The Influence of Total Quality Management, Concurrent Engineering and Knowledge Management in a Semiconductor Manufacturing Firm 1
P. K. Ng1, G. G. G. Goh2, U. C. Eze2 Faculty of Engineering and Technology, Multimedia University, Malacca, Malaysia 2 Faculty of Business and Law, Multimedia University, Malacca, Malaysia (
[email protected],
[email protected],
[email protected])
Abstract - For many years, total quality management, concurrent engineering and knowledge management have won considerable attention from industrial practitioners and academia. However, few studies have been conducted on the influence of these three practices among Malaysian manufacturing firms. Hence, the objective of this study is to analyse the influence of TQM, CE and KM on engineering performance in a Malaysian semiconductor manufacturing firm. For this study, surveys were used to obtain empirical data on these three practices. The data was analysed using multiple linear regression analysis. The findings indicated that TQM, CE and KM significantly influences the firm’s engineering performance with the three predictors explaining up to 53.9% of the variance in engineering performance. The findings of this study are useful to managers, engineers and researchers as it provides insights on specific areas that require adequate attention to ensure effective engineering performance.
practices and affective commitment within the context of the Malaysian semiconductor manufacturing industry [6]. In addition, very few studies have been conducted on the factors that promote or impede engineering in the Malaysian manufacturing firms. Few studies have been carried out on the systemic affects of industrial practices such as TQM, CE and KM on engineering performance in Malaysian manufacturing firms. Hence, the research question developed is: ‘How do TQM, CE and KM influence engineering performance in a semiconductor manufacturing firm?’ II. LITERATURE REVIEW A. Total Quality Management (TQM) TQM basically refers to structures of applications with organized or methodical outcome on organization practices and effectiveness [7]. In addition to that, Kaynak [1] suggests that it refers to comprehensive organizational philosophies striving in continual and functional development that is realized by quality concepts employed commencing at resource acquisitions right up toward consumer relations subsequent to retails. Also, TQM refers to management techniques on promoting a firm’s main viable capability along with attaining whatever that is best in the economy contained by it’s business [8]. According to Forza and Filippini [9] the TQM concept is characterized with directions toward value that assists in preventing crises as well as producing continual enhancement in current conditions. This notion should be disseminated to every level in a firm. A firm that plans to employ TQM as a rudimentary approach for its actions must bear in mind various components of the environment possess attributes that can possibly be damaging to accomplish its objectives [10]. Fuentes-Fuentes et al. [10] expostulate that environment-organization interfaces cause the divergences in the outcomes reached by TQM implementation will also be stipulated by the environment factor. Prajogo and Hong [11] suggest that in spite of plentiful research concerning links connecting quality with organizational performances amid immense acceptances in quality philosophies, there is but little thorough experiential research regarding associations connecting quality along with research and development performances. Furthermore, even if it is acknowledged
Keywords – TQM, concurrent engineering, KM, engineering performance
I. INTRODUCTION For nearly two decades, both the media and academia have published numerous reports and evidence relating the success and failures at employing TQM [1]. Bou-Llusar, Escrig-Tena, Roca-Puig and Beltran-Martın [2] mention that in the past 20 years, firms have used famous models such as the Deming Prize Model, Malcolm Baldrige National Quality Award Model and European Quality Award Model to develop a framework for succesfully employing TQM activities. Apart from TQM, as an advanced approach for manufacturing and new product development, concurrent engineering (CE) has been researched thoroughly and applied widely in the manufacturing industry to considerably speed up production schedules and reduce costs [3]. Initiatives to attain a knowledge-based competitive advantage have begun to intensify the significance of bridging actions and tactical coalitions where the importance of institutional knowledge, system improvement and knowledge management is more emphasized [4]. The development of TQM practices should provide useful measures for investigating the relationship between TQM practices and job satisfaction particularly in relation to the major Malaysian semiconductor organization where studies are yet to be conducted [5]. There is still a lack of studies on the link between TQM
978-1-4244-8503-1/10/$26.00 ©2010 IEEE
240
Proceedings of the 2010 IEEE IEEM
that TQM is capable of engendering a maintainable competitive edge, there is, astoundingly, hardly any postulation to strengthen that confidence [12]. While there is no doubt about subjects concerning TQM as well as innovativeness being handled comprehensively at numerous angles, investigations regarding association linking these notions prove to be minimal [13]. According to Prajogo and Hong [11] employing total quality management at research and development contexts becomes taxing in contrast with additional organizational implementations because research and development functions are mainly dependable on innovativeness. Therefore, this study focuses on the effects of TQM in the context of a Malaysian semiconductor manufacturing firm that consists of various manufacturing and development departments.
doubtful to obtain absolute design outcomes from the upstream design workflow. Haque, Pawar and Barson [22] indicate that the basic obstacle in attaining crossfunctional assimilation is the level whereby firms can acclimatize to their organization formations and methods to fit demands. Thus, this study is to gauge the effects of concurrent engineering applications in the context of a Malaysian semiconductor manufacturing firm which possesses multiple levels of management and administration. C. Knowledge Management (KM) Gold, Malhotra and Segars [23] distinguish that the more frequently a company carries out its KM processes, the more routine the norms and more efficient the integration process. According to Liebowitz [24] knowledge management refers to processes in generating importance using a firm’s indefinable advantages, uniting conceptions of practical AI, computer technology, industrial re-engineering, organisational performance as well as other IT related areas. The greatest KM undertakings are emphasized more on enabling than on managing the flow of knowledge [25]. Knowledge sharing is observed to be a groundwork on continual enhancement in IT processes as well as accordingly in resultant manufactured goods [26]. Liu, Chen and Tsai [27] theorize that firms with good KM approaches will have flourishing NPD performance. They also conceive that knowledge shared within communities enable technologists to employ process enhancement and assume new measures and new products, making Communities of Practice (COP) a helpful formation for technology and KM. According to Faniel and Majchrzak [28], an engineer is apt in using facts sourced from supplementary areas once these facts are obtainable by synopsis as well as thorough levels so as for them to better understand the knowledge and establish roles the knowledge may contain in current problems. Additionally, even though substances in KM between customer as well as supplier in the context of NPD has been acknowledged, moderately little investigation is done on the interorganizational socialization methods that make it possible [29]. According to the aforementioned literature review on TQM, CE and KM, the following hypothesis is able to be proposed: Hypothesis 1:TQM, CE and KM influences engineering performance in a semiconductor manufacturing firm.
B. Concurrent Engineering (CE) CE refers to interdisciplinary collaborations as well as corresponding efforts to achieve universal targets in NPD, production along with product retailing [14]. CE which differs from the customary sequential design technique proves to be of methodical approaches for assimilating concurrent product designs as well as the associated process involved [15]. In reality, lead time, costs, economic conditions as well as technical performance are interconnected characteristics and the CE tactic aims in blending every single characteristic and bestow a general outline for firms [16]. The basic concern in CE is to make available all relevant information to an agent involved in the design process before the design task is begun, whereby the full exploitation of this discernment and the facility to disseminate constructive information on a quick basis is mandatory [17]. Koufteros, Vonderembse and Doll [18] believe that concurrency acts as the instrument to decrease improbability as well as vagueness for improving a firm’s competing advantages by encouraging debates, clarifications, enactments and enabling knowledge dissemination throughout firms rapidly as well as efficiently. In CE processes, there is not only an overlap among the upstream design work and the downstream design work, but also reengineering of NPD by involving process design engineers in product design engineering at the beginning phase [19]. CE projects engage the institution of cross-functional design groups to concomitantly reflect on a variety of actions all through the whole product life cycle [20]. However, Valle and Vazquez-Bustelo [21] point out that latest study demonstrates CE’s inabilities in attaining optimistic outcomes along with the level in equivocality as well as difficulty that exist during innovative processes, causing it to affect impacts on development attributes in performances. Zheng, Wang and Yan [19] also stress that if downstream design workflow launches prematurely, it is
III. METHODOLOGY A. Survey-based method In this study, the survey method is an appropriate approach as it involves the development of a questionnaire based on the hypotheses emerging from the literature review. The use of survey research therefore
241
Proceedings of the 2010 IEEE IEEM
allows the determination and evaluation of the relationships between the variables identified in the literature review. A survey is capable of generating quantitative descriptions of several characteristics of a studied population [30, 31]. For this study, an explanation type of survey is adopted. This is because, the explanation method of survey will test the hypothesis derived from the literature review. The test result will then validate the existence of the association among the variables in the framework. The surveys were handed out to all project leader personnel in the firm. The population of the study consists of all the project leaders, managers and development personnel in the firm. Based on figures provided by this firm on projects in the last 2 years (since 2009), the firm had 3000 projects in total. Due to high turnover rate, transfers and resignation of project leaders, some projects are discontinued. As such, the unit of analysis for this study is the leader’s respective projects in the organization. Thus a total of 2100 surveys were handed out to the respondents of the firm according to workable projects. Duration of 6 weeks was used to gather the data. The response attained was 226 usable surveys collected back out of the 2100 surveys handed out, which produced a response rate of 11 percent. The data was gathered and analysed using the SPSS 18, a quantitative analysis application used for statistical analysis. Due to the relatively large population of project leader personnel in the firm, this study will conduct sampling and also a census of the human resources division in the firm. In this study, the attitudinal scales used are sevenpoint Likert-type scales. This scale is chosen because psychological research has proven that people will have complexities reliably making more than seven distinctions [32]. Furthermore, Linton [33], Editor-inChief of Technovation highlights that ‘past practices such as binary (yes/no) variables, five-point Likert-type scales … is increasingly unwelcome in journals’.
IV. RESULTS A multiple linear regression using the stepwise method was conducted to evaluate ‘Hypothesis 1: TQM, CE and KM influences engineering performance in a semiconductor manufacturing firm.’ Multiple regression is a family of techniques that can be used to explore the relationships between one continuous dependent variable and a number of independent variables or predictors [36]. The core outcome of the regression analysis is the R2 value, which denotes how much of the variance of the dependent variable is explained by the model [37]. Tabachnick and Fidell [38] provide a formula for calculating sample size (N) requirements for regression. They explain that the adequate sample size should be N > 50 + (8 x the number of independent variables). With each individual regression analysis that was conducted the researcher justified the sample size based on this recommendation and the number of variables that were included in the equation. The total number of respondents is 226 while the total number of independent variables tested is 3 (TQM, KM and CE) for Hypothesis 4. Using the formula provided by Tabachnick and Fidell, the minimum sample size required would be 50 + (8 x 3) or 74 respondents. As such, the sample size criterion is met for this study. In addition to that, there is a need for the nonexistence of multi-collinearity in multiple regression analysis. ‘Multi-collinearity” refers to the relationships among the independent variables. Thus, multicollinearity exists when the independent variables are highly correlated at a correlation coefficient (r) of 0.9 and above. Correlation matrixes are recommended to be derived prior to regression analysis to test this [38]. Having satisfied the assumptions for regression analysis, both independent variables (TQM, CE and KM) were regressed against engineering performance and the results are summarized in Table I. TABLE I MULTIPLE LINEAR REGRESSION FOR ENGINEERING PERFORMANCE
B. Multiple Linear Regression Multiple linear regression is based on dichotomous, linear combinations of interval or dummy independent variables and it is used to predict the variance in an interval dependent variable [34]. In other words, ‘multiple regression is an analysis of association in which the effects of two or more independent variables on a single, interval-scaled or ratio scaled dependent variable are investigated simultaneously’ [35]. Multiple linear regression is chosen among other multivariate techniques for this study because it can examine the result of several independent variables on a dependent variable [35]. Therefore, it can ascertain the influence of the TQM, CE and KM (independent variables) towards engineering project performance (the dependent variable).
Predictor
β
Std. Error 0.172
t
F
R
R2
8.409*** (Constant) 1.443 TQM 0.154 0.075 2.060* 0.539 0.734 86.680*** CE 0.218 0.081 2.683** KM 0.283 0.069 4.134*** (Notes: * p
