Shared knowledge and product design glitches in integrated product development

ARTICLE IN PRESS

Contents lists available at ScienceDirect

Int. J. Production Economics

Int. J. Production Economics 114 (2008) 723– 736

0925-52

doi:10.1

� Cor

E-m

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

Shared knowledge and product design glitches in integratedproduct development

Rauniar Rupaka,�, Doll Williamb, Rawski Gregc, Hong Pauld

a Department of Management and Marketing, College of Business, University of St. Thomas, Houston, TX 77006, USAb Department of Management, College of Business, University of Toledo, Toledo, OH 43606, USAc Department of Management, College of Business, University of Evansville, Evansville, IN 47722, USAd Information Operations Technology Management, College of Business, University of Toledo, Toledo, OH 43606, USA

a r t i c l e i n f o

Article history:

Received 21 April 2007

Accepted 20 March 2008Available online 3 April 2008

Keywords:

Product glitches

IPD process

Shared knowledge

Project performance

73/$ - see front matter & 2008 Elsevier B.V

016/j.ijpe.2008.03.005

responding author.

ail address: [email protected] (R. Rauniar

a b s t r a c t

Product development processes based on the joint collaboration of the cross-functional

team, suppliers, and customers can minimize project glitches. Glitches in the product

development project can cause project cost over-runs and delay a project past when first

mover advantages are possible. While previous theoretical work has suggested a negative

relationship between shared knowledge and product development glitches, empirical

studies have not identified how different types of shared knowledge are associated with

each other and the design glitches. This study proposes a model of the relationship

between specific types of shared knowledge and design glitches in integrated product

development (IPD) projects. We test our model using a sample of 191 projects from the

automotive industry in the United States. The major findings were that: (1) shared

knowledge of the development process can be built by improving a team’s shared

knowledge of customers, suppliers, and internal capabilities, (2) shared knowledge of the

development process for a project reduces product design glitches, and (3) reduced

product design glitches improve product development time, cost, and customer

satisfaction.

& 2008 Elsevier B.V. All rights reserved.

1. Introduction

The competitive business environment requires thedesign and development of high-quality innovative pro-ducts that are glitch-free and also mandates that theprocess of introducing new products to the market bestructured and managed appropriately. Glitches in productdevelopment projects are the differences between therequirements of customers, suppliers, and manufacturing/assembly and the actual deliverables. Such differences inrequirements on planned vs. realized means that theparticular stage(s) in the product development processfailed to deliver its target requirements. Glitches translate

. All rights reserved.

).

into the deficiency of product quality and can hamperboth the project and product performance.

In order to improve the project performance, manu-facturing firms are increasingly relying on integratedproduct development (IPD) processes for product devel-opment. IPD include some of the best practices such asconcurrent engineering (Krishnan and Ulrich, 2001;Roemer et al., 2000), customer involvement (Griffin andHauser, 1993), supplier involvement (Dowlatshahi, 1998),and the use of cross-functional team (Clark and Wheel-wright, 1993). The strategic benefits of IPD have beenfound to include reducing time (Gupta and Wilemon,1990; Blackburn, 1991), cutting costs (Hartley, 1990;Handfield, 1994), enhancing quality (Zairi, 1994), effectivedesign of product and process (Rosenthal, 1992), andmanufacturability (Swink, 1999).

The key to fast and effective product development is tolearn quickly about and shift with uncertain environments

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736724

and create structures accordingly (Eisenhardt and Tabrizi,1995). In order for team to learn, knowledge that residesamong its members need to be shared and then mappedinto a shared knowledge of process to exploit thisemergent knowledge base. Product development is a formof problem solving (Clark and Fujimoto, 1991; Thomkeand Fujimoto, 2000) where experiments are conducted todetermine an unknown solution space of process para-meters, which optimize or satisfy a set of processingobjectives (Pisano, 1996). This experiential learningprocess may avoid tradeoffs between cost, quality, andcustomer satisfaction.

The design and execution of highly inter-dependentconcurrent tasks in IPD projects to simultaneously meetvarious goals and expectations can be challenging and thepossibilities of glitches are therefore high. The learningchallenges (Arlati et al., 1995; Zha et al., 2003) include: (1)poor and inadequate description of the parts, components,and the inter-dependence by the customers, (2) the lack ofupfront representation of all the necessary informationand knowledge, (3) the inclusion of new and emergentexpectations, process improvements, and new technolo-gies during project execution, and (4) lack of standardized,conventional, and unified approach for decision makingand problem solving.

Against the challenges and difficulties of IPD projects,Bhuyian et al. (2006) highlight the importance of having aclearly identified IPD process. As a knowledge-intensiveactivity (Lang et al., 2002; Thomke and Fujimoto, 2000;Sureyskar and Ramesh, 2001), IPD process reflects whatthe team knows about the customers, products, pastsuccesses and failures, complex processes, and handoffsbetween the functions (Sureyskar and Ramesh, 2001).When the cross-functional team has a shared under-standing about customers, suppliers, and their own cross-functional capabilities, project processes can be plannedthat effectively integrates the inter-dependent teamknowledge. This should help the project in minimizingglitches and improve the project performance. Previousstudies have focused on the importance of shared teamknowledge of customers, internal capabilities, and sup-plier requirements (Hong et al., 2004). However, we donot know to what extent this shared knowledge ofcustomers, suppliers, and internal capabilities help a teamdevelop a shared knowledge of the product developmentprocess that improves multiple measures of projectperformance.

Although Hoopes and Postrel (1999) and Hoopes(2001) identify the importance of shared knowledge inprojects to resolve design glitches, they do not suggest anyrelationships between types of shared knowledge andglitches. They do not examine whether specific types ofshared knowledge (customers, internal capabilities, andsupplier) work through a shared knowledge of process toimpact project outcomes. Based on the knowledge-basedview of product development, this paper examines therelationship between three types of shared knowledge (ashared knowledge of the customers, suppliers, internalcapabilities) and a shared knowledge of the IPD processesthat is essential to reducing design glitches and improvingproject performance.

2. Theory development

IPD project process reflects a comprehensive networkof work breakdown structures or various stages and theirinter-dependencies. Such inter-dependent relationshipamong various concurrent stages is defined in terms oftask inter-dependencies and information inter-dependen-cies. At each stage of the IPD project, development teamsare deeply involved in problem solving and decisionmaking. Individual team members are assigned to theIPD projects because of the functional and technicalknowledge, skills, and experiences that is relevant forproject execution. Due to the reasons of boundedrationality, individuals in cross-functional IPD team haveto rely heavily on the available information and knowl-edge of other team members. IPD projects are designedand implemented to exploit such individual assets bytransforming it into collective team knowledge andstrength. Knowledge sharing is therefore an importantactivity in IPD project environment.

In the current study, we refer shared knowledge as theshared, common understanding of the IPD team. Langet al. (2002) indicated that successful collaboration inproduct development project requires cognitive synchro-nization/reconciliation, developing shared meaning, de-veloping shared memories, negotiation, communicationof data, knowledge, information, planning of activities,tasks, and methodologies, and management of tasks.Through shared knowledge, teams develop ‘‘transactivememory,’’ in which members have knowledge aboutwho knows what (Moreland, 1999). Shared team knowl-edge of the project process represents shared mentalmodels of the task domain, procedures, and task inter-dependencies (Lang et al., 2002; Cooke et al., 2000) thatare used in the collective problem solving and innovativesolution finding (Davenport et al., 1996; Kogut and Zander,1992). According to Cannon-Bowers et al. (1993), suchshared knowledge helps the team to understand theexpectations of one another and is directly related to theteam performance.

Within the knowledge management literature, re-searchers (e.g., Alavi and Leidner, 2001; Kim, 1993;Polanyi, 1983) differentiate between declarative knowledge

and procedural knowledge. Declarative knowledge orknowledge on something (‘‘know-what’’) refers to thefacts and objects. Within the context of IPD project, itrefers to the fragmented, individual’s, specialized func-tional knowledge that is used to complete varioustechnical activities in the project. Procedural knowledgeor process knowledge (‘‘know-how’’) concerns the waycognitive process and actions are performed. It is throughthe improvisations, practices or ‘‘doing’’, that knowledgeis best utilized for organizational advancements as in thecase of new product development (Orlikowski, 2002). The‘‘cognitive process’’ facilitates the cross-functional team totrigger different concurrent stages and activities basedupon the availability and inter-dependencies of informa-tion and knowledge. In addition to know-what and know-how, researchers (for example, Alavi and Leidner, 2001;Zack, 1998) have also emphasized on conditional knowl-

edge (or ‘‘know-when’’) which refers to the timing and

ARTICLE IN PRESS

Know-How andKnow-When Emerging Knowledge of IPD Process

(Activities for criticalpath, key steps,key deliverables, etc.)

Idea Generation, Concept Development, Product Planning

Product Design

Engineering andPrototype

Production and Ramp-up

Know-What Prior or LegacyKnowledge + New Knowledge (Customers, Internal Capabilities, Suppliers)

Reduced Glitches

Improved Customer Satisfaction, Development Time and Development Cost

Fig. 1. An illustration of IPD process and knowledge transfer (Modified from Soderquist, 1997, all dotted arrows represent flow of information and

knowledge).

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 725

sequencing of knowledge to be applied for successfuloutcomes.

Knowledge of customers, suppliers, and the cross-functional team (or know what) is valuable only to theextent that such knowledge is used and exploited in theproject activities. As Spender (1996, p. 64) observes,knowledge is less about truth and reason and more aboutthe practice of intervening knowledgeably and purpose-fully in the world’’. Therefore, within the IPD context, it isnot only important that all the necessary specialistknowledge domains are present, but also that such

knowledge is shared so that it is exploited knowledgablyand purposefully by the IPD team.

Fig. 1, which is modified from Soderquist (1997),captures key stages of the IPD process from a knowl-edge-based perspective. First, team members may havegained experience and knowledge from prior develop-ment projects (Roemer et al., 2000). This legacy or priorknowledge is shared among the various team members tocreate a shared knowledge base of the requirements ofcustomers, suppliers, and operations (i.e., know-what).Second, as the development process begins, emerging

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736726

knowledge of the IPD process (e.g., activities in the criticalpath, key steps, timing of handoffs, and key deliverables)is created by an experiential process of new productdevelopment (Eisenhardt and Tabrizi, 1995). In thisiterative learning process, activities take place concur-rently and feedback from each step informs and improvesconcurrent steps. Third, this learning process reducesglitches and, thereby, improves product developmentperformance outcomes such as time, cost, and customersatisfaction.

Shared knowledge of customers refers to the extent of ashared understanding of current customers’ needs andfuture value to customers (Hong, 2000; Griffin andHauser, 1993; Calantone et al., 1995). The extent of theshared knowledge of customers is an indication of acontinuous, proactive disposition toward creating highcustomer values across the project and is the mostfundamental aspect of product development (Deshpandeet al., 1993). Several studies show that a firm can acquireknowledge from its customers that can be used for furthermarket entry and expansion.

Shared knowledge of internal capabilities refers to theextent of a shared understanding of the firm’s design andengineering, process, marketing, manufacturing, andother functional capabilities among product developmentmembers (Hong, 2000; Clark and Wheelwright, 1993;Garvin, 1993). A clear understanding of each other’sstrengths can help the IPD team develop product devel-opment processes that maximize the knowledge re-sources of the team members and makes informationavailable as and when needed during the various inter-dependent stages of the project.

Shared knowledge of suppliers refers to the extent of theshared understanding of suppliers’ design, process, andmanufacturing capabilities among product developmentteam members (Hong, 2000; Slade, 1993). According toSharma and Johanson (1987), the firm’s relationships withits suppliers are the most important assets of the firm.Dowlatshahi (1998) developed a framework for imple-menting early supplier involvement, which addressed thestages and interactions among procurement, manufactur-ing, marketing, and design during the product develop-ment process for qualified suppliers. Developmental firmscombine existing knowledge with knowledge from otherpartners to create new knowledge (Hakansson andSnehota, 1995). Shared knowledge of suppliers can assistthe IPD team to seek different types of feed-forward andfeed-backward information and knowledge regardingsuppliers’ technological and process capabilities andconstraints. Such integrative efforts of suppliers with thefocal firms can lead to better business performance(Koufteros et al., 2005).

Shared knowledge of IPD process is the extent of a sharedunderstanding of firm’s concurrent and cross-functionalproduct development process among IPD team members(Rauniar, 2005). Careful management of overlappedactivities requires the detailed representation of informa-tion exchange between individual tasks and a deeperunderstanding of the properties of information (Krishnanet al., 1997; Loch and Terwiesch, 1998; Bhuyian et al.,2006). Team members can thus have a clear knowledge of

the timing and sequence of development activities(Krishnan et al., 1997), project milestones and plannedprototypes (Terwiesch and Loch, 1998), and the relativepriority of development objectives (Ittner and Larcker,1997). In their meta-study Krishnan and Ulrich (2001)focused on various kinds of decisions that need to bemade during the development process. Insights about thenature of development tasks can foster communicationwhere it is most valuable (Moenaert and Soulder, 1996;Henderson, 1994; Griffin, 1992; Clark and Fujimoto, 1991)leading to superior project performance.

Glitches hamper the product development process.These glitches are an indication that the basic processes ofproduct development (know-what, know-how, and know-when) are not well understood. Lack of shared knowledgeabout a particular concurrent activity and its inter-dependencies with other upstream and downstreamconcurrent stages can lead the project to proceed withoutbeing able to meet the expectations of customers,suppliers, or functional specialists. This failure to meetthese expectations often results in rework, cost over-runs,delays, and unsatisfied customers. The negative conse-quences of glitches in overlapping stages are amplifiedwhen glitches go undetected across multiple concurrentstages of the project.

According to Lin et al. (2008), re-work in concurrentprocess can stem from development errors (incorrecttasks) and from corruption (the effect of incorrect taskfrom upstream activities on the downstream activities). Inorder to fix the glitch, the IPD team has to revisit variousstages of the process to investigate the cause(s) andeffect(s) of a particular glitch for remedy; taxingthe project with valuable project time, cost, and raisingcustomer issues. Such glitches can propagate to lossin sales and market share (Hendricks and Singhal, 2003).According to Clark and Fujimoto (1991), product quality,development time, and efficiency, measured in termsof product development cost, should be used to measurethe success of any product development process. Accord-ingly, when glitches and associated reworks in down-stream stages of IPD process are reduced, projectperformance in terms of customer satisfaction, productdevelopment time, and product development cost can beimproved.

3. Research model and hypotheses

Fig. 2 presents our research model that identifies therelationships of such know-what (shared knowledge ofcustomers, internal capabilities, suppliers), know-howand know-when (shared knowledge of process), productdesign glitches and project performance. We also ac-knowledge that, as past studies have pointed out, knowl-edge sharing of customer, suppliers, and others are alsoduring project execution.

Unlike the previous studies on knowledge sharing insupply-chain management (for example, Armistead andMapes, 1993; Anderson and Fine, 1999; Bhaskaran, 1998,etc.), the focus of current study is on the processknowledge in the popular IPD project context and its

ARTICLE IN PRESS

Shared Knowledge of

Customers

Shared Knowledge of

Internal Capabilities

Shared Knowledge of

Suppliers

Shared Knowledge of IPD Process

Product Design Glitches

H1: +

H2: +

H3: +

H4: -

Product Development

Time

Product Development

Cost

Customer Satisfaction

H5: -

H6: -

H7: -

Fig. 2. Proposed research model.

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 727

impact on product design glitches and project perfor-mance. As suggested by Clark and Fujimoto (1991), we useproduct development time, product development cost,and customer satisfaction to investigate the effect ofglitches on project process performance.

Based on the research model, we present testablehypotheses next.

3.1. Hypothesis 1

Shared knowledge of customers’ requirements en-hances the likelihood of meeting the changing needs ofcustomers and coping with the internal dynamics of theproject (Holak and Lehmann, 1990). Compare to business-to-consumers, customers’ involvement in business-to-business is obvious and important (Garvin, 1993). Grunerand Homburg (2000) found a positive effect of newproduct success on the intensity of customer interactionin the first two stages of IPD, idea generation and productconcept development. Careful and thorough attention ofthe team to the translation of customer’s wants, needs,and issues into various stages of the IPD process isessential to the development of high-quality productsthat meet or exceed customers’ expectation and (Griffinand Hauser, 1993). Such shared knowledge of customers(or know-what of customers) would ensure that deliver-ables from each stage to the successive stage(s) of the IPDproject paths are properly aligned toward the customer’sexpectations. We therefore propose:

H1. Shared knowledge of customers has a positive effecton the shared knowledge of IPD process.

3.2. Hypothesis 2

The impacts of cross-functional integration on thestages of IPD have been studied fairly extensively(Bhuyian et al., 2006; Griffin and Hauser, 1993; Song andParry, 1993). To be successful, the inter-dependent stagesof the IPD process require access to deep, concentratedsources of functional expertise and technological knowl-edge (Susman and Dean, 1992), or the know-what of

project team. Each functional expert on the IPD team canprovide others the specific information that is needed tomake coherent decisions due to inter-dependenciesamong functional tasks in the concurrent environment(Bhuyian et al., 2006). This information exchange reducesambiguity in such projects (Koufteros et al., 2005). A moresubtle advantage occurs when one functional groupengages the other toward a more thorough and completeanalysis of its own issues. The later group receives thebenefit of an outsider’s perspective offered by the formergroup. This stimulates creativity and the creation of newknowledge (Ford and Randolph, 1992). When the teammembers develop shared knowledge of each other, like ajigsaw puzzle, each solution developed in the variousconcurrent stages should be able to integrate well withother solutions. This should help the project maintainproduct (or internal) integrity and meet customer ex-pectations (or external integrity). Team efforts and theexpectations can be aligned at various stages in theconcurrent development environment if the teams areaware of each other’s critical roles and responsibilities.

H2. Shared knowledge of internal capabilities has apositive effect on the shared knowledge of IPD process.

3.3. Hypothesis 3

Proactive management of suppliers’ involvement inproduct development can influence the competitivesuccess of organizations (Koufteros et al., 2005; Brownand Eisenhardt, 1995). Suppliers’ involvement must becoordinated internally within the project (Takeishi, 2001)for timely and cost-effective decision making (Evans andLindsay, 1993). Holmen and Kristensen (1998) proposetwo alternative approaches of supplier involvementthrough (1) task partitioning and (2) an interactiveproduct development process. The choice for either taskpartitioning or interaction is primarily related to thecharacteristics of the underlying product including itsdependencies and interfaces, and the knowledge of thesupplier about these characteristics. Based upon theproject requirement, shared knowledge of supplier, or

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736728

the know-what of suppliers, can help the IPD team tounderstand the opportunities, constraints, and expecta-tions of the supplier at different stages of the project.

H3. Shared knowledge of suppliers has a positive effect onthe shared knowledge of IPD process.

3.4. Hypothesis 4

Effective management of product design is central toany project to prevent glitches from creeping their wayacross IPD stages. When the design process is integratedwith knowledge and understanding of customers’ require-ments and suppliers’ capabilities (or the know-what),superior project process can be developed to match thenew product designs (Bhuyian et al., 2006). Such knowl-edge of the IPD process, or know-how and know-when,can help in developing a robust product that is free ofglitches. At each stage, because of the collaborativeenvironment, a cross-functional team ensures that thedeliverables are aligned with the expectations beforethe process owner signoff the responsibility to initiate thesuccessive stage(s). Shared knowledge of process canallow the developers to reduce ambiguity betweenfunction definitions, clarify technical language fromminimal vocabulary, and provide uniformity of informa-tion exchange between IPD stages (Szykman et al., 1999).When there is a shared understanding of project process,glitches in the project can be minimized. Therefore, wepropose:

H4. Shared knowledge of the IPD process has a negativeeffect on the product design glitches.

3.5. Hypothesis 5

Product development time is the time required fromproduct concept to product introduction (Gupta et al.,1992). A product development team that values time tomarket would strive to get products to market ahead ofcompetitors (Blackburn, 1991) develop products on sche-dule (Cohen, 1996) and keep improving on the previoustime to market (Mabert et al., 1992). When glitches areminimized, team spends less time on re-work or correct-ing the glitches. This can help the team to executeconcurrent activities on time and move the project aheadas per the schedule.

H5. Product design glitches have negative effect onlowering the product development time.

3.6. Hypothesis 6

Product development cost is the total cost associatedwith the NPD project to develop and manufacture newproducts. Cost reduction measures the success level of thedevelopment team to reduce product costs (Clark, 1989).A low product cost signifies efficiency in the developmentof the product, in handling uncertainty, and in efficientproblem solving by the cross-functional team members.Lower glitches can help the project in minimizing the

consumption of resources in terms of the cost of materials,labor, and overhead (Garrison and Noreen, 1997).

H6. Product design glitches have negative effect onlowering the product development cost.

3.7. Hypothesis 7

Customer satisfaction measures the satisfaction of thecustomer for the product designed for a certain targetmarket (Cooper and Kleinschmidt, 1987). Glitch-freeproducts add value to customers in terms of meetingcustomer needs and expectations (Clark and Fujimoto,1991; Clark and Wheelwright, 1993), enhancing perceivedproduct quality (Clark and Wheelwright, 1992) anduniqueness (Zirger and Maidique, 1990) and thereforecan lead to product process success (Slater and Narver,1995). Satisfied customers translate into lower handlingcost managing customer complaints, lower warrantycosts, and can help the company to attract new customers(Hong et al., 2004).

H7. Product design glitches have negative effect onimproving customer satisfaction.

4. Research methods

The conceptual model proposed in Fig. 2 provides thefoundation for the empirical research for this study. Anextensive literature review and case studies on previousworks on IPD (Hong, 2000; Rauniar, 2005) and structuredinterviews with product development professionals (man-agers and team-members) and academicians helped todefine the domain of constructs and facilitate itemgeneration. Items generated for each construct in theproposed research model are presented in Appendix A.Shared knowledge of customers, internal capabilities,suppliers, IPD process, product design glitches, and projectperformance have been coded to customers (CT), internalcapabilities (IT), suppliers (ST), IPD process (TP), glitch(DG), and PP respectively. Similarly, product developmenttime, cost, and customer satisfactions were coded asproduct development time (PDT), product developmentcost (PDC), and customer satisfaction (CS). In addition, theitems for DG were coded negatively in respect to otherconstructs. Item descriptions are presented in Appendix A.

For all items, a five-point Likert scale was used; where1 ¼ strongly disagree, and 5 ¼ strongly agree. A differentscale was used for the general demographic questions.Modified items from previous studies were presented totwo product development managers, three product devel-opment team members, and three academicians for theirfeedback. Items were added, modified, deleted, andfinalized on the basis of their qualitative feedback.

The research method in our empirical study includes apilot study and a large-scale study.

4.1. Pilot study

A pilot study from a small sample size can providepreliminary statistical indication on the finalized items for

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 729

the research for any modification before the large-scalestudy. For our research, the Society of AutomotiveEngineers (SAE) provided mailing list of 3200 membersthat included professionals involved in IPD projects. A listof 200 individuals from the SAE mailing list wasgenerated. These individuals were randomly selected fromtheir membership with parameters such as: productdevelopment managers, product development team mem-bers, position titles, functional diversity, product complex-ity and industry position (OEM vs. suppliers). The pilotstudy responses and respondents were excluded in thelarge-scale study.

A total of 34 usable responses (17.7% response rate)were received for the pilot study. All constructs reported ahigh Cronbach’s a value (40.80) and all the items for eachconstruct reported high corrected-item total correlations(CITCs) value, which indicated a good reliability of theitems. Exploratory factor analysis (principal componentanalysis with oblimin rotation) of each construct resultedin separate factors with no cross-loadings among theconstructs.

4.2. Large-scale study

After the pilot study, a survey was administered for thelarge-scale sample to empirically investigate the proposedconceptual research model of Fig. 2. Out of 3000(3200–200) surveys administered, a total of 220 responseswere obtained from the two waves of mail conducted intwo weeks apart. Out of 220 responses received, 191 wereusable resulting in a response rate of 6.3% (191/3000). Thelow response rate prompted us to examine the respondentbias in our targeted sample. A w2-test of differencesbetween observed and expected (population) frequenciesfor company size measured by the number of full-timeemployees (1–499, 500–999, 1000–4999, 5000–9999,and more than 10,000) was conducted. The w2-test showedthat the distribution of our sample fits very well withthe distribution of population (calculated w2 of 6.451ocritical w2 of 9.48). The result helped us to concludethat there existed no such bias between the samplescollected two weeks apart. Additionally, t-tests wereconducted on each of the constructs based upon theearly vs. late respondents. At a ¼ 0.05 level, we did notfind any statistical difference between the early and laterespondents.

Of the total responses received, 28% of the respondentsworked for companies with upto 499 employees, 8% withcompanies having 500–999 employees, 24% with compa-nies having 1000–4999 employees, 12% with companiesbetween 5000 and 9999 employees and 27% withcompanies having over 10,000 employees. These respon-dents represented manufacturing firms that developedand produced diverse product categories for the auto-motive industry. In addition, about 67% representedsupplier companies out of which 78% belonged to first-tier suppliers. Responses to questionnaire in the demo-graphic questions indicated that the respondents hadworked on separate IPD projects.

5. Results

Statistical analysis of our large-scale data includedtests for reliability and factorial validity. We then testeddiscriminant validity to determine if the constructs werestatistically different from one another or not. Finally, weused structural equation modeling (SEM) to test ourmeasurement and structural models.

5.1. Reliability and factor analysis

Composite reliability, for shared knowledge of CT, ST,IT, TP, DG, PDC, PDT, and CS was found to be 0.879, 0.912,0.725, 0.856, 0.836, 0.890, 0.910, and 0.850 respectively.All the items for each construct reported high CITC values,which indicated good reliability of the entire instrument.

Two separate factor analysis, using principal compo-nent analysis and oblimin rotations, was conducted. In thefirst factor analysis, items for shared knowledge ofcustomers, suppliers, and internal capabilities loaded onfactor 1, 2, and 3 with no cross-loadings. Except for IT1,which had a factor loading of 0.46, all the remaining itemshad high loadings compared to the customary value of0.60. In the second factor analysis, items for sharedknowledge of process, design glitches, customer satisfac-tion, product development time, and product develop-ment cost loaded on factor 1, 2, 3, 4, and 5 with no cross-loadings. TP3 had a low loading of 0.39 while the rest ofthe items reported high factor loadings. Based on therelevancy of IT1 for the IT construct, and TP3 for TP, it wasdecided to retain these items but pay close attentionduring the subsequent data analysis. Overall, the factoranalysis demonstrated the factorial validity of our instru-ment. Result of the factorial validity is presented inAppendix A.

5.2. Discriminant validity, correlation matrix, and

descriptive statistics

Discriminant validity is demonstrated when a measuredoes not correlate very highly with another measure fromwhich it should differ (Venkatraman, 1989). The differencein w2 values between restricted and freely estimatedmodels provides statistical evidence of discriminantvalidity (Segars, 1997). To assess discriminant validity,we first computed the differences in w2 values for eachpair of the constructs. The w2 difference between restrictedand freely estimated models was high and significant atpo0.01 for all 28 comparisons illustrated in Table 1, whichsuggested that the constructs are distinct and that theirunderlying scales exhibit the property of discriminantvalidity (Cohen and Cohen, 1983).

To fully satisfy the requirement for discriminantvalidity, average variance extracted for each constructshould be greater than the squared correlation betweenconstructs. Such results suggest that the items sharecommon variance with their respective constructs thanany variance the construct shares with other constructs(Fornell and Larcker, 1981). Table 1 represents the

ARTICLE IN PRESS

Table 1Composite reliability, average variance extracted (AVE), correlation and w2 difference

CT ST IT TP DG PDT PDC CS

Customers 0.88a

(CT) 0.71b

Suppliers 0.12c 0.91a

(ST) 114.5d 0.72b

Internal capabilities 0.33c 0.63c3 0.72a

(IT) 131.9d 51.9d 0.47b

Process 0.47c 0.47c 0.8c 0.86a

(TP) 111d 64.2d 59.9 0.55b

Glitches �0.06c�0.16c

�0.29c�0.31c 0.83a

(DG) 143.9d 114.8d 165.6d 172.1d 0.56b

Product development time 0.33c 0.33c 0.59c 0.52c�0.29c 0.91a

(PDT) 92.4d 50.8d 54.9d 61.2d 143.5d 0.73b

Product development cost 0.12c 0.38c 0.55c 0.29c�0.27c 0.52c 0.89a

(PDC) 120.6d 46.7d 62.9d 86.9d 139.4d 37.6d 0.63b

Customer satisfaction 0.26c 0.35c 0.43c 0.33c�0.34c 0.6c 0.35c 0.88a

(CS) 101.6d 50.5d 77.6d 81.8d 154.2d 32.2d 54.1d 0.66b

Mean 4.25 3.5 3.88 3.98 2.06 3.57 3.6 3.67

Standard deviation 0.71 1.05 0.78 0.8 0.97 0.98 0.98 0.89

Note: po0.01.a Composite reliabilityb Average variance extractedc Correlationd w2 difference.

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736730

correlation matrix with average variance extracted andCronbach’s-a represented on the diagonal.

All the correlations were significant at po0.01 levels.As indicated in Table 1, the average variance extracted is40.50 except for internal capabilities for which theaverage variance extracted was reported to be 0.47.However, the average variance extracted for all constructswas greater than the square of the correlation betweenconstructs. Our results in Table 1 demonstrated thediscriminant validity of the constructs used in theresearch model. The descriptive statistics of the constructsare also presented in Table 1. The mean for each constructranged from 2.06 to 4.25.

5.3. Measurement model

Following Gerbing and Anderson’s (1988) paradigm oftesting SEM models, the measurement model is testedfirst followed by the complete structural model. Thisshould be done to avoid the possible interactions betweenthe measurement and structural models. SEM, usingAMOS 5.0 (Arbuckle, 2003) was implemented to analyzethe measurement and structural models. Multiple indicesof model-data fit should be considered in SEM in assessingoverall model-data fit. We report CMIN/df, CFI, TLI, andRMSEA values for the measurement and structuralmodels. For each of these indices, good model-data fit isindicated by CMIN/df value o2.00 (Wheaton et al., 1977),CFI value 40.9 (Bentler, 1992), and TLI value closer to 0.95(Hu and Bentler, 1999). RMSEA values ranging from 0.05 to0.08 are considered to be adequate model-data fit(Browne and Cudek, 1993).

Measurement models for all the constructs wereanalyzed simultaneously (i.e., a eight-factor correlated

model). The output of this first order measurementmodels for all the constructs is summarized in Table 2.The unidimensional direction, for example from CT1 to CT,suggests that the score values are each influenced by theirrespective underlying factors (Byrne, 2001). The secondcolumn presents the standardized regression weight, thethird column presents the standard error or S.E, the fourthcolumn presents the critical ratio or C.R. (interpreted as z-scores), and the last column presents the significance atpo0.001 level as indicated by ‘‘***’’.

The CMIN/df, TLI, CFI, and RMSEA goodness of fit forthe measurement model were reported to be 1.430, 0.942,0.949 and 0. 048 indicating adequate model-data fit. Theoverall result from measurement model analysis gave usthe confidence to proceed to the next phase on analyzingoverall structural model, without any modification.

5.4. Structural model

Once the satisfactory results for each construct andunderlying items were established in the measurementmodel, we tested the complete structural model for theseven hypotheses presented earlier. The result of thestructural model data analysis is presented in Fig. 3.

The model fit indices for the overall model were CMIN/df ¼ 1.891, TLI ¼ 0.879, CFI ¼ 0.889, and RMSEA ¼ 0.068.The indices indicated adequate model-data fit. The resultsfrom Fig. 3 indicate that all the seven hypotheses weresupported by the data. The first hypotheses posited apositive relationship between the shared knowledge ofcustomers and the shared knowledge of IPD process. Thestandardized regression weight of 0.358 was found to bestatistically significant. The second hypotheses suggesteda positive relationship between the shared knowledge of

ARTICLE IN PRESS

Table 2Results from measurement model

Indicators Construct

Standardized

regression

estimates

S.E.a C.R.b Pc

Shared knowledge of customers

CT1 ’ CT 0.801

CT2 ’ CT 0.852 0.091 12.56 ***

CT3 ’ CT 0.87 0.09 12.677 ***

Shared knowledge of internal capabilities

IT1 ’ IT 0.764

IT2 ’ IT 0.598 0.096 7.670 ***

IT3 ’ IT 0.559 0.099 6.949 ***

Shared knowledge of suppliers

ST1 ’ ST 0.863

ST2 ’ ST 0.743 0.072 12.036 ***

ST3 ’ ST 0.896 0.068 16.544 ***

ST4 ’ ST 0.888 0.066 16.109 ***

Shared knowledge of IPD process

TP1 ’ TP 0.745

TP2 ’ TP 0.724 0.096 9.728 ***

TP3 ’ TP 0.646 0.108 8.558 ***

TP4 ’ TP 0.748 0.101 10.055 ***

TP5 ’ TP 0.829 0.103 11.364 ***

Product design glitches

DG1 ’ DG 0.674

DG2 ’ DG 0.792 0.100 9.047 ***

DG3 ’ DG 0.730 0.104 9.802 ***

DG4 ’ DG 0.867 0.098 8.324 ***

Product development time

PDT1 ’ PDT 0.796

PDT2 ’ PDT 0.773 0.081 11.850 ***

PDT3 ’ PDT 0.927 0.076 15.048 ***

PDT4 ’ PDT 0.903 0.078 14.453 ***

Project development cost

PDC1 ’ PDC 0.873

PDC2 ’ PDC 0.750 0.068 12.328 ***

PDC3 ’ PDC 0.764 0.071 12.661 ***

PDC4 ’ PDC 0.672 0.069 10.536 ***

PDC5 ’ PDC 0.892 0.065 16.458 ***

Customer satisfaction

CS1 ’ CS 0.792

CS2 ’ CS 0.730 0.087 10.523 ***

CS3 ’ CS 0.867 0.076 12.697 ***

CS4 ’ CS 0.842 0.083 12.563 ***

a Standard error.b Critical ratio.c po0.001.

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 731

internal capabilities and the shared knowledge of IPDprocess for which the standardized regression weight of0.587 was also found statistically significant. The stan-dardized regression weight for the third hypothesis, thatstated a positive relationship between the shared knowl-edge of supplier and the shared knowledge of process, wasfound to be 0.229. Similarly, The standardized regressionweight for hypothesis 4, the relationship between theshared knowledge of process and product design glitches,

was �0.353. The negative weight indicated that the lowerthe shared knowledge of process, the more glitches.Similarly, hypothesis 5, 6, and 7 which predicted therelationships between product design glitches and pro-duct development time, product development cost, andcustomer satisfaction had standardized regressionweights of �0.385, �0.340, �0.412, and �0.411, respec-tively. This means that the fewer the glitches, the greaterthe project performance. All the hypothesis were found tobe significant at po0.001.

6. Discussion

Based on knowledge theories, the study proposed andempirically validated relationships of shared knowledgeand glitches in IPD projects. There are, however, severallimitations of our current study. One limitation is that thedata was gathered from the automotive industry wherecustomer and supplier involvement is a common practice.Such practices for product development may not be aspopular in other industries and may require a differentresearch model to study the relationships between sharedknowledge and glitches. The current study excludesimportant contextual factors such as product and tech-nology complexity, degree of innovation, and environ-mental uncertainty. Another shortcoming of this researchcan be viewed from the context of common method biasas data collected on the dependent and independentvariables was obtained from the same respondents.

Against these limitations, the current study never-theless, makes valuable contributions in the area of IPDprojects. Our study highlights the importance of sharedknowledge of process (i.e., know-how and know-when) inIPD projects to meet the expectations and goals andaccordingly minimize glitches in the project. We extendtheories of knowledge management literatures on knowl-edge classifications and empirically map the relationshipsof know-whats and process knowledge, i.e., know-howsand know-whens in the IPD projects. Our study indicatesthat in a highly integrated, overlapped IPD projectenvironments, it is important to have specialists’ knowl-edge actively involved and engaged for the project toprogress in a timely manner.

Shared knowledge of the IPD team helps the team inunderstanding one another’s expectations, motivations,contributions, constraints, and others that can help theteam to proceed with the project in a disciplined andfocused manner. With a collective understanding of goalsand expectations, a collaborative effort can also beexpected for cross-functional problem solving and deci-sion-making activities. Decisions at each stage need to beconsistent regardless of who makes the decisions or whensuch decisions are made in the development process. Thecurrent research also drew its theory from Kofman’s (Kim,1993) OADI (observe, assess, design, and implement) tohypothesize the relationship between shared knowledgeof customers, suppliers, internal capabilities, and productdevelopment process.

During the initial project team formation, the productdevelopment team spends countless hours in refining the

ARTIC

LEIN

PRESS

CMIN/df = 1.891TLI = 0.879CFI = 0.889

SharedKnowledge of

Customers

SharedKnowledge of

InternalCapabilities

SharedKnowledge of

Suppliers

SharedKnowledge ofIPD Process

Product Design Glitches

H1: +, R= 0.358

H2: +, R=0.587

H3: +, R=0.229

H4: -, R=0.353

ProductDevelopment

Time

ProductDevelopment

Cost

CustomerSatisfaction

H5: -, R=-0.385

H6: -, R=-0.340

H7: -, R=-0.412

RMSEA = 0.068

Fig. 3. Results from structural model analysis. Notes: R ¼ Standardized regression weight, significant at po0.001.

R.

Ra

un

iar

eta

l./

Int.

J.P

rod

uctio

nE

con

om

ics114

(20

08

)7

23

–7

36

73

2

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 733

new product concept. Team observations and assessmenthelps the team to develop a shared knowledge baseof know-whats. Following the OADI cycle, the teamthen enters in design and implementation. Productdevelopment process is developed and the projectprogresses from the early planning stage to the executionstage.

Hypotheses 1 and 3 of our research model empiricallydemonstrate the importance of shared knowledge ofcustomers and suppliers in the IPD process. Thesehypotheses reinforce other research works on earlycustomer–supplier involvement in product developmentprojects. For further readings on early customer involve-ment, we recommend to the works of Maidique and Ziger(1985) and Kaulio (1998). Similarly, research by Takeishi(2001), Eisenhardt and Tabrizi (1995) and Culley et al.(1999), provides excellent arguments on early involve-ment of suppliers in product development projects. Theimportance of cross-functional integration to developshared knowledge of internal capabilities is underscoredby the support of hypothesis 2.

Because of high inter-dependencies among the activ-ities and decisions, changes made to fix a glitch in aparticular stage may require subsequent changes in otherstages. A good IPD process should minimize occurrencesof glitches and improve the project performance.Throughout the IPD process, the development teamshould ensure that the project is a multi-functionalactivity, and should ensure that the customer andmanufacturing requirements are being met. Our empiricalresults provide support to our hypotheses that sharedknowledge of the IPD process can minimize productdesign glitches (H4) and improve product developmenttime, cost, and customer satisfaction (H5–H7).

Against the above research findings, it is important toacknowledge the possible tradeoffs among the projectperformance measures if too much project time is spent insharing the know-what knowledge of customers, suppli-ers, and the internal capabilities and not focusing on howand when such knowledge would be utilized during theconcurrent process. Development of collective under-standing can tax the project valuable project time if toomuch time is spent on know-what shared knowledge andtoo little is spent on understanding the project steps,inter-dependencies among the parallel stages, and deli-verables of each stage that constitute the know-how andknow-when or shared knowledge of process.

It is also important to point out that the data used inour analysis came from the US automotive industry.Project-related information and knowledge sharing insuch complex and multi-tier supply-chain industry couldbe expected to be project-goal driven. It can be thereforeexpected that the knowledge sharing of the customers,suppliers, and the development team members can takeplace when there is a clear understanding among themembers that such shared knowledge of customers,suppliers, and internal capabilities has implications onthe project process and performance.

Further, the total quality management practices haveshown that cost, time, and quality are not always tradeoffdecisions if the process is well planned. Process improve-

ment is essential to resolving these tradeoff issues anddoing better on cost, time, and quality. The Japanesepractices has demonstrated that the experiential learningthrough feed-forward and feedback learning of the teamduring the continuous improvement of product andproduction processes facilitates the overall improvementof the performance in the long-run of such initiatives.Extra time and resources spent to facilitate sharedknowledge of customers, suppliers, internal capabilities,and IPD process at the upfront can, as our resultsdemonstrate, lead to superior project performance.

An implication of the current study is that the IPDproject management team must design a logical andflexible workflow process that integrates knowledgeand promotes a shared understanding of the organiza-tion’s products, process technologies, suppliers andcustomers. New requirements in product developmentproject can evolve from various sources. IPD projectscould be dynamic if the new requirements and opportu-nities are included and integrated with the initialproject plan. At a strategic level, the team must designeffective and efficient processes to capture and incorpo-rate knowledge of customers and suppliers in the project.Therefore process knowledge can help in the integrationof cross-functional effort in the concurrent process,minimize the project glitches, and improve the projectperformance.

At a tactical level, managers should encourage theteam to develop shared knowledge of the process so thateveryone has a clear idea about important informationexchange points in and across the stages. Such sharedknowledge of a process can help the team with clarity ofgoals, objectives, and targets for each stage and for theoverall project.

7. Conclusion

An effective and efficient IPD project requires cross-functional collaboration for work and knowledge integra-tion of the team. A lack of shared knowledge in the teamcan lead to occurrences of glitches. For any operations,glitches can be inferred as re-work and wastage ofvaluable resources. The present research investigated theimportance of shared knowledge of process in the IPDproject environment that uses cross-functional team andconcurrent activities. As our research indicates, suchshared knowledge of an IPD process can be developedon the basis of shared knowledge of customers, internalcapabilities, and suppliers.

While the current study focuses on the importance andthe drivers of shared knowledge of an IPD process and itseffect on glitches, it suggests another critical researchquestion: what drives the team members, customers,and suppliers to share knowledge? Based on an initialreview of the literatures, it seems plausible that trust,power, and rewards (financial, non-financial) may beadditional factors related to knowledge sharing that needsfurther study. Another meaningful way to extend thecurrent study is to identify the relationship of work

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736734

integration and knowledge integration mechanism for IPDprojects.

Appendix A. Pattern matrix results from factoranalysis using oblimin rotation (n ¼ 191)

See Table A1 for further details.

Table A1

Constructs Code Item description

Customers (CT)

CT1 customers requirements

CT2 what our customers wanted

CT3 current customers needs

Internal capabilities (IT)

IT1 the capabilities of process technologies we use

IT2 the strength of engineering design capabilities

IT3 the strength of our engineering staffs

Suppliers (ST)

ST1 our suppliers’ process capabilities

ST2 our suppliers’ manufacturing facilities

ST3 our suppliers’ capabilities to meet time requirements

ST4 our suppliers’ capabilities to meet quality requirements

Constructs Code Item description

IPD process (TP)

TP1 The steps in the product development process

TP2 The points in the product development process where inform

TP3 where key deliverables in the product development process w

activities

TP4 the activities in the product development process that were

TP5 key decision points in the product development process

Glitches (DG)

DG1 the product design did not meet customer requirement (s)

DG2 the product design did not meet supplier requirement (s)

DG3 the product design did not meet manufacturing requirement

DG4 the product did not meet assembly requirements

Customer satisfaction (CS)

CS1 The new product has more loyal customers

CS2 The new product generated more new customers

CS3 The new product was more highly valued by customers

CS4 The new product was more successful in the marketplace

Product development time (PDT)

PDT1 The NPD team enabled our company to start volume product

PDT2 The NPD team brought product to the market before our com

PDT3 The NPD team developed product from concept to commerci

PDT4 The NPD team made better progress in reducing total produc

Product development cost (PDC)

PDC1 The NPD team reduced product costs successfully

PDC2 The NPD team reduced material costs successfully

PDC3 The NPD team successfully reduced assembly cost

PDC4 The NPD team reduced equipment cost successfully

PDC5 The NPD team reduced manufacturing cost successfully

References

Alavi, M., Leidner, D., 2001. Knowledge management and knowledge

management systems: Conceptual foundation and research issues.

MIS Quarterly 25 (1), 107–136.

Anderson, E., Fine, C., 1999. Business cycles and productivity in capital

equipment chains. In: Tayur, S., et al. (Eds.), Quantitative Models for

Supply Chain Management. Kluwer Academic Publishers, Boston.

Arbuckle, J.L., 2003. Amos 5.0 to the Amos User’s Guide. Small-Waters

Corp., Chicago, IL.

Factor analysis 1

Factor1 Factor2 Factor3

0.892

0.902

0.899

0.460

0.848

0.905

0.928

0.794

0.921

0.899

Factor Analysis 2

Factor1 Factor2 Factor3 Factor4 Factor5

0.893

ation needed to be exchanged 0.744

ere essential to subsequent 0.390

on the critical path 0.601

0.766

0.764

0.794

(s) 0.865

0.779

0.851

0.758

0.792

0.813

ion faster 0.842

petitors 0.820

al production faster 0.846

t development time 0.778

�0.873

�0.778

�0.676

�0.726

�0.837

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736 735

Arlati, E., Bottelli, V., Christian, F., Tirassa, M., 1995. Modeling processknowledge in architectural design: A case-based approach. In:Proceedings from Eighth International Conference on SystemsResearch, Informatics and Cybernatics, Baden-Baden.

Armistead, Mapes, 1993. The impact of supply chain integration onoperating performance. Logistics Information Management 6, 9–14.

Bentler, P.M., 1992. EQS: Structural Equations Program Manual. BMDPStatistical Software, Los Angeles.

Bhaskaran, S., 1998. Simulation analysis of a manufacturing supply chain.Decision Sciences 29 (3), 657–663.

Bhuyian, N., Thomson, V., Gerwin, D., 2006. Implementing concurrentengineering. Research Technology Management, 38–43.

Blackburn, J., 1991. Time Based Competition. Business One, Irwin,Homewood, IL.

Brown, W., Eisenhardt, K.M., 1995. Product development: Past research,present findings and future directions. Academy of ManagementReview 20, 343–378.

Browne, M.W., Cudek, R., 1993. Alternative ways of assessing model fit.In: Bollen, K.A., Scott Long, J. (Eds.), Testing Structural EquationModels. Sage Publications, Newbury Park, CA.

Byrne, M.B., 2001. Structural Equation Modeling with Amos. LEAPublishers, Mahwah, NJ.

Calantone, R.J., Vickery, S.K., Droge, C., 1995. Business performance andstrategic new product development activities: An empirical investi-gation. Journal of Product Innovation Management 12 (3), 214–223.

Cannon-Bowers, J.A., Salas, E., Converse, S., 1993. Shared mental modelsin expert team decision making. In: Castellan, Jr., N. John. (Ed.),Individual and Group Decision Making: Current Issues. LawrenceErlbaum Associates, Hillsdale, NJ, pp. 221–246.

Clark, K.B., 1989. Project scope and performance: The effect of partsstrategy and supplier involvement on product development. Man-agement Science 35 (10), 1247–1263.

Clark, K.B., Fujimoto, T., 1991. Product Development Performance.Harvard Business School Press, Boston.

Clark, K.B., Wheelwright, S.C., 1992. Organizing and leading ‘‘Heavyweight’’development teams. California Management Review 34, 9–28.

Clark, K.B., Wheelwright, S.C., 1993. Managing New Product and ProcessDevelopment. Free Press, New York.

Cohen, J., Cohen, P., 1983. Applied Multiple Regression/CorrelationAnalyses for the Behavioral Sciences. Lawrence Erlbaum, Hillsdale, NJ.

Cohen, M.A., 1996. New product development: The performance andtime to market tradeoff. Management Science 42 (12), 1735–17511.

Cooke, N.J., Sales, E., Cannon-Bowers, J.A., Stout, R.J., 2000. Measuringteam knowledge. Human Factors 42 (1), 151.

Cooper, R.G., Kleinschmidt, E.J., 1987. New products: What separateswinners from losers. Journal of Product and Innovation Management4, 169–184.

Culley, S.J., Boston, O.P., Christopher, A., 1999. Suppliers in new productdevelopment: Their information and integration. Journal of Engi-neering Design 10 (1), 59–75.

Davenport, T., Jarvenpaa, S., Beers, M., 1996. Improving knowledge workprocess. Sloan Management Review (Winter), 43–57.

Deshpande, R., Farley, J.U., Webster, F.E., 1993. Corporate culture,customer orientation, and innovativeness in Japanese firms: Aquadrad analysis. Journal of Marketing 57 (1), 23–37.

Dowlatshahi, S., 1998. Implementing early supplier involvement: Aconceptual framework. International Journal of Operations &Production Management 18 (2), 143–167.

Eisenhardt, K.M., Tabrizi, B.N., 1995. Accelerating adaptive processes:Product innovation in the global computer industry. AdministrativeScience Quarterly 40, 84–110.

Evans, J.R., Lindsay, W.L., 1993. The Management and Control of Quality.West Publishing Company, New York, NY.

Ford, R.C., Randolph, W.A., 1992. Cross-functional structures: A reviewand integration of matrix organization and project management.Journal of Management 18, 267–294.

Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models withunobservable variables and measurement error. Journal of MarketingResearch 18 (2), 39–50.

Garrison, R.H., Noreen, E.W., 1997. Managerial Accounting, eighth ed.Irwin, Homewood, IL.

Garvin, D.A., 1993. Building a learning organization. Harvard BusinessReview 71 (4), 78–91.

Gerbing, D.W., Anderson, J.C., 1988. An updated paradigm for scaledevelopment incorporating unidimensionality and its assessment.Journal of Marketing Research 25, 186–192.

Griffin, A., 1992. Evaluating QFD’s use in US firms as a process fordeveloping products. Journal of Product Innovation Management 9,171–187.

Griffin, A., Hauser, J.R., 1993. The voice of the customer. MarketingScience 12, 1–27.

Gruner, K.E., Homburg, C., 2000. Does customer interaction enhance newproduct success? Journal of Business Research 49, 1–14.

Gupta, A.K., Wilemon, D.L., 1990. Accelerating the development oftechnology-based new products. California Management Review22, 24–44.

Gupta, A.K., Brockhoff, K., Weisenfeld, U., 1992. Making trade-offs in thenew product development process: A German/US Comparison.Journal of Product Innovation Management 9 (1), 11–18.

Hakansson, H., Snehota, I., 1995. Developing relationships in businessnetworks. Journal of Applied Psychology 74, 884–891.

Handfield, R.B., 1994. Effect of concurrent engineering on make-to-orderproducts. IEEE Transaction on Engineering Management 41 (4),384–393.

Hartley, J.R., 1990. Concurrent Engineering Shortening Lead Times,Raising Quality and Lowering Costs. Productivity Press.

Henderson, R.M., 1994. The evolution of integrative capability: Innova-tion in cardiovascular drug discovery. Industrial and CorporateChange 3, 607–630.

Hendricks, K.B., Singhal, V.R., 2003. The effect of supply chain glitches onshareholder wealth. Journal of Operations Management, 501–522.

Holak, S.L., Lehmann, D.R., 1990. Purchase intentions and the dimensionsof innovation: An exploratory model. Journal of Product InnovationManagement 7 (1), 59–73.

Holmen, E., Kristensen, P.S., 1998. Supplier roles in product development:Interaction versus task partitioning. European Journal of Purchasing& Supply Management 4, 185–193.

Hong, P., 2000. Knowledge Integration in Integrated Product Develop-ment. Doctoral Dissertation. College of Business, University ofToledo, Ohio.

Hong, P., Doll, W.J., Nahm, A.Y., Li, X., 2004. Knowledge sharing inintegrated product development. European Journal of InnovationManagement 7 (2), 102–112.

Hoopes, D.G., 2001. Why are the glitches in product development? R&DManagement 31 (4), 381–389.

Hoopes, D.G., Postrel, S., 1999. Shared knowledge, ‘‘glitches,’’ and productdevelopment performance. Strategic Management Journal 20,837–865.

Hu, L., Bentler, P.M., 1999. Cutoff criteria for fit indices in covariancestructure analysis: Conventional versus new alternatives. StructuralEquation Modeling 6, 1–55.

Ittner, C.D., Larcker, D.F., 1997. Product development cycle time andorganizational performance. Journal of Marketing Research 34 (1),13–23.

Kaulio, M.A., 1998. Customer, consumer, and user involvement in productdevelopment: A framework and a review of selected methods. TotalQuality Management 9 (1), 141–149.

Kim, D.H., 1993. The link between individual and organizational learning.Sloan Management Review, 37–50.

Kogut, B., Zander, U., 1992. Knowledge of the firm, combinativecapabilities and the replication of technology. Organization Science3 (3), 383–397.

Koufteros, X.A., Vonderembse, M.A., Jayaram, 2005. Internal and externalintegration for product development: The contingency effects ofuncertainty, equivocality, and platform strategy. Decision Sciences36 (1), 97–133.

Krishnan, V., Ulrich, K., 2001. Product development decisions: A review ofthe literature. Management Science 47 (1), 1–21.

Krishnan, V., Eppinger, S.D., Whitney, D.E., 1997. A model basedframework to overlap product development activities. ManagementScience 43 (4), 437–451.

Lang, S.Y.T., Dickinson, J., Buchal, R.O., 2002. Cognitive factors indistributed design. Computers in Industry 48, 89–98.

Lin, J., Chai, K.H., Wong, Y.S., Brombacher, A.C., 2008. A dynamic modelfor managing overlapped iterative product development. EuropeanJournal of Operational Research 185, 378–392.

Loch, C.H., Terwiesch, C., 1998. A Model of Preliminary InformationExchange in Concurrent Development. INFORMS, Seattle.

Mabert, V.A., Muth, M.F., Schmenner, R.W., 1992. Collapsing new productdevelopment times: Six case studies. Journal of Product InnovationManagement 9 (3), 200–212.

Maidique, M., Ziger, B., 1985. The new product learning cycle. ResearchPolicy 14 (6), 299–313.

Moenaert, R.K., Soulder, W.E., 1996. Context and antecedents ofinformation utility at the R&D/marketing interface. ManagementScience 42 (11), 1592–1610.

Moreland, R.L., 1999. Learning who knows what in work groups andorganizations. In: Thompson, L.L., Levine, J.M., Messick, D.M. (Eds.),

ARTICLE IN PRESS

R. Rauniar et al. / Int. J. Production Economics 114 (2008) 723–736736

Shared Cognition in Organizations: The Management of Knowledge.Erlbaum, Mahwah, NJ, pp. 2–3.

Orlikowski, W.J., 2002. Knowing in practice: Enacting a collectivecapability in distributed organizing. Organization Science 13 (3),249–273.

Pisano, G.P., 1996. Learning-before-doing in the development of newprocess technology. Research Policy 25, 1097–1119.

Polanyi, M., 1983. The Tacit Dimension. Peter Smith, Gloucester, MA.Rauniar, R., 2005. Knowledge integration in integrated product develop-

ment: The role of mutual trust, mutual influence, team vision onshared knowledge. Doctoral Dissertation. College of Business,University of Toledo, Toledo, OH.

Roemer, T.A., Ahmadi, R., Wang, R.H., 2000. Time-cost trade-offs inoverlapped product development. Operations Research 4 (6),858–865.

Rosenthal, S.R., 1992. Effective Product Design and Development, How toCut Lead Time and Increase Customer Satisfaction. Business OneIrwin, Homewood, IL.

Segars, A.H., 1997. Assessing the unidimensionality of measurement: Aparadigm and illustration within the context of information systemsresearch. Omega 25 (1).

Sharma, D., Johanson, J., 1987. Technical consultancy in internationaliza-tion. International Marketing Review 4, 20–29.

Slade, B., 1993. Compressing the Product Development Cycle. AmericanManagement Association.

Slater, S.F., Narver, J.C., 1995. Market orientation and the learningorganization. Journal of Marketing 59 (3), 63–74.

Soderquist, K., 1997. Inside the Tier Model: Product DevelopmentOrganization and Strategies in Automotive Expert Supplier Firms,Doctoral Dissertation. Henley Management College, Groupe ESCGrenoble, Brunel University.

Song, X.M., Parry, M.E., 1993. R&D-marketing integration in Japanesehigh-technology firms: Hypotheses and empirical evidence. Journalof Academy of Marketing Science 21 (2), 125–133.

Spender, J.C., 1996. Organizational knowledge, learning, and memory:Three concepts in search of a theory. Journal of OrganizationalChange 9, 63–78.

Sureyskar, K., Ramesh, B., 2001. On managerial incentives for processknowledge capture and reuse. In: Proceedings of the Thirty-Fourth

Annual Hawaii International Conference on Systems Sciences. IEEEComputer Society Press.

Susman, G.I., Dean Jr., J.W., 1992. Development of a model for predictingdesign for manufacturability effectiveness. In: Susman, Gerald (Ed.),Integrating Design and Manufacturing for Competitive Advantage.Oxford University Press, New York, pp. 207–227.

Swink, M., 1999. Threats to new product manufacturability and theeffects of development team integration processes. Journal ofOperations Management 17 (6).

Szykman, S., Racz, J.W., Sriram, R.D., 1999. The representation of functionin computer-based design. In: The Proceedings of the 1999 ASMEDesign Engineering Technical Conferences (11th International Con-ference on Design Theory and Methodology).

Takeishi, A., 2001. Bridging inter- and intrafirm boundaries: Managementof supplier involvement in automobile product development.Strategic Management Journal 22, 403–433.

Terwiesch, C., Loch, C.H., 1998. Communication and uncertainty inconcurrent engineering. Management Science 44 (8), 1032–1048.

Thomke, S., Fujimoto, T., 2000. The effect of ‘‘Front-Loading’’ problem-solving product development performance. Journal of ProductInnovation Management 17, 128–142.

Venkatraman, N., 1989. Strategic orientation of business enterprises: Theconstruct, dimensionality and measurement. Management Science35, 942–962.

Wheaton, B.B., Muthen, B., Alwin, D.F., Summers, G.F., 1977. Assessingreliability and stability in panel models. In: Heise, D.R. (Ed.),Sociological Methodology. Jossey-Bass, San Francisco.

Zack, M., 1998. What knowledge-problems can information technologyhelp to solve. In: Hoadley, E., Benbasat, I. (Eds.), Proceedings of theFourth Americas Conference on Information Systems, Baltimore, MD,pp. 644–646.

Zairi, M., 1994. Benchmarking: The best tool for measuring competitive-ness. Benchmarking for Quality Management and Technology 1,11–24.

Zha, X.F., Lim, S.Y.E., Lu, W.F., 2003. A knowledge intensive multi-agentsystem for cooperative/collaborative assembly modeling and processplanning. Transactions of SDPS 7 (1), 99–122.

Zirger, B.Z., Maidique, M.A., 1990. A model of new product development:An empirical test. Management Science 36 (7), 867–883.