1、The measurement of technical competencies$Steven Walsha,1, Jonathan D. Lintonb,*aAnderson School of Management, University of New Mexico, Albuquerque, NM, USAbDepartment of Management, Polytechnic University, Brooklyn, NY 11201, USAAccepted 15 October 2001AbstractTechnical competencies are operation
2、alized and tested on a census of the international semicon-ductor industry. This study is an important step in moving competence and capability from a populartopic of trade publications and theorists to a quantifiable management science. A series of questionsthat can be used for the study of other c
3、ompetencies is developed and tested. Competence is shown tobe a single trait measure, with a high reliability. The reliability and validity of technical competenciesare also considered. D 2002 Published by Elsevier Science Inc.Keywords: Core competence; Competence theory; Process innovation; Technol
4、ogy; Measurement1. IntroductionThe study of core competencies has been hampered by the lack of an accepted way ofoperationalizing this construct. This is unfortunate, since the theoretical roots of core com-petencies is very old (Coase, 1937; Schumpeter, 1934; Smith, 1776) and is currently of intere
5、stto both academe (Morone, 1993; Prahalad Walsh, Boylan, Paulson, Coase, 1937; Frohman, 1980; OBrien, 1962; Penrose, 1959; Schumpeter,1934). Competencies and capabilities are frequently categorized as separate but related con-structs (Barney, 1995; Hitt Marino, 1996). However, confusion results from
6、the different definitions proposed in the literature. Prahalad and Hamel (1990) use the termscompetence and capability interchangeably. Whereas, Javidan (1998) suggests that the termscapability, competence, and core competence refer to the span of the advantage within thefirmwithin a department, wit
7、hin a single SBU or across multiple SBUs. Marino (1996) andothers suggest that competencies are firm-specific technologies and production technologies.Whereas, capabilities are firm-specific business practices, processes, and culture. We havebased our study on Marinos definition; this decision is ba
8、sed on the use of the terms tech-nology and production skills in Prahalad and Hamels (1990, pp. 8182) discussion ofcore competencies. To date, research has identified firms that have achieved success by usingstrategies that focus on a firms core capabilities and competencies (examples include Hitt M
9、orone, 1994; Teece, Pisano, Flynn Von Hippel, 1976; Voss, 1985). Consequently,we propose:Hypothesis 1: Respondents from suppliers and a firm will agree on the presence orabsence of a competence at the firm.Customers are also indicated in the innovation literature as having depth of knowledge incompe
10、tencies related to their field and an interest in understanding the strengths and gaps inthe technical competence of their suppliers (Von Hippel, 1976). Therefore, we propose:Hypothesis 2: Respondents from customers and a firm will agree on the presence orabsence of a competence at the firm.Other fi
11、rms within the industry require the same or similar types of technical expertise(Hamel, Doz, Khanna, Gulati, Schrader, 1991; Teece,1989; Von Hippel Hamel Khanna et al., 1998; Kogut, 1988; Teece, 1989; Von Hippel, 1976,1987). Consequently, we propose:Hypothesis 3: Respondents from a firms competitor(
12、s) and the firm will agree on thepresence or absence of a competence at the firm.Consultant and industry experts within an industry have depth of knowledge in thetechnical competencies that they deal with (Zairi, 1992a, 1992b). Due to their work withfirms and their continued exposure, involvement, a
13、nd monitoring of the industry, they be-come aware of the strength and weaknesses of firms within not only the competencies thatS. Walsh, J.D. Linton / Journal of High Technology Management Research 13 (2002) 638666they specialize in, but other competencies that are important to the industry. Therefo
14、re, wepropose that:Hypothesis 4: Responses from industry experts and the firm will agree on the presenceor absence of a competence at a firm.The presence of a core competence offers competitive advantage (Hamel Prahalad ICE, 1984, 1988, 1994;Mead, 1994; Noyce, 1980; Robinson, 1984) and semiconductor
15、 silicon (OMara, 1989;Runyan, 1965; Walsh, Boylan, Cook, Walsh, Boylan, Warrington, Walsh, Cook, also see Kim, Nie, and Verba (1977). In this study, the responses aredichotomous. However, the value that is factored is a value between 0 and 1. In the casesof companies that had 20 separate respondents
16、, there are 20 intervals. Whereas, in thecases of a captive company (one customer), the data are dichotomous. An examination ofcorrelations (Table 1) shows that there is a high correlation between variables. If allmissing values are considered to be zero, then the correlation between variables is at
17、 least.45 and as high as .83 (Table 1). If all cases with missing values are deleted, then thecorrelation between variables is at least .54 and as high as .86 (Table 1). Consequently, itis no great surprise that there is a single factor. A factor analysis of the technical com-petence scale was condu
18、cted for the 595 cases that had no missing data (Table 2) and for the732 cases, in which any missing data were considered to be the same as stating that nocompetence existed (Table 2). In both cases, there is only one factor. In all cases with missingdata deleted, the eigenvalue for the single facto
19、r is 3.88 and explains 77.6% of the variance.And in the case in which all missing values are considered to be equivalent to stating that thecompetence is absent, the single factor has an eigenvalue of 3.55 and explains 70.6% ofthe variance.There are no differences between factor analysis and theoret
20、ical expectations; thereliability and validity are now considered.6.2. ReliabilityThere are three methods of estimating reliability: testretest, internal consistency, andalternative forms (Peter, 1979). The survey participants are busy practitioners. Conse-quently, it is not practical to test reliab
21、ility either using testretest or alternative forms.Internal consistency is possible to calculate, however, from the data. Therefore, Crohn-bachs alpha is used to examine internal consistency of the technical competence scale(Table 3).The reliability of the technical competence scale is very high (Ta
22、ble 3). This isunsurprising since the eigenvalue of the factor and the correlations between its scale itemsare both high. In fact, one may argue that the scale reliability is too high. This issue will beaddressed in the future by using a 10-point Likert scale, instead of questions that require yesor
23、 no as an answer. (Recall, yes or no measures are used since some of the competenciesunder study were acquired and mastered decades ago.) Having considered reliability, validityis now considered.Table 3Crohnbachs alpha for technical competence measureCase Crohnbachs alphaif missing data is deleted (
24、n=599) .9201if missing data equals lack of competence (n=732) .8795S. Walsh, J.D. Linton / Journal of High Technology Management Research 13 (2002) 6386 716.3. Content validityContent validity refers to the representativeness or sampling adequacy of the content. Thequestions used to gauge technical
25、competence (Appendix A, Table B) are very similar toeach other in wording. The questions were asked to respondents as part of an interviewprocess. The interviewer clarified the meaning of the question and answered all queries therespondents had. The questions are consistent with previous work. Earli
26、er studies do notoperationalize technical competence, but they do examine the presence of technical strengthswithin an organization. Fusfeld (1978) proposes a Technological Audit to identify syner-gies throughout an organization. Technology subsets are described as being distinctivetechnologies or s
27、trategic technological assets (Bitindo Frohman,1980). Rosenbloom (1978) categorizes technologies based on differences in their innova-tion processes. SEST-Euroconsult. (1984) suggests that the success of many Japanese firmsis based on their pursuit and competence with specific technological subsets.
28、 Havingdemonstrated that previous literature supports the content validity of the measure, otherevidence of content validity is considered.Internal consistency and correlation of scales with other measures of the constructprovide statistical evidence of content validity. Internal consistency has bee
29、n establishedand discussed in the previous section (see Table 3). The lack of other measures of theconstruct prevents using of correlations with similar instruments as a further test ofcontent validity. Content validity is supported through: (1) the consistency of wording ofall the questions, (2) co
30、nsistent with earlier works, and (3) because the scale isinternally consistent.6.4. Predictive validityPredictive validity is demonstrated by correlating a measure against other measures of thesame construct. Since this is the first operationalization of technical competence, devel-opment of alterna
31、tive measures was not feasible at the time of survey development.Predictive validity needs to be pursued in a future study.6.5. Construct validityConstruct validity is demonstrated by showing the validity of the theory behind theinstrument (Kerlinger, 1986). In order to do this, the theory has to de
32、monstrate consistentfindings over a number of studies. Two challenges to demonstrating construct validity areoperationalization being limited to case studies and different definitions for competence andcore competence.As stated earlier, competency theory has been discussed in great detail, but opera
33、tion-alization has been limited to case studiesexamples include Prahalad and Hamel(1990)Honda, Quinn (1992)Walmart, Morone (1993)Corning, Hamel and Prahalad(1994)Sony. Consequently, more studies are required prior to the establishment ofconstruct validity for the technical competence measure. Howeve
34、r, of greater concern areS. Walsh, J.D. Linton / Journal of High Technology Management Research 13 (2002) 638672substantive differences in definition. Recall, Hamel and Prahalad (1994) describe a corecompetence from the perspective of technology and production skills and then quitecorrectly describe
35、 Sony as having a core competence in miniturization. The simplicityof the term miniturization is engaging and is accurate at the time of writing. However, itis incorrect several years later. Sony does have a competence of miniturization oftraditional mechanical systems. But technology is dynamic and
36、 leaders in miniturizationtoday, firms like Analog Devices and Texas Instruments, have competencies in micro-electromechanical systems an area in which Sony is involved, but not at the forefront(Nexus, 1998). And leaders in miniturization in the future are likely to be firms likeMotorola and IBM tha
37、t are active in the nanotechnology field (Rotman, 1999). Con-sequently, it is important to be very precise in identifying the basis of a competence orcore competence. It is quite likely that subtle, but evidently important, differences indefinition have hindered investigations involving operationali
38、zation and the establishmentof instrument validity.Two methods, suggested by Kerlinger (1986), for examining construct validity, arefactor analysis and correlation between item scores and total scores. Factor analysissupports the construct validity. Correlations between item scores and total scores
39、wereconducted after removing the item score from the total score (Cohen Chris-tensen, 1997). Consequently, it is suggested that bundles of competencies and/or core com-petencies be described so that the knowledge base that the competence represents is clearlystated. Having considered the implication
40、s of this work to the identification of core com-petencies, we consider whether the SIC code assists in understanding the presence orabsence of competence.6.7. SIC codes as an indicator of competenceWe attempted using SIC codes as a proxy for identification of competencies. A firm wasgiven a value o
41、f 1 for an SIC code, if products related to the technology were beingproduced. Otherwise, a value of zero was assigned. Correlations between .51 and .86(Table 1) suggest that SIC code offers a good indication with minimal effort. It appearsthat the SIC succeeds as an indicator for incremental or con
42、tinuous innovation (Morone,1994; Zaltman, Duncan, Christensen, 1997; Zaltman et al., 1973).S. Walsh, J.D. Linton / Journal of High Technology Management Research 13 (2002) 638674If competencies that rely on incremental innovation are under study, the SIC code issufficient. However, Leonard-Barton (1
43、995) warns of competency traps that result in firmsbeing blind to the value of new technologies (Christensen, 1997; Foster, 1986) thateventually overthrow the existing technical regime. Consequently, SIC codes should onlybe used as indicators of competence cautiously and may result in the mis-specif
44、ication ofcore competencies.6.8. Insights for future studies of competenceA tremendous amount of effort was required to conduct this analysis of the technicalcompetencies associated to this one industry. If such an effort is required to identify thetechnical competencies in an industry, it is unders
45、tandable why the discussion of competence,capability, and core competence tends to be theoretical and case study based. The experienceof collecting and analyzing this massive data set has given us some insights into simplerand easier methods of conducting a competence analysis in other industries. A
46、n invest-igation into the technical competencies that are of importance to firms and industries can bebroken down into two parts: identification of competencies and analysis of the presence orabsence of competencies. For the identification of technical competencies, two methods arerecommended. (1) A
47、n examination of industry forecasts, technical roadmaps, and tradepublications will identify many of the technical competencies associated with theindustry. (2) An examination of either the SOPs or a detailed consideration of the processof a few leading firms identifies additional technical competen
48、cies that would be missed ifthe literature alone is to be considered. Once the technical competencies have beenidentified, industry experts should be invited to identify additional competencies. (Thisstep may also result in the elimination of some competencies if they are redundant withother compete
49、ncies on the list.) Having identified the technical competencies, one mayeither conduct an analysis on all competencies or rate of competencies to eliminate onesof low importance. We will consider how to expedite the analysis of the competencies.The analysis of technical competencies in part depends on the purpose of the inves-tigation. For many, the interest will relate to questions like what technologies, if any, areimportant for competitive advantage? For questions such as this, the companies that areleaders in terms of such metrics as g
