組織管理研究中測量工具的跨文化可輸出性

行政院國家科學委員會專題研究計畫 成果報告

組織管理研究中測量工具的跨文化可輸出性

研究成果報告(精簡版)

中 華 民 國 98 年 12 月 31 日

行政院國家科學委員會補助專題研究計畫

行政院國家科學委員會補助專題研究計畫

行政院國家科學委員會補助專題研究計畫

行政院國家科學委員會補助專題研究計畫

成 果 報 告

成 果 報 告

成 果 報 告

成 果 報 告

□

□

□

□期中進度報告

期中進度報告

期中進度報告

期中進度報告

組織管理研究中測量工具的跨文化可輸出性

組織管理研究中測量工具的跨文化可輸出性

組織管理研究中測量工具的跨文化可輸出性

組織管理研究中測量工具的跨文化可輸出性 (I)

Exportability of Organizational Research Measures Across Cultures (I)

計畫類別：

個別型計畫 □ 整合型計畫

計畫編號：

NSC

97-2410-H-003-123-執行期間：97 年 8 月 1 日至 98 年 9 月 30 日

計畫主持人：葉俶禎

共同主持人：

計畫參與人員：

林朱燕、陳淑媛、葉宜菁、易俞廷

成果報告類型(依經費核定清單規定繳交)：

精簡報告 □完整報告

本成果報告包括以下應繳交之附件：

□赴國外出差或研習心得報告一份

□赴大陸地區出差或研習心得報告一份

出席國際學術會議心得報告及發表之論文各一份

□國際合作研究計畫國外研究報告書一份

處理方式：除產學合作研究計畫、提升產業技術及人才培育研究計畫、

列管計畫及下列情形者外，得立即公開查詢

□涉及專利或其他智慧財產權，□一年□二年後可公開查詢

執行單位：

國立臺灣師範大學

中 華 民 國 九十八 年 十二月 三十一日

National Science Council Founded Research

NSC 97-2410-H-003-123-

Exportability of Organizational Research Measures Across Cultures (I)

Final Report (Short)

C. Rosa Yeh

Assistant Professor

Graduate Institute of International Human Resource Development

National Taiwan Normal University

Abstract

It is common for organizational researchers to adopt existing measures in their research. However, the exportability of organizational research measures across borders may be limited, making the validity of study results questionable. The extent of this practice in Taiwan’s academic society will need to be surveyed to determine the severity of the situation. A full scale investigation was performed to understand the practice of scale development in Taiwan’s organizational research society and the extent of practice in adopting existing western measurement scales in Taiwan’s organizational research. The target of this

investigation was research articles published in major Taiwanese academic journals in the last five years that included latent constructs in an organizational research. Farh, Cannella, & Lee’s (2006) framework of scale development approaches in Chinese management research was used to categorize the development of organizational research constructs into four approaches -- translation, de-contextualization, adaptation, and contextualization.

This effort uncovered 146 organizational studies in Taiwan with 791 research constructs in these studies. Of the 791 constructs, 263 (33.2%) constructs were translated from an existing foreign scale, 352 (44.5%) constructs were measured using an adaptation of an existing foreign scale, 59 (7.5%) constructs used a measurement scale developed for the study in a de-contextualization manner, 36 (4.6%) constructs used a measurement scale developed specifically for the study in a contextualization manner. Research constructs that were measured with a direct translation or adaptation of a Western research instrument were tallied to uncover the most frequently adopted instruments by organizational researchers in Taiwan. As it turns out, organizational commitment, job satisfaction and performance were three most frequently researched constructs in Taiwan. Mowday, Steers, & Porter’s (1979) measurement on organizational commitment, and Michigan Organizational Assessment and Minnesota Satisfaction Questionnaire for job satisfaction were the most frequently cited instruments.

Introduction

The field of organizational research deals with many constructs that are latent in nature. That is, these constructs are abstract, implicit, and cannot be directly measured. Examples of frequently researched latent constructs include organizational climate, leadership,

organizational commitment, organizational learning, organizational citizenship behaviors, employee satisfaction, just to name a few. These latent constructs are important in both academic and practical sense in understanding the inner working of organizational

management and organizational behaviors. However, the development of valid measures or measurement instruments for these latent constructs has always been a challenge for

organizational management researchers.

To develop a valid measurement instrument for a latent construct is a daunting task. It takes a lot of knowledge and experience in both the theory behind the construct and the most current methodological savvy to pull it off, not to mention extra time and effort. Moreover, as social science methodology progresses, the bar is raised higher and higher for the academic society to accept a newly developed measure. As a result, most researchers are discouraged in developing on their own and are more inclined to adopt or adapt measurement scales that have been published before when taking on a research involving latent constructs.

This phenomenon is worsened when the researchers are working under the pressure of getting published. Many researchers rushed to undertake substantive research without any intent to devote time in crafting valid research instruments (Farh, Cannella, & Lee, 2006). In Taiwan, that pressure is not only real, it is also doubled because of an urgent need to publish in international journals, specifically SSCI ranked journals for social science researchers. To increase their chance of getting manuscripts accepted by the international journals,

organizational researchers in Taiwan often adopt or adapt those measurement scales that had appeared in internationally renowned journals, and most of which were originally developed

by scholars in the US or other western countries. This has become a common practice for researchers who intent to 1) devote valuable research time to tests of substantive hypotheses, 2) avoid being challenged in their validation process of a new research instruments, or 3) compare results with other studies using the same research construct.

This practice, however, has been under the attack by many research methodologists in many years because of the potential threats to measurement equivalence/invariance (ME/I) involved in this type of research (Cheung & Rensvold, 1999, 2002; Farh, Cannella, & Lee, 2006; Little, 1997; Mullen, 1995; Reise, Widaman, & Pugh, 1993; Singh, 1995; Steenkamp & Baumgartner, 1998; Vandenberg & Lance, 2000). Measurement equivalence/invariance is needed to ensure that items used in survey-type instruments mean the same things to members of different groups (Cheung, in press; Cheung & Rensvold, 2002). Researchers invoke

assumptions about measurement equivalence when they adopt or translate an existing measurement scale in their substantive research.

There are many sources of threats to ME/I in this type of research, for instance, culture difference, language/translation problem, organization influence, and response context (Robert, Lee, & Chan, 2006). Culture difference is a major source of threats to ME/I. It stems from the adoption of a measure for an emic construct without testing first whether the

measure was developed under an etic or an emic assumption. An emic assumption is culture-specific while an etic assumption is culturally neutral. Theoretically, “scientists interested in the local construction of meaning, and local rules for behavior, will rely on emic accounts”, while “scientists interested in facilitating comparative research and making

universal claims will rely on etic accounts” (Wikipedia, 2007). When social science researchers adopt measures that were developed and validated in the US or other western cultures and use them in their local studies without first examining whether the construct under investigation is culture-specific, or that ME/I is achieved, they do not know whether a

measurement scale actually measures the same construct they want to measure. Thus, the validity of the practice of adopting western research instruments in local organizational studies in Taiwan is seriously undermined.

Schoenfeldt (1984) stated “The construction of the measuring devices is perhaps the most important segment of any study. Many well-conceived research studies have never seen the light of day because of flawed measures.” Korman (1974) said “The point is not that adequate measurement is ‘nice’. It is necessary, crucial, etc. Without it we have nothing.” (p. 194) Nunnally & Bernstein (1994) believe that “validation is an unending process…Most psychological measures need to be constantly evaluated and reevaluated to see if they are behaving as they should.” (p. 84) Many researchers would agree that the issue of construct validation is by far the foremost important in the research planning process. However, as adopting existing measures has become a very common practice among researchers, tests of measurement equivalence/invariance are now a necessary step in ensuring construct validity.

Almost all the theories and data of contemporary management come from Western populations (e.g., Europeans, North Americans, Australians, etc.). Yet about 70% of humans live in non-Western cultures (Triandis, 1995). If management is to become a universal discipline it will need both theories and data from the majority of humans. This is especially true for the international business discipline as it deals with cross-cultural issues and concerns emerged from the management of humans with different cultural context. As Taiwan enters into the global village, more and more Taiwanese corporations are faced with decisions requiring in-depth understanding of the cultures involved. Many organizational research measures have been proven to be helpful in understanding organizational phenomenon. However, most measurement scales were developed and validated entirely in the western context and possibly with student samples. When these measures were adopted in research involving Taiwanese business sample without knowing the influence of culture, inference

from the study result may be invalid.

This research proposes a large scale study to survey the extent of adoption of western research instruments in organizational research in Taiwan. Investigation will focus on the scale development approach in major Taiwanese academic journals, specifically those listed in the TSSCI index, targeting latent constructs in organizational research. The survey will use Farh, Cannella, & Lee’s 2006 framework of scale development approaches in Chinese management research which categorizes the development of organizational research constructs into four approaches -- translation, de-contextualization, adaptation, and

contextualization -- according to two dimensions: 1) source of the scale, and 2) expectations about cultural specificity. Research constructs that were measured with a direct translation of a western-developed research instrument will be tallied to uncover the most frequently adopted instruments by organizational researchers in Taiwan. This research hopes to achieve the following two objectives: to understand the practice in scale development in Taiwan’s organizational research society; and to investigate the extent of practice in adopting western measurement scales in Taiwan’s organizational research.

Research Background

Farh, Cannella, & Lee (2006) classified survey scales or measures currently used in Chinese management research along two dimensions – the source of the scale and

expectations about its cultural specificity. Based on these two dimensions, they differentiated four approaches to scale development: translation, adaptation, de-contextualization, and contextualization, each having different key assumptions, strengths, limitations and roles in Chinese management research. See table 1 for the classification. “Translation” involves using a direct translation of an existing Western scale to create a Chinese language version of the scale. “Adaptation” also involves translating an existing scale, but with some modification to create a more meaningful Chinese version. “De-contextualization” involves assembling a

scale from scratch in a Chinese context, under the assumption that the target construct is etic (i.e., universal or culturally invariant). “Contextualization” involves developing a scale from scratch in a Chinese context, but under the assumption that the target construct is emic (i.e., relevant to China only).

Table 1 – Four Types of Scale Development Approaches in Chinese Management Research

Expectations about Cultural Specificity Source of the Scale

Etic Orientation Emic Orientation

Use or Modify an Existing Scale Translation Adaptation

Develop a New Scale De-contextualization Contextualization

Source: Farh, Cannella, & Lee (2006)

The translation approach adopts two key assumptions: (1) target construct is equivalent across cultures in terms of overall definition, content domain, and empirical

representations of the content domain; and (2) a high quality culturally unbiased western scale for the target construct is available. Most researchers use the direct translation approach to develop their scales because this approach not only saves development time and costs on a validation procedure required by a brand new scale, a properly translated scale also allows researchers to compare their findings directly with those from Western settings. However, with a direct translation approach, it is sometimes difficult to achieve semantic equivalence between the Chinese and Western scales. Culturally unbiased Western scales are hard to come by.

The adaptation approach assumes that target construct is equivalent between cultures in terms of overall definition and content domain, and that there are high quality Western scales for target construct available for adaptation. Strengths of this approach include low to moderate developmental time and costs and ease of scholarly exchanges of research findings with the Western literature. However, major drawbacks of an adaptation approach include

difficulty in conducting cross-cultural research and that drastic adaptation may create new scale that requires extensive validation in the Chinese context.

Key assumptions of the de-contextualization approach are (1) target construct is etic or universal or culturally invariant; (2) High quality scale for the target construct is unavailable in the literature. This approach provides opportunity to develop universal measure for target construct and ease of scholarly exchange of research findings with the Western literature. However, it takes much longer developmental time and high developmental costs. Also, items developed under this approach tend to be phrased at a more abstract level, which may limit its informational and practical value.

Contrary to the de-contextualization approach, the contextualization approach assumes target construct to be emic or culture specific, and that high quality emic scales for the target construct are unavailable in the literature. Contextualization provides opportunity to develop scales that are highly relevant for the Chinese context, and contribute context-specific knowledge to Chinese management. Similar to the de-contextualization approach, it takes longer developmental time and higher developmental costs. New scales developed under this approach have limited generalizability. Also, it is hard to communicate research finding with the Western literature.

Farh, Cannella, & Lee’s classification provides a very good framework from which to observe and categorize scale development approaches in Taiwan’s organizational research.

Research Method

This study selected 7 most important and representative management journals in Taiwan (including six TSSCI journals), namely, Asia Pacific Management Review, NTU Management Review, Sun Yat-Sen Management Review, Journal of Management, Chiao Da Management Review, Management Review, and Journal of Human Resource Management.

The purpose was to examine the characteristics of the organizational management

measurements used in the empirical studies of these journals, targeting specifically the fields of organization behavior and human resource (OB/HR). We examined the articles published in these six journals in the time period from 2004 to 2008. The criteria used to select papers are the following: (1) empirical studies, (2) organizational management studies, (3) studies that involve the use of a measurement scale to measure the research construct. All three criteria have to be met to be included in this review.

Table 2 – Number of Articles Reviewed from 2004 to 2008 by Journal

APMR 台大管 理論叢 中山管 理評論 管理學 報 交大管 理學報 管理評 論 人力資 源管理 學報 Total 2004 4 0 8 3 1 4 15 35 2005 1 4 1 6 0 1 16 29 2006 2 1 6 6 2 1 21 39 2007 3 3 6 0 1 1 13 27 2008 1 7 1 0 3 2 2 16 Total 11 15 22 15 7 9 67 146

As table 2 reveals, this study included 146 papers, with a range from16 to 39 papers each year. In these journals, Journal of Human Resource Management, with its specific focus on human resource management field, has the most number of papers, 67, being selected than any other journals in this study; while Chiao Da Management Review has the least, only 7, number of papers that match the criteria of this study.

After the sample was determined, a team of three graduate students went through a comprehensive review of each of these 146 articles using a coding table (see Appendix I) to record important information related to the purpose of this study. The coding table asked the research team for specific information such as basic reference data, field of study, study population and sample, instruments used, whether the instrument is a measurement scale,

whether the measurement scale is developed for the study or adopted from existing scale, source of the measurement scale if adopted, name of the measurement scale, constructs being measured in the scale, development/adoption process, validation process, and evidences of validity and reliability of each scale.

Once the comprehensive review was completed, the resulting 146 review tables along with their respective articles were sent to two doctoral students for cross-checking. All pieces of information that were coded incorrectly, incomplete, or in doubt were discussed and

reconciled in review meetings. To uncover the nature of the measurement scales used by these 146 studies, the research team then went through another coding effort targeting at each measurement scale used in these studies. Measurement scales were identified and attributes of each measurement scale were coded in an excel spreadsheet using pre-determined coding scheme as shown in Appendix II. The coding task was divided among the two doctoral students who also cross checked the other person’s coding for accuracy. Again, review meetings were used to reconcile any differences in judgment.

Attributes of measurement scales that used a numerical coding scheme were research field, study sample, scale development approach, the use of back-translation, cited reason for scale modification, evidence of validity, and evidence of reliability. Research field included the fields of OB, HR, and others. Study sample included Taiwan, Taiwan plus others, and others. Scale development approach combined Farh, et al.’s (2006) framework of translation, adaptation, de-contextualization, and contextualization, and three other categories uncovered in the course of this study, the adoption of a Taiwanese scale, the adaptation of a Taiwanese scale, and not specified. The use of backward translation was coded as yes, no and not applicable. Cited reason for scale modification included cultural differences,

background/context differences, population differences, not stated, and not applicable. Evidence of validity included seven categories, exploratory factor analysis, confirmatory

factor analysis, content validity, criterion-referenced validity, convergent validity,

discriminant validity, and others, each was dummy coded as yes or no. Evidence of reliability included Crobach’s alpha and others; also dummy coded as yes or no.

Findings and Discussions

Findings and Discussions are divided into two sections. The first section reports the results of our analysis on the most frequently studied constructs and the measurement scales used to measure these constructs in organizational management studies by Taiwanese researchers. The second section provides a detail analysis of the attributes of measurement scales used in Taiwan’s organizational management studies.

Popular Research Constructs and Their Measures

According to our data analysis result, we tabulated several most frequently used measures as shown in table 3, 4 and 5. The result indicates that measure of organizational commitment was the most frequently used in OB/HR studies. As we can see in table 3, 12 out of 42 studies adopted Porter, Steers, & Mowday’s (1974) organizational commitment

questionnaires in their studies. Ten studies adopted Mowday, Steers, & Porter’s (1979) version. Other organizational commitment questionnaires, such as Allen and Meyer (1990) (5 out of 42 studies), Meyer, Allen and Smith (1993) (6 out of 42 studies) were also quite often used in OB/HR studies. From the citation sources, we found that some studies adopted the

organizational commitment scales Mandarin version translated by Taiwanese scholars. For examples, Chang (1995), Huang and Tsai (1998), Chung (2000), and Yang (2002) translated the organizational commitment scale from Porter et al.’s (1974) original one.

Table 3 – Measures of Organizational Commitment

* adopted questionnaires developed (translated) from studies by Taiwan scholars.

1

count for Porter, Steers & Mowday’s (1974) citations.

Measures of job satisfaction and performance were quite frequently used in OB/HR studies as well. For job satisfaction measures, the result in table 4 indicated that Minnesota

Satisfaction Questionnaire (MSQ) developed by Weiss, Dawis, England & Lofquist (1967)

was the most frequently used (7 out of 36 studies). Another equally frequent measures used in job satisfaction studies was job satisfaction subscale of Michigan Organizational Assessment Questionnaire (7 out of 36 studies) developed by Cammann, Fichman, Jenkins, & Klesh (1979). We also found that some studies adopted job satisfaction scale that were developed by Taiwan scholars such as Huang (2002), Yen (2004); Lai (2005), Yang (2002), Lin (2000), Hsu (2001), Liao (1982), Yu (2001), and Jing (2002).

Measures Source Count

Mowday, Steers, & Porter (1979) 10 Porter, Steers, & Mowday (1974)1 8

Allen and Meyer (1990) 5

Meyer, Allen, and Smith (1993) 6

Huang and Tsai (1998),* Porter, Steers & Mowday (1974)1 3

Mowday, Porter, & Steer (1982) 2

Allen and Meyer (1996) 1

Aranya, Kushnir, & Valency (1986), Blau (1999); Chuang (2002)* 1

Blau (1985) 1

Gu (2003); Kuo, & Tsaur (2004); Vredenburgh (2002) 1 Meyer and Allen (1984); Fukami and Larson (1984) 1

Morgan and Hunt (1994) 1

Porter and Smith (1976) 1

Organizational Commitment

Porter, Steers, & Mowday (1974)1; Yang(2002)* ;Chung (2000)*;

Chang(1995)* 1

Table 4 – Measures of Job Satisfaction

Measurement Source Count

Minnesota Satisfaction Questionnaire (MSQ), Weiss, Dawis,

England & Lofquist (1967) 6

Job Satisfaction subscale of Michigan Organizational assessment Questionnaire , Cammann, Fichman, Jenkins, & Klesh (1979); Seashore and Lawler (1975)

4

Job Satisfaction subscale of Michigan Organizational assessment

Questionnaire , Cammann, Fichman, Jenkins, & Klesh (1979) 3

Etezadi-Amoli and Farhoomand 2

Graham (1985) 1

Cammann, Fichamn, Jenkins, & Klesh (1983) 1

Fullerton (2005) 1

Ganzach, (1998) 1

Gregson (1990) 1

Hackman and Oldham(1975) 1

Hackman and Oldham (1975); Churchill, Ford, & Walker (1974) 1

Heneman and Schwab(1985) 1

Masterson, Lewis, Goldman, & Taylor (2000) 1

Mohr and Spekman (1994); Daugherty, Myers, & Richey (2002) 1 Nerkar, McGrath, & MacMillian（1996） 1 Rich (1997); Hackman and Oldham (1975) 1 Seashore, Lawler, Mirvis & Cammann (1982) 1 Slavitt, Stamps, Piedomte, & Hass (1978) 1

Smith, Kendall & Hulin (1969) 1

Spector (1985) 1

Wexley and Yukl (1977) 1

Wang, Yang, & Wang (2001)*; Minnesota Satisfaction

Questionnaire (MSQ), Weiss, Dawis, England & Lofquist (1967) 1

Huang(2002)*; Yen(2004)*; Lai(2005)* 1

Yang(2002); Lin(2000)*; Hsu(2001)* 1

Job satisfaction

Liao(1982)*; Yu(2001)*; Jing(2002)* 1

Total 36

* adopted questionnaires developed (translated) from studies by Taiwan scholars.

The measures of job performance are listed in table 5. These studies adopted scales that measure performance at different level of analyses such as team, individual job and

organizational performance. The result in table 5 implied that most studies (32 studies) still focus on individual job performance. Kostova’s (1999), Zahra and George (2002), Kao and Huang’s (2000) scales were the most frequently used.

Table 5 – Measures of Job Performance

Measurement Source Count

Team

performance Robbins (1990) 1

Kostova (1999) ; Zahra and George (2002) 4

Kao and Huang(2000)* 4

Kirkpatrick (1994), Parry(1997) 2

Pulakos, Arad, Donovan, & Plamon (2000) 2

Fernando and Marshall (2004) 2

Utterback (1994); Souder and Song (1997) 2 Harrigan (1988); Lewis (1990); Geringer and Hebert (1991) 2 Robinson, Marshall, Moncrief and Lassk (2002); Sujan, Bettman, &

Baumgartner (1993) 2

Celly, Spekman, & Kamauff (1999); Hackman and Lawler (1971);

O'Donnell (2000) 2

Bucklin and Sengupta (1993); Reukert and Walker (1987) 1 Motowidlo and Van Scotter (1994); Yu(1996) 1

Fukami and Hopking (1993) 1

Farh, Dobbins, & Cheng (1991) 1

Singh (1995) 1

Yousef (1991); Utterback (1994) 1

Wang(1999)*; Daft(1978); Tsai(1997)* 1

Motowidlo and Van Scotter (1994) 1

Goodman and Svyantek (1999); Williams and Anderson (1991) 1 Individual job

performance

Allworth and Hesketh (1999) 1

Subtotal 32

Buckley (2000); Burn, Marshall, & Barnett (2002) 1 Cran(1994); Homburg , Hoyer, & Fassnacht (2002); Keillor, Parker,

& Pettijohn (1999); Gonzalez and Garazo(2006); Liao and Chuang (2004); Locke (1976); Lytle and Timmerman (2006)

1

Govindarajan (1984) 1

Organizational performance

Huselid and Jackson (1997); Becker, Poreda, Hunt, Bunch, &

Kaplan and Norton (2004) 1

Chou (2004)* 1

Motowidlo and Van Scotter (1994) 1

Borman and Motowidio (1993) 1

Delaney and Huselid (1996) 1

Subtotal 9

Total 43

* adopted questionnaires developed (translated) from studies by Taiwan scholars.

Attributes of Measurement Scales

As shown in table 6, our study uncovered 791 research constructs in the selected sample, of which 547 (69.2%) can be categorized as OB constructs while 179 (22.6%) as HR constructs. There were 65 (8.2%) constructs that were multi-disciplinary combining OB/HR constructs with constructs from other research fields such as marketing, operation, etc. It is obvious that OB remains the biggest and most interested research field among Taiwan's organizational management scholars.

Table 6 – Number of Constructs Reviewed by Management Field

Research_field Research_fieldResearch_field Research_field 547 69.2 69.2 69.2 179 22.6 22.6 91.8 65 8.2 8.2 100.0 791 100.0 100.0 OB HR Others Total Valid

Frequency Percent Valid Percent CumulativePercent

The samples of the studies of the papers we selected included both Taiwanese and subjects from other regions. As the analysis result in table 7 shows, most constructs in the studies used samples from Taiwan only, which is 725 (91.7%), while 14 (1.8%) constructs used samples from both Taiwan and other regions. A very small number of constructs, 52 (6.6%), collected data from samples that resided in regions outside of Taiwan, such as Hong Kong, Mainland China, etc. Therefore, we can conclude that the OB/HR researchers in Taiwan conduct their research mainly on Taiwanese samples.

Table 7 – Samples Studied by Taiwanese Organizational Management Researchers 樣本 樣本 樣本 樣本 725 91.7 91.7 91.7 14 1.8 1.8 93.4 52 6.6 6.6 100.0 791 100.0 100.0 Taiwan only Taiwan+others other countries Total Valid

Frequency Percent Valid Percent CumulativePercent

We classified the study measures into 7 categories: translation, adaptation,

de-contextualization, contextualization, the adoption of a Taiwanese scale, the adaptation of a Taiwanese scale, and not specified. The frequency analysis result in table 8 shows that of the 791 constructs, 263 (33.2%) constructs were translated from an existing foreign scale, 352 (44.5%) constructs were measured using an adaptation of an existing foreign scale, 59 (7.5%) constructs used a measurement scale developed for the study in a de-contextualization manner, 36 (4.6%) constructs used a measurement scale developed specifically for the study in a

contextualization manner, 46 constructs (5.8%) were measured using an existing Taiwanese scale, 31 constructs (3.9%) were measured using an adaptation of an existing Taiwanese scale, and the source of measurement scale of 4 (.5%) constructs were left unspecified. The above analysis shows that the majority of the measurement scales, a total of 615 (77.7%), used by Taiwanese organizational management researchers in our sample were in the categories of translation and adaptation. Both involve the use of an existing foreign scale.

Table 8 – Scale Development Approach

量表類別 量表類別 量表類別 量表類別 263 33.2 33.2 33.2 352 44.5 44.5 77.7 59 7.5 7.5 85.2 36 4.6 4.6 89.8 46 5.8 5.8 95.6 31 3.9 3.9 99.5 4 .5 .5 100.0 791 100.0 100.0 Translation Adaptation Decontextualization contextualization 引用本土量表 修改本土量表 未說明 Total

Valid Frequency Percent Valid Percent

Cumulative Percent

Of the constructs that were a direct translation of or an adaptation from an existing foreign measure, only 41 constructs (5.2%) were reported using the backward translation procedure, while 567 constructs (71.7%) were not reported using this procedure to make sure the translation of a foreign scale into Chinese was accurate (see table 9). It seems most

Taiwanese researchers still do not take the backward translation procedure as a necessary step to ensure the validity of their translated measures.

Table 9 – The Use of Backward Translation

翻譯過程是否述明有 翻譯過程是否述明有 翻譯過程是否述明有

翻譯過程是否述明有backward_translationbackward_translationbackward_translationbackward_translation

41 5.2 5.2 5.2 567 71.7 71.7 76.9 183 23.1 23.1 100.0 791 100.0 100.0 yes no no need translation Total Valid

Frequency Percent Valid Percent CumulativePercent

Of the 557 measures that were modified from an existing scale, 8 measures (1.0%) were modified because of cultural differences, 61 measures (7.7%) were modified because of the differences in the study context, 96 measures (12.1%) were modified because of the

differences in population characteristics, while 392 (49.6%) measures did not specify a reason for modification (see table 10). Most Taiwanese researchers left the reasons for modification of an existing measure unstated in their papers.

Table 10 – Reason for Modifying from Existing Scales

量表修正原因 量表修正原因量表修正原因 量表修正原因 8 1.0 1.0 1.0 61 7.7 7.7 8.7 96 12.1 12.1 20.9 392 49.6 49.6 70.4 234 29.6 29.6 100.0 791 100.0 100.0 文化差異 研究情境不同 研究對象不同 未說明 不須修正 Total Valid

Frequency Percent Valid Percent CumulativePercent

management research reviewed. In regards to the validity indicators of measures, of the 791 constructs, 330 constructs reported using Exploratory Factor Analysis (EFA) as the validity indicator, 231 constructs reported using Confirmatory Factor Analysis (CFA) to ensure the validity, 71 constructs reported content validity analysis, 11 constructs reported

criterion-referenced validity, 140 constructs reported convergent validity, 144 constructs reported discriminant validity, 39 constructs reported other validity indicators. Most of the constructs that reported validity used multiple validity indicators, for example, convergent validity usually goes with discriminant validity. However, there were still 193 constructs studied without reporting any evidence of validity.

As for the reliability indicators of measures, most of the measures, 699 out of the 791 constructs, reported the internal consistency using Cronbach’s alpha as their reliability indicator, while 19 constructs used other reliability indicators. There were 79 constructs that did not report any reliability analysis. The Cronbach’s alpha is the most frequently reported reliability indicator.

Table 11 – Reporting of Scale Validity and Reliability

Yes No

Exploratory factor analysis 330

(41.7%)

461 (58.3%) Confirmatory factor analysis 231

(29.2%) 560 (70.8%) Content validity 71 (9.0%) 720 (91.0%) Criterion-referenced validity 11 (1.4%) 780 (98.6%) Convergent validity 140 (17.7%) 651 (82.3%) Discriminant validity 144 (18.2%) 647 (81.8%) Other validity 39 (4.9%) 752 (95.1%) Report of validity 598 (75.6%) 193 (24.4%) Cronbach's alpha 699 92

(88.4%) (11.6%) Other reliability 19 (2.4%) 772 (97.6%) Report of reliability 712 (90.0%) 79 (10.0%)

Taiwan Organizational Management Scholars’ Research Orientation

After reviewing all articles, we tried to explain the reviewing result by redefining a new framework according to two dimensions: 1) research field (OB/HR) 2) level of analysis (individual/team and organizations) which are elaborated as follows.

Research field (OB/HR): By focusing on different research filed, the studies will adopt different constructs accordingly. HR studies were those that focused on the HR functions (e.g., selection and recruitment, training, retention), HR policy and practices, and organizational strategy, and adopted the related constructs in the studies accordingly. OB studies were those that placed emphasis on employees’ psychology (e.g., expectation, psychological contract) and behavioral outcome (e.g., organizational commitment, job satisfaction, performance, OCB) and included those constructs in the studies.

Level of analysis: The level of analysis can be defined as individual level which focused on individual employee’s behavior and outcomes (e.g., individual job performance, organizational commitment, job burnout). Team and organizational level focused on team or organizational outcomes (e.g., organizational performance, team cohesiveness).

Based on the above two dimensions, we developed a new framework to place those constructs reviewed in a typology.

Individual (1) Functional orientation (3) Behavioral orientation Level of

analysis

Team/Org (2) Strategic orientation (4) Psychological orientation

HR OB

Research field Figure 1 – Typology of Research Orientation

1) Functional orientation: Those constructs with functional orientation focus on HR regular functions (e.g., selection and recruitment, training, retention). For example,

organizational attractiveness, training effectiveness are functional oriented constructs. Eighty eight (or 11.1%) constructs reviewed in this study can be classified in this category.

2) Strategic orientation: The constructs with strategic orientation focus on human resource management practices and organizational level performance. HRM practices, HRM practice system, and organizational performance are constructs classified as having strategic orientation. Eighty three (or 10.5%) constructs reviewed in this study can be classified in this category.

3) Behavioral orientation: The behaviorally oriented constructs focus on individual behavior and outcomes. For example, organizational commitment, individual performance, and OCB are behaviorally oriented constructs. The majority of constructs reviewed in this study, 539 (or 68.1%), belong to this category.

4) Psychological orientation: Psychologically oriented constructs focus on team or organizational level psychological conditions. For example, organizational climate, culture, and perception of justice are classified as psychologically oriented constructs. Thirty (or 3.8%) constructs reviewed can be categorized as psychologically oriented.

Conclusions and Suggestions

This study attempted to understand the practice of scale development in Taiwan’s organizational research society; and to investigate the extent of practice in adopting western measurement scales in Taiwan’s organizational research. After reviewing 146 OB/HR studies published in Taiwan’s major journals for the past five year, we found that most of the published OB/HR studies in Taiwan adapted or modified from existing scales. According to our analysis, 615 out of 791 constructs studied in the 146 research adapted existing scales, including 263 constructs reported by the researchers that they used an existing translated version of a western scale in their studies. In addition, only 141 out of 791 constructs reported efforts of the development of a new scale used in their studies. The result implied scale development in OB/HR studies is not the mainstream. Most researches still choose to adapt existing scales instead. The result is consistent with the condition in the current academic society in Taiwan where the researchers are trying to increase their chance of getting manuscripts accepted by international journals by using an existing scale to avoid being challenged in their validation process of a new research instruments.

From our analysis, some problems underlying the current OB/HR studies in Taiwan were surfaced. The first problem is the tendency of OB/HR researchers in Taiwan to adapt mostly the existing scales developed in the west. Though they adapted scales that were well validated in previous studies, attempts to validate in current research context are lacking. It is obvious that the potential threats to measurement equivalence/invariance (ME/I) are not paid much attention to. Secondly, most OB/HR studies did not indicate clearly how they adapted the existing western developed scales. As the analysis indicated, only 263 out of 615 researched constructs reported using scale translation. Moreover, only 41 constructs clearly explained the translation process (i.e., back translation process) and others did not report how scales were translated, an indication that the researchers might have performed the scale

translation without checking the accuracy of translated scales. Thirdly, most researchers did some or minor revision on the adapted scales for their studies without explaining how or why they did the revision process. Based on our analysis result, at least 50% of studies did not explain any revision process although they reported they revised the scales for the study needs. Finally, though most studies (75.6%) reported the scale validity, less than half (41.7%) reported EFA analysis and about 30% reported CFA analysis. The analysis result indicates that most of the OB/HR studies did not perform further scale validation when they adapted western scales for their studies.

In sum, this study provides concrete evidence to the academic society in Taiwan that the current OB/HR studies did not pay sufficient attention to the potential threats of measurement equivalence/invariance (ME/I). It also shows that OB/HR scholars in Taiwan lacked the awareness in a solid scale validation process when they adapt existing western scales.

Based on the result of the above literature survey, we propose an investigation of ME/I of those most frequently adopted western measures to understand the extent of validity of past research results and to ensure validity of future applications. Multi-group comparison procedures in LISREL can be used to test the measurement equivalence/invariance of these instruments between Taiwanese samples and samples from the country of the scale’s origin such as the US.

Self-Evaluation of Study Results

This report covers the result of the first-year of research effort of a two-year study proposed to the NSC. As the NSC only approved a one-year study, the minimal funding seriously limited the scope of this study and only allowed the research team to conduct the full scale investigation of research articles published in major organizational research journals in Taiwan in the last five years.

As seen in this report, only quantitative and descriptive analyses were conducted to reveal the status quo of scale development and adaptation in Taiwan’s OB/HR research society. In the two-year proposal, this researcher had also proposed a qualitative approach to be followed by the quantitative analysis to provide a more in-depth understanding from the perspective of experienced Taiwanese scholars of the circumstances and the reasons for using existing western scales. A measurement equivalence/invariance test of some of the most popular western measurement scales was also proposed as the major research effort in the following year.

As indicated in the Conclusions and Suggestions section above, more efforts in scale validation are needed in Taiwan’s academic research. Validation of measurement scale is the first step to prove the validity of study result. Therefore, this research hopes to be able to extend from the first-year result to carry out subsequent research efforts that were originally proposed. Funding for these subsequent efforts will be tremendously helpful.

References

Cheung, G. W., & Rensvold, R. B. (1999). Testing factorial invariance across

groups:Areconceputalization and proposed new model. Journal of Management, 25(1), 1-27.

Cheung, G. W., & Rensvold, R. B. (2002). Evaluating Goodness-of-Fit Indexes for Testing Measurement Invariance. Structural Equation Modeling, 9(2), 233-255.

Farh, J. L., Cannella, A. A. J., & Lee, C. (2006). Approaches to scale development in Chinese management research. Management and Organization Review, 2(3), 1-18.

Little, T. D. (1997). Mean and covariance structures (MACS) analyses of cross-cultural data: Practical and theoretical issues. Multivariate Behavioral Research, 32(1), 53-76. Nunnally, J. C., & Bernstein, I. H. (Eds.). (1994). Psychometric theory (3rd ed.). New York:

McGraw-Hill.

Reise, S. P., Widaman, K. F., & Pugh, R. H. (1993). Confirmatory factor analysis and item response theory: Two approaches for exploring measurement invariance.

Psychological Bulletin, 114(3), 552-566.

Robert, C., Lee, W., & Chan, K. (2006). An empirical analysis of measurement equivalence with the indcol measure of individualism and collectivism: Implications for valid cross-cultural inference. Personnel Psychology, 59(1), 65-99.

Singh, J. (1995). Measurement issues in cross-national research. Journal of International

Business Studies, 26, 597-619.

Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement invariance in crossnational consumer research. Journal of Consumer Research, 25(1), 78-90. Triandis, H. C. (Ed.). (1995). Individualism and collectivism. San Francisco, CA: Westview

Press.

Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement

invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3(4-69).

Appendix I

Paper Review Coding Table

期刊名 期刊名期刊名 期刊名 文章資訊 文章資訊文章資訊 文章資訊 卷期 卷期卷期 卷期 年月 年月年月 年月 篇名 篇名篇名 篇名 作者 作者作者 作者 文章範疇 文章範疇文章範疇 文章範疇 (OB or (OB or(OB or (OB or HR HR HR HR)))) 研究對象 研究對象研究對象 研究對象 樣本 樣本樣本 樣本 (誰去填問卷，多少人填) (填的人都在台灣嗎?) 研究工具 研究工具研究工具 研究工具 量表使用與否 量表使用與否量表使用與否 量表使用與否 量表來源 量表來源量表來源 量表來源((((自行自行自行自行 設計或引用 設計或引用設計或引用 設計或引用)))) 自行設計 自行設計自行設計 自行設計 引用引用 引用引用 量表名稱 量表名稱量表名稱 量表名稱 量表名稱量表名稱 量表名稱量表名稱 量表概念 量表概念量表概念 量表概念 量表概念量表概念 量表概念量表概念 量表發展之過程 量表發展之過程量表發展之過程 量表發展之過程 有無修改內容有無修改內容，有無修改內容有無修改內容，，，修改之修改之修改之修改之 內容為何 內容為何內容為何 內容為何???? 1 11 1.... 題項增刪題項增刪題項增刪題項增刪(增哪些題 項，減哪些題項) 2 22 2.... 文字修辭文字修辭文字修辭 文字修辭 3 33 3.... 構面增刪構面增刪構面增刪 構面增刪 量表驗證之過程 量表驗證之過程量表驗證之過程 量表驗證之過程 **須註明是質化(內容效度)or 量 化(eg,信度分析，探索性因素分 析，驗證性因素分析) 修改之理由 修改之理由修改之理由 修改之理由 (研究情境不同 or 文化不同) 引用原文或翻譯之版 引用原文或翻譯之版引用原文或翻譯之版 引用原文或翻譯之版 本 本本 本，，，如為翻譯請詳述其，如為翻譯請詳述其如為翻譯請詳述其如為翻譯請詳述其 如何翻譯 如何翻譯如何翻譯 如何翻譯 修改後之量表有無重新 修改後之量表有無重新修改後之量表有無重新 修改後之量表有無重新 驗證 驗證驗證 驗證 量表信效度驗證之過程 量表信效度驗證之過程量表信效度驗證之過程 量表信效度驗證之過程

Appendix II

Measurement Scale Coding Scheme

Measurement Scale Attribute Coding

期刊 text

西元年 text

Authors text

Research field 1. OB, 2. HR, 3. Other

樣本 1. Taiwan only, 2. Taiwan + Other, 3. Other

Construct 名稱 text Measurement (量表) 名稱 text 量表別名 text 量表類別 1.Translation, 2.Adaptation, 3. De-contextualization, 4. Contextualization, 5. 引用本土量表, 6.修改本土量表, 7.未說明 引用來源 text 翻譯過程: 是否述明有 backward translation?

1.Yes, 2. No, 3. N/A (不需要做翻譯))

量表修正原因 1.文化差異, 2.研究情境不同, 3.研究對象不 同, 4.未說明, 5. N/A (不需做修正) 效度 1. 因素分析(EFA) 1. 有, 0. 無 效度 2. 因素分析(CFA) 1. 有, 0. 無 效度 3. 內容效度 1. 有, 0. 無 效度 4. 效標關聯效度 1. 有, 0. 無 效度 5. 收斂效度 1. 有, 0. 無 效度 6. 區辨效度 1. 有, 0. 無 效度 7. 其他效度 1. 有, 0. 無 信度 1. Cronbach's alpha 1. 有, 0. 無 信度 2. 其他 1. 有, 0. 無

出席國際學術會議心得報告

出席國際學術會議心得報告

出席國際學術會議心得報告

出席國際學術會議心得報告

會議時間地點 October 19, 2008-October 22, 2008; Las Vegas, Nevada, USA 會議名稱 IABE-2008 Las Vegas Annual Conference

發表論文題目 A validation of a collective human capital strength measurement.

一、 參加會議經過

因本人預計在暑假期間轉換教職，本計畫將因此變更執行學校，而原本想參加的研 討會如 AOM 與 AIB 都在暑期舉辦，為避免產生報銷的困擾，故選擇在 10 月份舉辦的研討 會，並由 96 年度與 97 年度國科會計畫國際會議差旅費項下核銷。此次參加的研討會是由 International Academy of Business and Economics (IABE) 在美國拉斯維加斯所舉辦之 2008 年會。IABE 自 2005 年起在拉斯維加斯一年舉辦一次會議，今年改成兩次，另一次是 6 月在瑞典的斯德哥爾摩舉辦，兩場年會的人數都超越以往的與會人數，算是愈來愈成功的 一個國際學會組織，今年拉斯維加斯這一個研討會會議期間為 10 月 19 日至 22 日共四天， 在舊市區的 Plaza Hotel 舉行，19 日下午為大會報到時間，期間安排二天半的論文發表時 間，共有 12 場 concurrent sessions，每一場為 1 小時 50 分鐘，最多同時有 10 個會議室， 各進行 5 至 6 篇論文的發表。總計約有 450 篇論文在此次研討會發表。發表人來自世界五 大洲，是相當國際化的學術研討會。 本人的論文排在 21 日的早上 8 點到 9 點 50 的人力資源管理場次。本人的場次由 University of North Carolina at Charlotte 的 Dr. A. Yesim Sireli 主持，共有 5 篇論文發表，第 1 篇是由來自美國 Montclair State University 與 University of Wisconsin-Green Bay 的學者所 發表的 An examination of the antecedents of motivation to lead，是一領導理論的實證研究，

探討哪些前置變數可以激勵員工擔當領導的責任。第 2 篇也是來自美國 Montclair State University 的 學 者 ， 發 表 Employees’ distributive justice perceptions and organizational citizenship behaviors: a social exchange perspective，以社會交換理論探討員工對分配正義的 感受與組織公民行為之間的關係，並。本人為第 3 篇，發表 A validation of a collective human capital strength measurement，是本人所發展的組織人力資本強度量表的效度檢測。第 4 篇 由 Session chair, Dr. A. Yesim Sireli 發表 Workforce attrition and other risks for the future of US nuclear industry: similarities with other large industries，由風險管理的角度探討美國核能 產業人員流失所可能 帶來的負面影響。第 5 篇則是由立陶宛 Kaunas University of Technology 來的學者，發表 Globalization and human resource development: theoretical approach，探討立陶宛在逐漸全球化的過程中應加強哪些人力資源的發展。

本人的論文，經過兩位匿名審查人的審查結果，獲得入選刊登於 IABE 學會的期刊 Journal of Academy of Business and Economics, Volume 8, Number 3。除了論文發表，此次 研討會並安排一次自助式晚宴，以及一次午宴及 keynote speech，提供與會者自由認識 交談機會。本人在晚宴上有幸與來自 California State University at Chico 及 University of Tennessee 的學者同桌，暢談美國大選與校園教學研究要求等議題。午宴上則與來自加拿 大、澳洲、波多黎各與美國學者同桌，於世界政治、經濟之依存關係並有一番意見交換與 言辭辯論。 二、 與會心得 本次研討會的發表論文數雖然很多，但感覺上每一場次的參與人數都不太多，留在 研討會旅館的學者也較少，雖與選擇的旅館地點與設施有關(見下文)，但主辦單位的設計 與主持人選應該都有影響。每一場次只安排有主持人與發表人，沒有評論人的設計，而且 主持人也是有該場次其中一位發表人來擔任，較沒有 credibility。本人這一場次的主持 人並非在正統的人力資源管理領域，無法對發表的數篇論文做任何評論，而且也沒有控制 好時間， 5 篇論文發表完便宣布解散，沒有保留發問與討論的時間，使論文發表意見交 流的美意大打折扣，我的論文幸好有同場發表的學者在會後給我一些意見，否則真有入寶 山而空手歸的感覺。如果由其他相同領域但不在同一場次的學者來擔任主持人與評論人，

並且在每一場次平均分配發表與評論時間，不但可增加每一場次的參與人數，並可以讓發 表人得到更多、更深入的意見，討論也會較熱絡。另外因為這個研討會涵蓋的主題包含經 濟、商業與管理，雖然發表的論文數很多，但真正與自己領域相關的場次並不多，會後社 交時間碰面的學者也較難遇到相同領域可以就所研究題目進行討論的，與其他我參加過的 研討會比起來，IABE 提供的學術研討的效能是較少的。

研討會在拉斯維加斯的舊市區旅館 The Plaza Hotel 舉辦，也是本人居住的旅館，雖 然較新市區的 strip hotels 便宜許多，但旅館已經很老舊，且進出房間及研討會場次房間都 會經過 Casino 區域，所以每天衣服都是濃濃的煙味，感覺很不好。美國國內學者較熟悉 狀況的多半選擇新市區的 strip hotels 來居住，但也造成交通問題，例如本人的發表場次， 因為前兩篇論文的發表人從新市區的 strip hotels 搭公車過來非常耗時而遲到，所以居住在 Plaza 的我便成為第一位發表人。另外也有參與發表以外其他場次會議的意願問題，使得 大部分場次都只有發表人及少數的聽眾在場。在旅遊勝地舉辦研討會雖然有其吸引力，不 過，明年如果還在同一個旅館舉辦，相信並不吸引人，應該換到 新市區的 strip hotel，讓 參與研討會的場地和 Casino 可以有效分隔，且增加與會者住在 conference hotel 的比例， 應可增加每場次參與的人數及意見交換的討論機會。 三、 建議 如果 IABE 繼續維持相同模式，則對於想要得到比較多相同領域學術交流討論與對自 己論文的意見的國內學者，本人並不建議參加這個研討會，尤其是人力資源領域的，財務 經濟方面的領域可能好一點。如果目的只是想在具有相當規模的國際研討會發表，並結識 不同國家、不同領域學者，那這個研討會仍是相當好的一個選擇。本人因經費拮据住在市 區 Plaza Hotel 是相較於新市區的 strip hotels 便宜，雖可減輕經濟負擔，但居住品質很不 好。建議還是選擇新市區的 strip hotels，同時可以觀摩拉斯維加斯博弈事業的管理發展模 式與進步的程度，另外拉斯維加斯較著名的 show 也都在 strip hotels，個人經濟許可的話 可以選幾個 show 來看，也是難得的機會。

會議期刊論文資料及紙本議程

AVALIDATION OF A COLLECTIVE HUMAN CAPITAL STRENGTH MEASUREMENT

Chu-chen Rosa Yeh, National Taiwan Normal University, Taipei, Taiwan, ROC

ABSTRACT

Human capital (HC) has become one of the most researched topics in the knowledge economy. Yeh (2006) in a review on the concept of human capital revealed several problems in the research of human capital at organizational level. To correct these problems, Yeh (2006) developed a new measurement of organizational HC which included three new sub-constructs (i.e., human capital-organization fit, complementarity of human capital, and specificity of human capital) from the resource-based perspective. This study attempted to extend Yeh’s 2006 study in order to cross-validate the newly developed measure of collective human capital. Data were collected from a larger sample of knowledge-based companies in Taiwan. All analyses were targeted at the organizational level. A total of 153 highest-level executives of Taiwan’s knowledge-intensive companies provided ratings of their collective human capital strength using this instrument. After exploratory factor analysis, measurement items were screened for congruence with the theorized factor structure. The remaining items form a second-order measurement model which included four first-order constructs (human capital-organization fit, complementarity of human capital, cohesion of human capital and specificity of human capital) under collective human capital strength. The model was entered into a confirmatory factor analysis using LISREL. After minor modification, the model achieved acceptable fit on major fit indices reported by LISREL.

出席國際學術會議心得報告

出席國際學術會議心得報告

出席國際學術會議心得報告

出席國際學術會議心得報告

會議名稱 14th Annual Meeting of Asia Pacific Region of Decision Sciences Institute (APDSI)

發表論文題目 Personal Response System: A Model-Based Case Study in Taiwan

一、參加會議經過

七月四日由台北出發，下午抵達上海浦東機場，由旅行社安排的接機服務同時有另 幾位乘客，其中一位是由臺灣到上海參加工作面試的年輕人，頗有冒險精神。到達旅館安 頓後坐捷運再換搭計程車到本次研討會會場，同時也是承辦學校的中歐工商管理學院 (China Europe International Business School, CEIBS)進行會前註冊及參加接待會 (reception)。中歐工商管理學院的校園與建築物很新，空間設計得不錯，感覺很舒服。 接待會原本要在校園裡舉行，但承辦單位與外匯廠商沒連絡好，廠商未出現，面對現場數 十位國際學者，承辦單位須緊急危機處理，等待約半小時後，承辦單位召來一部遊覽車與 數輛計程車並安排大家到附近一家國際飯店的 lunge bar 參與接待會。親見承辦人的危機 處理方式，也是ㄧ次新鮮的經驗。

研討會正式會議期間為 7 月 5 日至 7 日共二天半，共有 Service and Systems Management, Technology, Knowledge and Information Management I, Supply Management I, Supply Management II, Technology, Knowledge and Information Management II, Marketing and Demand Chain Management, Technology, Knowledge and Information Management II, Quantitative Methods, Marketing and Demand Chain Management, Manufacturing Systems & Processes, Supply Chain Management III, Management Issues, Technology, Knowledge and

session 為 1 小時 30 分鐘，各進行 3 至 4 篇論文的發表。

本人的論文安排在七月六日是早上 10:30 開始到中午 12 點的場次，由大陸在美國 University of Massachusetts at Dartmouth 的 Zhenghong Shi 主持，是四篇論文發表。第一 篇是台灣元智大學研究生 Hong-Wei Yeh 報告”Knowledge Contribution of Information Development Team: An Empirical Research from Social Cognitive Perspecive”。第二篇則是成 大企研所的博二學生 Wen-Yen Hsu 報告的”Governmence Mechanisms of Inter-organizational Knowledge Transfer – A Conceptual Framework”。第三篇由韓國的 Sim Hyehjung 報告 “A Research on CSF for Successful Implementation of SOA: from Perspective of Exploratory Research”。 本人的論文“Personal Response System: A Model-Based Case Study in Taiwan” 是第四篇。但很不幸的，可能有些水土不服，論文報告前晚就開始有蠻嚴重的腹瀉狀況， 當天早上起床後腹瀉狀況雖有減輕，但身體已無法負荷由旅館到中歐工商管理學院的長途 跋涉，最後改由也有出席本次研討會的共同作者代為報告。

二、與會心得

本年度國際會議差旅費所剩不多，正好這篇與經常合作的學者合寫的論文投上了今 年在上海舉行的 APDSI，我曾在 2001 年參加新加坡的 APDSI，感覺不錯，2003 年 APDSI 與它所屬的 DSI 在上海舉行國際年會，我也投上了一篇論文，但很可惜那一年的研討會因 SARS 臨時取消舉行，所以一直沒有機會到上海，這次終於可以見見上海風華。 今年因為 H1N1 也是全球大流行，APDSI 研討會又是在上海舉行，實在很怕重導覆轍， 加上 program 到很晚才出來，使我不敢先預訂機票旅館。可能很多學者有共同想法，所以 今年參加發表的論文不多，研討會五天的 program 排得很空，每天都只有半天的演講與發 表。最後 program 是研討會兩周前才出來，使得機票旅館的預定有些措手不及。最後圖交 通方便，訂了市區捷運站旁的旅館，不過離中歐工商管理學院最近的捷運站也還得搭計程 車約 20-30 分鐘才能到，實在不甚方便。莫怪研討會在每天論文發表場次結束後會安排提 供巴士到市中心區。 總計今年只有約 50 篇論文在此次研討會發表。雖然也有國際學者參與，但發表人以 台灣、大陸學者居多，與往年參加的非常國際化的大型 APDSI 學術研討會(二、三百篇論

文發表)相比差距頗大。可能是因為今年是由中歐工商管理學院的 Operational Management 教授承辦，感覺上主題訂得比較狹隘，所以參與的人數也較以往減少許多，例如往年都有 至少一個至二個 track 的主題是有關組織管理或人力資源管理，今年就完全沒有。不過中 歐工商管理學院(CEIBS)本身是相當的國際化，除了相當大比例的國際化師資外，職員與 學生也來自世界各地。加上中國大陸經濟的發展，MBA 學生及 EMBA 學生在取得學位後 的薪資大幅提升，促使中歐工商管理學院在全世界 MBA 課程排名大幅提升，這點是值得 台灣的學校參考的。不過這些國際師資的薪水應該都遠超過本地的一般水準，才能吸引來 自世界各地尤其是已發展國家的優秀師資到此任教。CEIBS 是國際知名的學校，能有機會 到此參觀，親見它的校園、設施、教授與學生，也是ㄧ大榮幸。 三、建議

Decision Science Institute 是國際知名的學會，其旗下的各區域性組織以及所舉辦的研 討會皆有一定的品質與水準。本年度的 APDSI 研討會涵蓋主題較為狹隘，較不適合組織 與人力資源方面的學者參與，但如果可以回復往年的一些屬於組織管理的主題，增加不同 領域學者的參與，則還是不錯的一個研討會。另外在議程的訂定如能提早完成作業並通知 與會人士，將使旅程安排能更為順利。承辦人的主動積極回應問題也很重要，今年承辦的 教授可能沒有很多的奧援，自己疲於奔命不說，無法即時回答註冊發表人的問題，造成發 表人的許多不便，實需改善。 四、攜回資料名稱及內容 會議期刊論文資料及紙本議程 五、附件：發表論文

Personal Response System: A Model-Based Case Study in Taiwan

Yu-Hui Tao C. Rosa Yeh

National University of Kaohsiung National Taiwan Normal University [email protected] [email protected]

ABSTRACT

Personal response system (PRS) is increasingly adopted in Taiwan’s higher education. A small-scale case study was conducted to compare the perceptions of PRS usage of Taiwan’s college students on those issues reported by UK and USA literature, and at the same time to empirically test an integrated model that has not been tested in prior PRS studies. Results and implications of the study are summarized in this paper.

INTRODUCTION

Instant response system (IRS) includes the interactive white board (IWT) and personal response system (PRS), and is also called audience or student response system in the literature. There has been an increasing trend of PRS application in Taiwan’s higher education. However, to our knowledge, there is no related studies outside the continents of Europe, America, and Australia. This is a situation of concern since culture or region has been generally recognized as having a moderating effect in the technology acceptance model (TAM) and related theories. Meanwhile, only a few PRS studies touched upon theories in relation to education, and none on the technology-based theories for establishing students’ usage of such educational technology in classroom. We believe, in order to shorten the learning curve of PRS applications in Taiwan’s higher education, it is important to compare how Taiwan’s college students perceive PRS on those issues that had been reported in UK and USA so that more effective implementation strategies or practices can be adopted. A small-scale study was conducted with over fifty students registered in a course in a university in southern Taiwan. Due to page limitation, the literature review is omitted to accommodate research design, data analysis and conclusions in the remaining paper.

RESEARCH DESIGN

The research model is an integration of technology acceptance model (TAM) and expectation confirmation theory (ECT) since they are supplementary to each other as implied by Premkumar and Bhattacherjee (2008). The question is how these two models can be smoothly integrated? One drawback of existing TAM studies as summarized by Benbasat and Barki’s (2007) is that it only stresses on the frequency and the number of times of the usage, which is deficient and leads the TAM researchers to neglect other important user behaviors (Nambisan et al., 1999). Therefore, this research used learning performance in place of the usage intentions and actual usage of TAM, which not only was more meaningful in practice, but also provided a fine link from TAM to ECT. The research model can be seen in Figure 1 in the later section.

the eighteen-week course design, the instructor taught the first five chapters of the textbook “System Analysis and Design in a Changing World, 4th Edition” by Stazinger, Jackson and Burd, and 10 groups of students prepared and presented the remaining 10 chapters in the following 10 weeks. In each of the three-hour session, PRS was used to do the roll call at least twice and to perform at least three Q&A’s regarding the materials. The students first experienced how PRS was used in the first five meetings taught by the instructor before designing their own PRS application in their presentation. Fifty-three students registered in this course, but one dropped out after the first mid-term exam. Among the remaining fifty-two students, forty-seven students filled out the online questionnaire at the end of this course.

The questionnaire was divided into two sections. The first section collected data on questionnaire items from existing PRS literature and the second section contained items for measuring the TAM and ECT constructs adopted from the literature. Students were asked to respond to each questionnaire item (except item 22) using a 5-point Likert scale with 1 representing “extremely disagree” and 5 representing “extremely agree”.

ANALYSIS AND DISCUSSION

Descriptive statistics is used to analyze the questionnaire items in the first section, and simple regression is used to analyze the measurement items for the research model. Regression is used because of the small sample size of 47 in this study.

Table 1 enumerates the questions with supporting references and average scores calculated from the 47 observations. To give meaning to this survey result, we summarized our interpretations into nine points which are listed below.

Table 1 Questionnaire items, references and average scores

# Item & references Score

1 I prefer the Harvard style that students preview class materials and answer some questions at

the beginning of each class meeting, and then the instructor adjusts teaching outline and materials accordingly. In the class, PRS is still used to guide students to answer quick questions and discussions for clarifying some misconceptions and perceptions. (Anonymous, 2008)

3.66

2 I prefer not to preview class materials before each class meeting, but to use PRS during the class for identifying students’ understandings for reinforcing certain topics.

(http://www.habook.com.tw/default.as)

3.88

3 I prefer to preview class materials before each class meeting, and use PRS during the class for

dynamically adjusting class materials and schedule. (http://www.habook.com.tw/default.as)

4.02 4 I prefer that PRS is not used to monitor student attendance and cumulate points for

participation. (D’Arcy et al., 2007)

3.49

5 Using PRS increases students’ class attendance. (Briggs, 2006) 3.72

6 Using PRS increases students’ class concentration. (Murphy, 2008) 3.68

7 Using PRS helps the instructor quickly get the quick answers from students. (Skiba, 2006) 3.7

8 Using PRS helps the instructor discover students’ misconceptions. (Linsenmerier, et al., 2006) 3.13