. Sixteen participatorswere asked to evaluate the importance weight and performance-grade of each evaluation criterion according to the practicaloperating situation of each firm. To increase validity of perfor-mance measurement, all high tech firms in our samples havechecked whether utilized BSC framework as their performancemeasurement systems before performance evaluation work carryout. Evaluators were requested to evaluate the importance

weighting of criteria depending on the HBSC system using thefive fuzzy linguistic weighting variables. Fuzzy sets theory isespecially appropriate for expressing subjectivity and objectively,qualitative and quantitative, and ambiguity and precision such ashuman experiences, overall judgment, human psychology orbehavior and human feeling are easy to explain through fuzzysets theory. Researchers are especially attracted to fuzzy setstheory because it allow researchers to capture domain knowledgequickly using that contain fuzzy domain terms without muchrestriction such as number of data (Terano et al., 1992).Unlike probability statistical theory, they argued that fuzzy setstheory can expresses the amount of ambiguity under the littledata, the total number of the data has no meaning, and theoccurrence of events is not clear. In this study, the fuzzy modelwas chose that widely used in real-case high tech firmsperformance measurement within BSC performance evaluationsystem.

According to the criteria/sub-criteria of the bottom level, theimportance weight scores can be obtained in the case of the HBSCsystem. This is, from the bottom level, every item comprises ananswer to a question and the associated importance weighting.Likewise, internal business processes perspectives (IB) have nineanswered questions and the associated importance weight scores.The criteria importance scores are graded using the membershipfunction of importance weight (see Figs. 4–6) which depends onthe individual subjective perspectives and professional knowl-edge background of every individual evaluator. Each evaluatorobtains their l-value using Eq. (7) with corresponding measuredensity, gi (fuzzy measure density is apparent that the degree ofimportance weight of each criterion). In this study, each evaluatordetermined the importance weighting of criteria using thetriangular fuzzy number and aggregative fuzzy importanceweight obtained in Eq. (1). The defuzzifying approach in Eq. (4)can obtain the crisp importance weight. Furthermore, substitutingthe crisp number of the importance weight into Eq. (7) can yieldthe fuzzy li value. In Fig. 7, the aspects FP, CP, IB and LG generate a

T value in level 2 for the evaluator using Eq. (7) based on g1, g2, g3

and g4. Moreover, level 2 contains l1, l2, l3 and l4 for theevaluator using Eq. (7) based on the importance weight of eachcriterion in level 3. Additionally, level 3 contains l1, l2, l31, l32,l41, l42 and l43 for the evaluator based on the importance weightof each criterion of level 4. Consequently, every evaluatorpossesses ten li values that need to be determined. Moreover,the Choquet integral (c)

Rh dg in Eq. (10) is utilized to determine

the aggregated value of each criterion based on the sub-criteriaand the aggregated value of each aspect based on the criteria(see Fig. 7).

Using Fig. 7 and the bottom to top method based on theproposed approach it is possible to easily and efficiently obtainoverall high tech firm performance in situations involvinginteraction among criteria.

Table 3 presents sub-criteria for the importance weight (fuzzymeasure gi) of firm B and aggregated values ((c)

Rh(xi) dgl) of

internal business processes perspective (IB) and learning andgrowth perspective (LG) based on the evaluation criteria for theHBSC system. Two evaluators exist for each high tech firm, withone being a senior manager and the other being a senior internalauditor. The evaluators were requested to complete a checklistusing subjective judgments of the importance weighting of eachsub-criterion. The subjective judgments of participants wereintegrated to employ the fuzzy importance weights the fuzzymeasure density gi is apparent for each criterion as in Eq. (1) andthen the nonfuzzy importance weights were obtained usingEq. (4), as shown in Table 3. In Table 3, the l-value wasdetermined using Eq. (7) based on the gi of each sub-criteriaand the program was designed using Mathematical 5.0. Table 3

Level 1Objective

Level 2Aspect

Level 3Criteria

Level 4Sub-criteria

Ove

rall

perf

orm

ance

sco

re

g11

g12

g13

g14

g15

g16

fp1

fp2

fp3

fp4

fp5

fp6

(include 1)FP

1

∫hdg

∫hdg

CP

cp1

cp2

cp3

cp4

cp5

g21

g22

g23

g24

g25

(include 2)

2

∫hdg

LG

g411

g412

g413

g414

lg1

lg12

lg13

lg14

lg11

include ( 41)∫hdg

41

(include 42)

424

∫hdg ∫hdglg2

lg21

lg22

lg23

g421

g422

g42342

43

lg31

lg32

lg33

lg3

g431

g432

g433

(include 43)

∫hdg

43

ib1

ib11

ib12

ib13

ib14

ib15

g311

g312

g313

g314

g315

(include 31)

31

ib2

ib21

ib22

ib23

ib24

g321

g322

g323

g324

(include32)

32

IB

3

∫hdg

∫hdg

∫hdg

31

32

Fig. 7. The process of evaluation final synthetic performance values.


shows that there is a high degree of interactive mutual influence,namely l-values, with the IB and LG exceeding �0.95. Thel-values close to �1 demonstrated the existence of completedependent and mutual influence relationships among sub-criteria, and demonstrated the importance of sub-criteria to theinteractive effect among the other sub-criteria. Consequently,according to Eq. (10), the Choquet integral (c)

Rh(�) dgl in Eq. (10)

can determine the aggregated value of each criteria depending onsub-criteria and the aggregate value of each aspect dependingon criteria (see Table 4). Table 3 presents the crisp performancescores h(�) and the importance weighting of gi(�), which wasassessed by the senior manager and senior internal auditor, the lvalue of gl(�), obtained by Eq. (7) and program was designedusing Mathematical 5.0 and the aggregated values (c)

Rh(�) dgl

obtained by Eq. (10). In Table 3, the aggregated value (c)R

h(�) dglrepresents the overall perceived performance of the evaluatorperceptions of the five criteria ‘manufacturing processes (ib1)’,‘internal organization processes (ib2)’, ‘information systemcapability (ig1)’, ‘employee capability (ib2)’ and ‘motivation andempowerment alignment (ib3)’.

From Table 3, the Choquet integral values (c)R

h(�) dgl of eachsub-criterion can further employed to determine the nextevidence value h(�) of each criterion (see Table 4). Table 4 alsoshows h(�) and gi(�) as assessed by the senior manager and seniorinternal auditor, as well as the l value of gl(�) and (c)

Rh(�) dgl for

the four aspects, financial perspective, customer perspective,internal business processes perspective and learning and growthperspective based on the HBSC structure for the performanceevaluation of firm B.

Similarly, from Table 4, the Choquet integral values (c)R

h(�)dgl can further obtain the overall performance value for firm B,which is shown in Table 5, where the overall performance forfirm B is 0.813 (81.3%). In Table 5, the l-value equals �0.98(which is extremely close to �1 and indicated completedependence) implying a high degree of interaction amongvarious aspects of HBSC. Furthermore, accuracy informationshould be obtained during the performance measurementprocess to examine the overall perspective of particular firms.According to the above approach the operational process isrepeated using a bottom-up approach until each type of evidence

Table 3The fuzzy measure and aggregated values of IB and LG for the firm B.

Criteria Sub-criteria h(�) gi(�) (c)R

h(�) dgl (l-value) gl(�)

ib1 ib11 0.682 0.591 0.765 (�0.991) gl(ib11)¼0.591

ib12 0.682 0.409 gl(ib11,ib15)¼0.838

ib13 0.728 0.695 gl(ib11,ib15,ib12)¼0.941

ib14 0.682 0.591 gl(ib11,ib15,ib12,ib13)¼0.991

ib15 0.728 0.591 gl(ib11,ib15,ib12,ib13,ib14)¼1.00

ib2 ib21 0.728 0.591 0.762 (�0.974) gl(ib24)¼0.682

ib22 0.728 0.591 gl(ib24,ib21)¼0.910,

ib23 0.728 0.682 gl(ib24,ib21,ib22)¼0.977

ib24 0.773 0.682 gl(ib24,ib21,ib22,ib23)¼1.000

lg1 lg11 0.773 0.591 0.818 (�0.980) gl(lg14)¼0.682

lg12 0.773 0.682 gl(lg14,lg11)¼0.878

lg13 0.773 0.591 gl(lg14,lg11,lg12)¼0.973

lg14 0.839 0.682 gl(lg14,lg11,lg12,lg13)¼1.000

lg2 lg21 0.773 0.591 0.805 (�0.924) gl(lg23)¼0.591

lg22 0.728 0.682 gl(lg23,lg21)¼0.859

lg23 0.839 0.591 gl(lg23,lg21,lg22)¼1.000

lg3 lg31 0.728 0.591 0.818 (�0.976) gl(lg33)¼0.591

lg32 0.794 0.889 gl(lg33,lg32)¼0.967

lg33 0.839 0.591 gl(lg33,lg32,lg31)¼1.000

Table 4Fuzzy measure and aggregated values of IB and LG for the firm B.

Aspect Criteria h(�) gi(�) (c)R

h(�)dgl (l-value) gl(�)

FP fp1 0.728 0.591 0.720 (�0.981) gl(fp2)¼0.409

fp2 0.728 0.409 gl(fp2,fp1)¼0.839

fp3 0.728 0.591 gl(fp2,fp1,fp3)¼0.906

fp4 0.682 0.409 gl(fp2,fp1,fp3,fp6)¼0.966

fp5 0.682 0.318 gl(fp2,fp1,fp3,fp6,fp4)¼0.980

fp6 0.728 0.591 gl(fp2,fp1,fp3,fp6,fp4.fp5)¼1.000

CP cp1 0.682 0.409 0.824 (�0.969) gl(cp2)¼0.786

cp2 0.839 0.500 gl(cp2,cp3)¼0.905

cp3 0.839 0.786 gl(cp2,cp3,cp1)¼0.955

cp4 0.682 0.318 gl(cp2,cp3,cp1,cp4)¼0.979

cp5 0.682 0.409 gl(cp2,cp3,cp1,cp4,cp5)¼1.000

IB ib1 0.765 0.591 0.752 (�0.811) gl(ib1)¼0.786

ib2 0.765 0.786 gl(ib1,ib2)¼1.000

LG lg1 0.818 0.682 0.816 (�0.944) gl(lg1)¼0.682

lg2 0.805 0.591 gl(lg1,lg3)¼0.925

lg3 0.818 0.682 gl(lg1,lg3,lg2)¼1.000

Table 5Fuzzy measure and overall performance values for the firm B.

Firm Aspect h(�) gi(�) (c)R

h(�) dgl (l-value) gl(�)

B FP 0.720 0.682 0.813 (�0.980) gl(CP)¼0.682

CP 0.824 0.682 gl(CP,LG)¼0.877

IB 0.752 0.591 gl(CP,LG,IB)¼0.960

LG 0.816 0.591 gl(CP,LG,IB,FP)¼1.000


is obtained. To determine the overall performance result, theChoquet integral is utilized again to integrate FP, CP, IB and LG(see Fig. 6).

The different aspect performance values and overall performancescores are determined by analyzing all sets of performanceevaluation criteria (see Fig. 7) for various high tech firms by theirsenior managers and senior internal auditors depended on the HBSCsystem and proposed evaluation method. Therefore, adopting the

similar evaluation procedure and proposed approach can obtainoverall performance scores for selecting hi-tech firms. The individualaspect scores and global performance evaluation scores of each hightech firm are provided in Table 6. To clearly understand theperformance scores for the different firms under the HBSC fourdifferent aspects, the performance scores for each of eight cases arealso listed in Table 6.

From Table 6, for the global perceived performance score(or the value of (c)

Rh(�) dgl) of eight hi-tech firms only one firm

scored over 90%. Through analyzing and evaluating all sets ofperformance criteria evaluated by senior manager and seniorauditor based on the HBSC performance measurement system(see Fig. 1) each firm can determine the performance-grade fromvarious perspective. The evaluation results can provide decision-makers with a deep understanding of organizational advantagesand disadvantages during strategy development. Although firm Dachieved the highest global performance score, its score was nothighest in all aspects. Table 6 shows the performance values foreach aspect and the corresponding priority of alternatives. Clearly,

Table 6Firm performance value and their corresponding rankings.

Firm FP CP IB LG GPS

PV Ranking PV Ranking PV Ranking PV Ranking Score (%) Ranking

A 0.758 5 0.912 1 0.681 6 0.738 6 0.877 (87.7%) 2

B 0.729 6 0.824 4 0.752 4 0.816 2 0.813 (86.2%) 5

C 0.723 7 0.728 8 0.785 2 0.712 7 0.775 (77.5%) 6

D 0.912 1 0.887 2 0.792 1 0.805 3 0.902 (90.2%) 1

E 0.859 3 0.836 3 0.681 6 0.823 1 0.846 (84.6%) 4

F 0.899 2 0.785 5 0.765 3 0.780 4 0.871 (87.1%) 3

G 0.716 8 0.741 7 0.709 5 0.760 5 0.764 (76.4%) 7

H 0.762 4 0.770 6 0.383 8 0.690 8 0.760 (76.0%) 8

GPS is abbreviated from ‘global performance score’. The abbreviated PV indicates the performance value of different perspective on HBSC.


firm D has the best performance score for the four aspectsrespectively, based on the HBSC performance measurementsystem. However, the perceived performance score of the internalbusiness processes perspective for firm D only reaches 0.792,meaning the performance may falls short of requirements.This result indicates that managers must track back the manu-facturing processes and relevant departments to examine areassuch as continuous process improvement capability, conformingrate, total cost reduction/control capability, etc., to improve IBperspective performance.

7. Conclusions

Balanced scorecard is a tool for translating strategy into actionvia various sets of performance measurement indicators. Numer-ous studies and publications have designed procedures forevaluating performance measurements/indicators. However, fewsuch studies use standard-settings of measures in performance-grade in which is still problematic for high tech firms. Thus, thisstudy identifies the performance-grade setting depending onexpert consensus opinions from experts working in high techindustry. Furthermore, this study also constructed the HBSCsystem capable of providing a reference point and focus for theentire organization. Particularly, the balanced scorecard providesa set of evaluation indicators for monitoring and tracking back thefactors that require improvement to ensure that managers anddecision-makers remain ‘in control’ and can respond quickly toitems that require immediate attention. This study adopted a non-additive fuzzy set function and algorithm procedure to solve thebalanced scorecard, difficult to quantify and cause-and-effectrelationship among various perspectives. An important advantageof the non-additive measurement approach is that the interactionof the aspects and criteria can be clearly identified and expressedquantitatively. This identification enabled researchers and man-agers to understand the interaction of aspects will influence theperformance evaluation results. The application of the non-additive measurement model to evaluate the performance ofvarious high tech firms demonstrates that the effects of multiaspects on performance can be aggregated into a global perceivedperformance score. Moreover, the proposed performance mea-surement methods can provide firm self-auditing and third-partyexpert evaluations based on HBSC. In this study, the performancemeasurement is only tend to the inside perspective such as seniormanagers within an organization but may ignore outside aspects.Nevertheless, the pitfalls of this philosophy can improveaccording to propose HBSC framework. Through this framework,the outside experts such as scholars or third parties can measurethe high tech firms’ operational performance to reduce the

gaps and cognizance between inside and outside performancemeasurement perspectives. In subsequent research, there is aneed to focus on the outside information source within which theorganization is set, rather than just being concerned with internalperformance measurement. Decision-makers look increasinglyoutside information to improve performance measurementaccurately and then plan to move the firm’s strategic adjustmentaccordingly.

The main benefits of the hierarchical BSC performance evalu-ation system presented in this study can establish a communica-tion system that bridges the gap between goals established byhigh-level managers and the employees whose performance isultimately responsible for achieving organizational goals. Bygathering and processing information from existing organizationsit is possible to understand the manner in which these organiza-tions operate based on the HBSC performance evaluation system;HBSC determines whether the organization is achieving its goalsand assigns a grade together with a quantitative assessment thatmanager and employees can use to assess whether the organiza-tion is making progress. Second, adopting the HBSC performanceevaluation system with organizational operating enables con-trollers and managers to more easily establish comprehensive andeffective integrating perspectives from different departments oforganizations. Consequently, successful organization performancemeasurement systems can effectively integrate the informationflow within various organization functions and between businessprocesses, and then can cooperate and manage enterprise opera-tion practices, including R&D, finances, manufacturing, marketing,service and human resource management. Complete performancemeasurement systems are implemented to identify and validateexisting organization performance evaluations and help anorganization to further increase its competitive advantage.

Acknowledgments

The authors would like to thank to acknowledge the contribu-tion of the referees and Editor-in-Chief, Dr. Cheng T.C.E. for theirinvaluable comments and suggestions which improved thequality of the paper.

References

Amir, E., Lev, B., 1996. Value-relevance of nonfinancial information: the wirelesscommunications industry. Journal of Accounting and Economics 22, 3–30.

Atkinson, A.A., Waterhouse, J.H., Wells, R.B., 1997. A stakeholder approachto strategic performance measurement. Sloan Management Review 38 (3),25–37.


Abran, A., Buglione, L., 2003. A multidimensional performance model forconsolidating balanced scorecards. Advances in Engineering Software 34,339–349.

Ayag, Z., 2005. A fuzzy AHP-based simulation approach to concept evaluation in aNPD emvironment. IIE Transactions 37, 827–842.

Banker, R.D., Chang, H., Janakiraman, S.N., Konstans, C., 2004. A balanced scorecardanalysis of performance metrics. European Journal of Operational Research154, 423–436.

Bozarth, C.C., 1993. A conceptual model of manufacturing focus. InternationalJournal of Operations and Production Management 13, 81–93.

Barclay, C., Osei-Bryson, K.M., 2010. Project performance development framework:an approach for developing performance criteria and measures for informationsystems (IS) projects. International Journal of Production Economics 124,272–292.

Chakravarthy, B.S., 1986. Measuring strategic performance. Strategic ManagementJournal 7, 437–458.

Cameron, K., Whetten, D., 1983. Organizational Effectiveness: A Comparison ofMultiple Models. Academic Press, London.

Chen, M.J., 1996. The applications of fuzzy multiple attribute decision making instock selection. Journal of Management Science 13, 227–248.

Chen, S.J., Hwang, C.L., 1992. Fuzzy Multiple Attribute Decision Making Methodsand Applications.. Springer-Verlag, Berlin, Germany.

Chen, B.C., Klein, C.M., 1997. An efficient approach to solving fuzzy MADMproblems. Fuzzy Sets and Systems 88, 51–67.

Chan, Y.C.L., 2006. An analytic hierarchy framework for evaluating balancedscorecards of healthcare organizations. Canadian Journal of AdministrativeSciences 23, 85–104.

Chiou, H.K., Tzeng, G.H., 2002. Fuzzy multicriteria decision making approach forindustrial green engineering. Environmental Management 30, 816–830.

Dubois, D., Prade, H., 1980. Fuzzy Sets and Systems: Theory and Applications.Academic Press, New York.

de Haas, M., Kleingeld, A., 1999. Multilevel design of performance measurementsystems: enhancing strategic dialogue throughout the organization. Manage-ment Accounting Research 10, 233–261.

de Ron, Ad J., 1998. Sustainable production: the ultimate result of a continuousimprovement. International Journal of Production Economics 56/57, 99–110.

de Lima, E.P., da Costa, S.E.G., de Faria, A.R., 2009. Taking operations strategy intopractice: developing a process for defining priorities and performancemeasures. International Journal of Production Economics 122, 403–418.

Ecceles, R.G., 1991. The performance measurement manifesto. Harvard BusinessReview 69, 131–137.

Evans, J.R., 2004. An exploratory study of performance measurement systems andrelationships with performance results. Journal of Operations Management 22,219–232.

Fisher, J., 1992. Use on nonfinancial performance measures. Journal of CostManagement 6, 31–38.

Fortuin, L., 1988. Performance indicators—why where and how? European Journalof Operational Research 34 1–9.

Ghalayini, A.M., Noble, J.S., Crowe, T.J., 1997. An integrated dynamic performancemeasurement system for improving manufacturing competitiveness. Interna-tional Journal of Production Economics 48, 207–225.

Grabisch, M., 1995. Fuzzy integral in multicriteria decision making. Fuzzy Sets andSystems 69, 279–298.

Grabisch, M., 1996. The application of fuzzy integrals in multicriteria decisionmaking. European Journal of Operational Research 89, 445–456.

Kaplan, R.S., Norton, D.P., 1992. The balanced scorecard-measures that driveperformance. Harvard Business Review 70, 71–79.

Kaplan, R.S., Norton, D.P., 1996. The Balanced Scorecard: Translating Strategy intoAction. Harvard Business School Press, Boston.

Kaplan, R.S., Norton, D.P., 2001. The Strategy Focused Organization How BalancedScorecard Companies Thrive in The New Business Environment. HarvardBusiness School Press, Boston.

Karsak, E.E., Kuzgunkaya, O., 2002. A fuzzy multiple objective programmingapproach for the selection of a flexible manufacturing system. InternationalJournal of Production Economics 79, 101–111.

Keeney, R.L., Raiffa, H., 1976. Decision with Multiple Objectives: Preferences andValue Tradeoffs. John Wiley and Sons, New York.

Leszczynski, K.Penczek, Grochulski, P., Sugeno, W., 1985. Fuzzy measure and fuzzyclustering. Fuzzy Sets and Systems 15, 147–158.

Lee, A.H.I., Chen, W.C., Chang, C.J., 2008. A fuzzy AHP and BSC approach forevaluating performance of IT department in the manufacturing industry inTaiwan. Expert Systems with Applications 34, 96–107.

MOEA, 2004, Annual Report of Industrial Production. Department of Statistics,Ministry of Economic Affairs, Taiwan.

Marichal, J., Roubens, M., 2000. Determination of weights of interacting criteriafrom a reference set. European Journal of Operational Research 124, 641–650.

Niven, P.R., 2002. Balanced Scorecard Step By Step: Maximizing Performance andMaintaining Results. John Wiley and Sons, New York.

Nørreklit, H., 2003. The balanced scorecard: what is the score? A rhetoricalanalysis of the balanced scorecard. Accounting Organizations and Society 28,591–619.

Neely, A., Mills, J., Platts, K., Gregory, M., Richard, H., 1994. Realizing strategythrough measurement. International Journal of Operational and ProductionManagement 14, 104–152.

Olve, N.G., Roy, J., Wetter, M., 1997. Performance Drivers: A Practical Guide toUsing the Balanced Scorecard. John Wiley and Sons, New York.

Onisawa, T., Sugeno, M., Nishiwaki, Y., Kawai, H., Harima, Y., 1986. Fuzzy measureanalysis of public attribute towards the use of nuclear energy. Fuzzy Sets andSystems 20, 259–289.

Petrovic, R., Petrovic, D., 1996. Multicriteria ranking of inventory replenishmentpolicies in the presence of uncertainty in customer demand. InternationalJournal of Production Economics 45, 439–446.

Quezad, C., Cordova, F.M., Palominos, P., Godoy, K., Ross, J., 2009. Method foridentifying strategic objectives in strategy maps. International Journal ofProduction Economics 122, 492–500.

Ravi, V., Shankar, R., Tiwari, M.K., 2005. Analyzing alternatives in reverse logisticsfor end-of-life computers: ANP and balanced scorecard approach. Computersand Industrial Engineering 48, 327–356.

Schmidberger, S., Bals, L., Hartmann, E., Jahns, C., 2009. Ground handling servicesat European hub airports: development of a performance measurementsystem for benchmarking. International Journal of Production Economics 117,104–111.

Sugeno, M., 1974. Theory of fuzzy integrals and its applications. Ph.D. Dissertation,Tokyo Institute of Technology, Japan.

Sugeno, M., Terano, T., 1977. A model of learning on fuzzy information. Kybernetes6, 157–166.

Sattler, L., Sohoni, V., 1999. Participative management: an empirical study of thesemiconductor manufacturing industry. IEEE Transactions on EngineeringManagement 46, 387–398.

Sandstrom, J., Toivanen, J., 2002. The problem of managing product developmentengineers: can the balanced scorecard be an answer? International Journal ofProduction Economics 78 79–80.

Terano, T., Asai, K., Sugeno, M., 1992. Fuzzy Systems Theory and Its Applications.Academic Press Professional, Inc., San Diego, CA, USA.

Tahani, H., Keller, M.J., 1990. Information fusion in computer vision using the fuzzyintegral. IEEE Transactions on Systems, Man and Cybernetics 20, 733–741.

Watson, D.J.H., 1975. Contingency formulations of organization structure:implications for management accounting. In: Livingstone, J.L. (Ed.), ManagerialAccounting: The Behavioral Foundations, pp. 65–80.

Ward, P.T., Duray, R., 2000. Manufacturing strategy in context: environment,competitive strategy and manufacturing strategy. Journal of OperationsManagement 18, 123–138.

Wang, W., Wang, Z., Klir, G.J., 1998. Genetic algorithms for determiningfuzzy measures from data. Journal of Intelligent and Fuzzy Systems 6,171–183.

Youngblood, A.D., Collins, T.R., 2003. Addressing balanced scorecard trade-offissues between performance metrics using multi-attribute utility theory.Engineering Management Journal 15, 11–17.

Zadeh, L.A., 1975. The concept of a linguistic variables and its application toapproximate reasoning, Parts I–III. Information Sciences 8 (3), 199–249; 8(4),301–357; 9(1), 43–80.

Zirger, B.J., Hartley, J.L., 1996. The effect of product acceleration techniques onproduct development time. IEEE Transactions on Engineering Management43 (2), 143–152.

int. j. production economics - biu · maintain competitive advantage high tech ﬁrms must...

Documents