CENTRAL TENDENCY IN ANSWERING QUESTIONNAIRE FOR HONG KONG CHINESE SHUM SUK-CHONG MBA PROJECT REPORT. Presented to. The Graduate School

CENTRAL TENDENCY IN ANSWERING QUESTIONNAIRE FOR HONG KONG CHINESE by SHUM SUK-CHONG ~ }ff;>t #1:. MBA PROJECT REPORT Presented to The Graduate School In Partial Fulfillment of the Requirements for the Degree of MASTER OF BUSINESS ADMINISTRATION TWO-YEAR MBA PROGRAMME THE CHINESE UNIVERSITY OF HONG KONG May 1992 Dr WONG CHI SUM Advisor

~ L 348111 ta e s'5 H 6/, ~1 S~3

ii ABSTRACT Some researchers believed that reluctance among Chinese people in expressing opinions would affect the validity of data collected using odd-numbered response scale of structured questionnaire. If this cultural hesitation indeed exists, they may be more likely to respond to the mid point of an odd-numbered scale. This study tested this speculation by a variety of measures in a sample of recent graduates from the Chinese University of Hong Kong. Results of this study did not support this speculation.

ACKNOWLEDGEMENTS The author wishes to express his gratitude to her advisor, Dr. C S Wong, for his generous support and guidance. The author would also like to specially acknowledge Dr. C S Wong's assistance in the use of statistical packages and in the pre-test and distribution of questionnaires.

TABLE OF CONTENTS iv ABSTRACT............. ii ACKNOWLEDGEMENTS..... iii TABLE OF CONTENTS iv LIST OF FIGURES................... v LIST OF TABLES............ vi CHAPTER I INTRODUCTION....... 1 Statement of Problem........ 2 Focus of Research............. 3 Job Characteristics Model..... 4 Number of Response Categories... 6 Response Sets of the Chinese........ 7 Importance of this Study.......... 9 Development of the Research Hypotheses....... 10 Research Hypothesis in the Null Form... 10 11 III IV RESEARCH METHODOLOGY.. Design of Questionnaire Layout of Section I.... Layout of Section 11...... Layout of Section III...... Pre-testing and Adjustments...... Selection of Respondents........ Estimates for the Variables.... Transforming the Scale......... Response Rate....... RESEARCH FINDINGS................. Research Hypotheses...... Research Hypothesis in the Null Form... Reliability Coefficients...... Means and Standard Deviations... T-tests on Means and F-tests on Standard Deviations Correlation Among Factors... Conclusion.......... CONCLUS IONS.................. Response Scale........ Validity of Job Characteristics Model in Hong Kong. Characteristics of University Graduates in Hong Kong Limitations of Study............ Implications for Future Study......... 12 12 13 14 15 16 16 17 18 19 21 21 22 22 24 24 25 26 27 27 28 29 29 31 APPENDIX............. 33 44 Questions used in Questionnaire REFERENCES............. 48

v LIST OF FIGURES Figure 1 JOB CHARACTERISTICS MODEL 33

vi LIST OF TABLES Table 1 Sample Characteristics...... 34 Table 2 Coefficient Alphas for All Measures 36 Table 3 Means and Standard Deviations. 37 Table 4 T-values for Means...... 38 Table 5 F-values for Standard Deviations 39 Table 6 Correlation and Significance - 4-Point Scale 40 Table 7 Correlation and Significance - 5-Point Scale 42

1 CHAPTER I INTRODUCTION The American Marketing Association has defined marketing research as the "systematic gathering, recording, and analyzing of data about problems relating to the marketing of goods and services" (Churchill 1987, p.10). This definition suggests that research information is not intuitive because it requires a scientific and systematic approach. If the data or information gathered, recorded, and analyzed is to be accurate and unbiased, the researcher must be objective. Thus the design of the research must contain no bias. Objective business research can result in credible and valid information which can assist in making sound business decisions. This study will focus on investigating how a research design may generate bias or inaccurate information owing to the presence of cultural factors. In particular, this study will investigate the potential bias of central tendency created by the type of response scale for Hong Kong people when they are responding to structured questionnaire.

2 Statement of Problem It is well believed that Chinese people tend not to express extreme opinions (e.g., Yang 1982). If this cultural hesitation indeed exists, they may be more likely to respond to the mid point of an odd-numbered scale of the structured questionnaire. Thus in the case of an odd-numbered Likert-type scale, which is a popular means for measuring attitudes in a lot of business research, it may not be appropriate for the Chinese because information obtained may not truly reflect the respondents' opinions. With the Likert-type scale, respondents are asked to indicate their own attitudes by checking how strongly they agree or disagree with carefully constructed statement, ranging from very positive to very negative, toward the attitudinal object. For example, five response alternatives could be: Strongly disagree Disagree Uncertain Agree Strongly Agree 123 4 5 Usually the number of alternatives on a Likert-type scale may range from three to nine. If an odd-numbered scale with anchors ranging from strongly agree to

3 strongly disagree was used, Chinese respondents might tend to avoid choosing the strongly agree or disagree answers. This particular type of central tendency is commonly hypothesized as a specific cultural characteristic of the Chinese. Researchers in Hon~ Kong and Taiwan usually recommend using an even-numbered response scale when structured questionnaire was administered (e.g., Yang 1982). Although widely discussed by researchers, this characteristic of the Chinese in responding to odd-numbered scale has not been tested empirically. The purpose of this study is to investigate whether Chinese respondents will express their opinions differently when the scale is of even number anchors, i.e., without a mid point. Focus of Research In this research, the Job Characteristics Model (Hackman & Oldham 1976, 1980) will be used in designing the questionnaire. This model, as shown in Figure 1, was chosen because it has been widely researched in the past two decades. Items measuring the job characteristics are standardized and proven to be reliable. The relationship between the job characteristics and job satisfaction have been found not only in Western countries but also for the Chinese people. Many researches have shown that core job

4 characteristics would influence the workers' motivation and satisfaction. Job Characteristics Model The Job Characteristics Model traces motivation and satisfaction to psychological states experienced by individuals, and then links these states to the job characteristics. It hypothesizes that core job characteristics would significantly influence the behaviours and attitudes of workers. Hackman and Oldham (1976, 1980) identified five job dimensions or core job characteristics, which influence employees' attitudes and behaviours, as follows: 1. Skill variety - the extent to which the work requires workers to perform activities that challenge their skills or requires them to use diverse abilities ' to perform work; 2. Task identity - the extent to which the work requires workers to complete a "whole" or an identifiable piece of work or requires them to do a job from beginning to end with a visible outcome;

5 3. Task significance - the extent to which the work has a significant impact on lives or work of other people within or outside the organization; 4. Autonomy - the extent to which the job provides substantial freedom, independence, and discretion to the employees in scheduling the work and in determining the procedures to be used in carrying it out. 5. Feedback from the job itself - the extent to which information about the effectiveness of their efforts in performing the work is channelled to the workers. According to the Job Characteristics Model, these five core job dimensions will affect personal and work outcomes such as internal work motivation, quality of work performance, satisfaction with work, absenteeism and turnover. The Job Diagnostic Survey (JDS; Hackman and Oldham, 1975) is the standardized and widely used questionnaire to measure the five core job characteristics. It contains seven three-item scales to measure employees' perceptions of each job characteristics, including two additional characteristics of feedback from agents and dealing with others. Respondents are required to answer

6 questions in two sections. The first section asks respondents to mark on a seven point continuum the amount of each job characteristic they perceive to be present in their job. In the second section, responses are in terms of the accuracy of a number of statements about features of their job. A mean score is taken across the three items in each sub-scale, with a possible range of scores from 1 to 7, the latter indicating the substantial presence of a job characteristic. One item in each subscale is reverse-scored. The questionnaire is written in English with different types of response scale. To investigate our research question, the questionnaire was modified slightly so that a Likert-type response scale could be applied to all items. Two types of response scale are used. One is designed with a fourpoint scale of "strongly disagree", "disagree", "agree" and "strongly agree". The other is a five-point scale of "strongly disagree", "disagree", "neither disagree nor agree", "agree" and "strongly agree", that is, a mid point is provided. Number of Response Categories Studies by Bendig (1952a, 1952b, 1953, 1954a, 1954b) revealed that the most efficient number of response

7 categories for rating formats is from 3 to 5. Considering both scale and rater reliability, there is no gain in efficiency when the number of categories increases from 5 to 9. In fact, reliability drops with 3 categories or less and with 11 categories or more. Finn (1972) found out that reliability even dropped with less than 3 or more than 7 response categories. Lissitz and Green (1975), in a Monte Carlo study of the effect of response categories on scale reliability, concluded that there was little increase in reliability when there were more than 5 scale points or response categories. Finally, Jenkins and Taber (1977) in another Monte Carlo study of factors affecting scale reliability, agreed with Lissitz and Green (1975) that there is little utility in adding scale categories beyond five. They even discovered that excessive number of categories can have negative effects on scale reliability. In this study, we have two sets of questionnaire which are designed with either 4 or 5 response categories. This will fall within the efficient range of 3 to 5 response categories for rating formats. Response Sets of the Chinese Yang (1982) has indicated the potential problem of

8 external validity in cross-cultural investigations. When research designs' with high internal validity and tested/used in Western countries are applied to the Chinese, the results obtained may not be valid because of lack of external validity. The problem of generalizability exists when results or research designs widely used in Western countries are applied to the Chinese. Reasons include cultural specific response set of the Chinese. Yang (1982) argued that the Chinese would be influenced by (1) social desirability set - regardless of personal real situations, Chinese respondents may respond or answer questions according to what their society expects; (2) acquiescence set - regardless of personal real situations, Chinese respondents tend to answer "yes" or "agree" rather than "no" or "disagree"; (3) mediocracy set - regardless of personal real situations, Chinese respondents tend to respond in a medium level, i.e. avoid giving extreme opinions, such as "extremely agree" or "extremely disagree"; (4) response avoidance mind set - if possible, Chinese respondents tend not to respond or express personal opinions. As a result of these cultural specific response sets for the Chinese, the Western research scale may not be directly applicable to the Chinese. Chinese may tend to choose the mid-point of the research scale because of the

9 mind set of mediocracy. In addition, there may exist semantic difference in the meaning of "strongly agree" or "strongly disagree". Chinese may view the English word "strongly" as being too "strong" and not subtle enough to represent their feelings of "completely agree" or "completely disagree". As a result, in designing research instrument, we may consider omitting the midpoint to obtain more representative answers. Some studies discovered that in case of sensitive questions such as political research, Chinese will answer "I don't know" to avoid expressing their opinions (Hoadley, 1970; Podmore et al., 1975). Therefore, Yang (1982) suggests the use of an open-ended "others" to replace "I don't know", and indicates the importance of probing when Chinese answer "I don't know". Importance of this Study Structured questionnaire is a popular tool of soliciting opinions and attitudes in business studies. This research provides valuable insights into the application of Likert-type response scale in questionnaire design, especially for Chinese samples e If the widely believed central tendency bias do exist among

10 Hong Kong Chinese, researchers in conducting studies using questionnaires have to cater for this cultural factor and use even-number scale. However, if this bias is not empirically valid, then either even-number scale or odd-number scale can be administered to Chinese respondents. Development of the Research Hypotheses Based upon the arguments, findings and beliefs of different researchers (e.g., Hoadley, 1970; Podmore et ai, 1975; Yang, 1982), we hypothesize: The responses of the Hong Kong Chinese on questionnaire with odd-number scale would be different from those on even-number scale. Research Hypothesis in the Null Form Specifically, this hypothesis can be tested upon the followings: Hypothesis la: The mean scores of the job characteristics and job

satisfaction level between the 4-point scale set and the 5-point scale set would be different. 11 Hypothesis 1b: The standard deviations of the job characteristics and job satisfaction measures would be smaller for the 5- point scale set than the 4-point scale set. Hypothesis 1c: Due to the restriction in range (i.e., smaller standard deviations), the correlation coefficients among the job characteristics, individual characteristics and job satisfaction measures would be smaller for the 5-point scale set and the 4-point scale set. The remainder of this study is organized into three chapters. In Chapter II, it outlines the research methodology and presents the instrument used. Field work and other aspects are also discussed. Chapter III is devoted to the discussion of data analysis and interpretation. Also statistical tests are explained. Results of hypotheses testing and other relevant results are presented in greater details. Chapter IV contains evaluation of the findings, business research implications and conclusion drawn from this study.

12 CHAPTER 11 RESEARCH METHODOLOGY This chapter explains the research methodology used in this study. The questionnaire, sample, and the operation of the survey are explained. Design of Questionnaire The research instrument consists of a questionnaire which was divided into three sections. To study whether odd-numbered or even-numbered response scales make a difference in research results done by Hong Kong Chinese, we have designed a closed-ended structured questionnaire. One set provided a 4-point response scale while the other set with a 5-point scale. The questionnaire was designed to be a mail-back self-administered one. Instructions for completing the questions are provided at the beginning. The first section is designed to collect

13 respondent's opinions on his job nature. The second section asks for' their level of job satisfaction. The third section is used to collect personal information of respondents. Layout of Section I Section I can be further divided into two parts. Questions concerning job characteristics included 5 aspects, namely skill variety, task identity, task significance, autonomy, and feedback from the job itself. These are standard questions taken from Job Diagnostic Survey_ In addition, questions concerning respondent's individual characteristics are included, namely need for achievement (5 items from Edwards, 1959), intention to leave (3 items from Cammann, Fichman, Jenkins and Klesh, 1979) and self-perceived performance (5 items assessing the extent to which the respondents thought they were good performers) tailored for this study. A total of 26 questions are included in section I. All questions are standard questions well-tested in Western countries. For respondents receiving a questionnaire with 4- point response scale, they are asked to express their opinions of their job characteristics by circling one of the four answers:

1. Strongly disagree 14 2. Disagree 3. Agree 4. Strongly agree For respondents receiving a questionnaire with 5- point response scale, they can express opinions by choosing one of the five answers: 1. Strongly disagree 2 Disagree 3. Neither disagree or agree 4. Agree 5. Strongly agree Layout of Section 11 Section 11 is devoted to asking for job satisfaction items taken from the Minnesota Satisfaction Questionnaire (20 items from Weiss, Dawis, England & Lofquist, 1967: the short form of the Minnesota Satisfaction Questionnaire). A total of 20 questions are included. Questions are designed to ask how satisfied the respondent is concerning business, working alone, variety, status in community, the way superior handles others, competence of superior, conscience, stability,

15 working for others, giving instruction, using abilities, practice of company policies, pay and work load, promotion, judgement, practice of judgement, working environment, relationship with colleagues, praise and feeling of accomplishment. In this section, respondents are asked to express their satisfaction/dissatisfaction towards their jobs by circling one of the following four answers: 1. Very dissatisfied 2. Dissatisfied 3. Satisfied 4. Very satisfied For the set of questionnaire designed with a 5-point response scale, the answers available are: 1. Very dissatisfied 2. Dissatisfied 3. Neither dissatisfied nor satisfied 4. Satisfied 5. Very satisfied Layout of Section III

The third section consists of eight multiple choice 16 questions and two open-ended questions. This section collects demographic daia of respondents, including sex, age, years of service, monthly salary, number of job changes since graduation, absenteeism, graduation year from the Chinese University of Hong Kong, major field of study, present job title and name of their organization. Pre-testing and Adjustments During the initial phase of survey, a pretest of the questionnaire conducted with forty respondents who were conveniently chosen from the employees of the Chinese University of Hong Kong. Only minor adjustments of the wordings in some items were suggested by these respondents. Selection of Respondents The survey was conducted in Hong Kong in October 1991. The two sets of questionnaires (one with 4-point and the other 5-point scale) are sent out randomly to all graduates of the Chinese University of Hong Kong who graduated from 1986 to 1990. Respondents are required to

17 mail back the completed questionnaire anonymously. Estimates for the Variables 1. Core Job Characteristics For each question of the three-item sub-scale of the core job characteristics questions, the average score are calculated to yield the level of core job characteristics. 2. Motivating Potential Score An overall MPS score is used to represent the overall motivational job features. MPS was calculated by averaging skill variety, task identity, task significance, and multiplied by autonomy and feedback as indicated by Hackman and Oldham (1976). This serves as an estimate of the overall motivating potential of the job. Motivating Potential Score = (skill variety + task identity + task significance / 3 * Autonomy * Feedback from Job Itself 3. Individual Characteristics

18 For the three-item sub-scale of "intention to leave" and "need for achievement" and the five-item sub-scale of "self-perceived performance", again the average scores are calculated to yield the level of intention to leave, the need for achievement and self-perceived performance. 4. Job Satisfaction For the twenty-item sub-scale, the average score is calculated to yield the general level of job satisfaction. Transforming the Scale The average scores obtained from the 5-point response scale set are recorded and converted to scores in corresponding to a 4-point response scale line. That is, the averaged scale of "2" from the 5-point scale would be equivalent to a score of "1.25" in a 4-point scale. And a "3" of the 5-point scale would be equivalent to "2.5" of a 4-point scale. The corresponding scores of the converted 4-point scale are listed in next page:

19 5-point scale 1 2 3 4 5 Transformed 4-point scale 1 1.25 2.5 3.75 4 Then the mean scores and standard deviations obtained from the 4-point scale set are compared with the mean scores and standard deviations obtained from the transformed 4-point scale set. Paired t-tests and F tests are performed respectively for the means and standard deviations of the 4-point scale and transformed 4-point scale set to find out whether the differences in means and standard deviations are statistically significant or not. The Alpha value is set at 0.05 level. Response Rate A total of 2137 questionnaires were sent out with 825 returned. The response rate was 38%. Respondents were asked not to complete the questionnaire if they did not hold a full time job. Out of the 410 returned questionnaires from the 4-point set, four questionnaires were incomplete and were discarded, producing a total of

20 406 valid questionnaires. For the 5-point set, out of the 415 returned', 12 are discarded for incompleteness producing a total of 403 valid questionnaires. A total of 809 questionnaires are useful. The demographic variables of the respondents for the two sets of questionnaire were depicted in Table 1 - Sample Characteristics. Characteristics for the respondents from the two samples with one receiving the 5-point scale and the other receiving the 4-point scale were compared. As shown in Table 1, the sex distribution skewed towards male, 52.9% for the 5-point scale and 51.5% for the 4-point scale. Majority of the two samples lies within the age range of 23 to 28. Most of the respondents have 1 to 2 1/2 years of working experience with the company they are working at. Their pay levels are also similar. As a whole the two samples appeared to be reasonably representative to the whole population.

21 CHAPTER III RESEARCH FINDINGS In this chapter, the results of the survey are presented and t~e four hypotheses are tested accordingly. The statistical results are tabulated and the findings are explained. Research Hypotheses The hypothesis to be tested in this research is: Hypothesis: the responses of the Hong Kong Chinese on questionnaire with odd-number scale would be different from those on even-number scale.

22 Research Hypothesis in the Null Form The above hypothesis to be tested in this research are rewritten in the null format and summarized as follows: Hypothesis la: The mean scores of the job characteristics and job satisfaction level between the 4-point scale set and the 5-point scale set would be different. Hypothesis lb: The standard deviations of the job characteristics and job satisfaction measures would be smaller for the 5- point scale set than the 4-point scale set. Hypothesis lc: The correlation coefficients among the job characteristics, individual characteristics and job satisfaction measures would be smaller for the 5-point scale set and the 4-point scale set. Reliability Coefficients various statistics were calculated for the two

23 samples to assess the potential bias of the odd-number response scale.. The reliability coefficients (Alphas) were shown in Table 2 of the Appendices. The Alphas calculated for the 4-point scale set of questionnaires were reasonably high for the core job characteristics like Skill Variety (0.7317), Task Identity (0.8021), Task Significance (0.7880) and Autonomy (0.7062), except for Feedback from Job Itself (0.3292). The reliability coefficients for other variables, namely Intention to Leave (0.7951) and Need for Achievement (0.7358) were also reasonably high except for Perceived Performance (0.5991), which was at a marginal level. For job satisfaction level, the coefficient Alpha was high at 0.8852. This indicated each question was consistently testing for the same aspect as other questions. Similarly, the reliability coefficients obtained for the 5-point set were also reasonably high in Skill Variety (0.7091), Task Identity (0.8103}, Task Significance (0.7346) and Autonomy (0.77), except for Feedback from Job Itself (0.3910). The reliability coefficients for individual characteristics were high with Intention to Leave (0.8318) and Need for Achievement (0.6911) while Perceived Performance (0.5782) was a marginal level. For job satisfaction level (0.8646), the coefficient was also high. There appeared no systematic differences among the two samples with one receiving the 5-point scale questionnaire and the other receiving the

4-point scale questionnaire. 24 Means and Standard Deviations Table 3 presented the means and standard deviations for all the multiple item measures for the two samples. The raw data of the 5-point scale sample were transformed to a 4-point scale by recording 1 as 1, 2 as 1.75, 3 as 2.5, 4 and 3.25, and 5 as 4. The means obtained from 4- point scale set for each of the five core job characteristics ranged from 2.66 to 3 while standard deviations were within the range of 0.40 to 0.64. Similarly, the means and standard deviations for the transformed 4-point scale set ranged from 2.72 to 2.97 while for standard deviations, the range was from 0.47 to 0.62. T-tests on Means and F-tests on Standard Deviations With these results, paired t-tests and F-tests were calculated for each pair of means and standard deviations from the two response sets. The t-values for means were listed in Table 4. If the t-value is greater than 1.96 at p < 0.05, the difference between two means is deemed

25 as significant. The t-values calculated for each of mean score for the core job characteristics, individual characteristics and job satisfaction level were found to fall within 0.1098 to 0.1946. That means, no systematic differences could be observed for the two samples. As such, Hypothesis la was not supported. The mean scores of the job characteristics and job satisfaction level between the 4-point scale set and the 5-point scale set were found to have no significance difference. F-tests were performed on the standard deviations obtained for the two samples. The F-values were listed in Table 5. If the F-value is smaller than 0.76 or greater than 1.31 at p < 0.05, the two standard deviations are considered to have significant difference. The results indicated that no systematic differences could be found for the two samples. Hypothesis 1b was not proved. Correlation Among Factors The correlation coefficients for each of the job characteristics, MPS, individual characteristics, job satisfaction, sex, age, years of service, monthly salary, number of job changes, absenteeism, year graduated were

26 calculated. Table 6 and 7 provided the correlation coefficients and their level of statistical significance for the measures in the 'two samples. Results indicated that there were no systematic differences between the two samples. Conclusion Results of this study did not support the belief that Chinese people might be biased by odd-number response scale. The results that there are no systematic differences between the means, standard deviations and correlation coefficients reveal that Chinese people, especially for the Hong Kong Chinese, do not possess the mind set as described by previous researchers (e.g., Yang, 1982).

27 CHAPTER IV CONCLUSIONS We cannot discover central tendency in our sample of Hong Kong Chinese and as a result, we cannot conclude that even-number response scale is better. Response Scale From our study, we cannot discover central tendency in our sample. As a result, beliefs about central tendency and non-expression mentality of Chinese people cannot be proved. Therefore, we cannot conclude that even-number response scale is better in collecting information from Chinese. Neither can we say that oddnumber response scale used in research instrument is unsuitable for soliciting opinions from the Chinese people. However, it should be noted that this study is conducted on a specific group of Chinese, University

28 graduate, who are considered to be highly educated. In generalizing thi's finding to other samples and measures, we must proceed with cautions. For instance, it may be too early to say that Chinese people, especially for Hong Kong Chinese, are no longer reluctant to express their political opinions on sensitive issues. Also, some other topics such as immigration may still be considered as sensitive. Validity of Job Characteristics Model in Hong Kong Results of this study further confirm that the validity of Job Characteristics Model in Hong Kong, which relates the core job dimensions, the critical psychological states, and on-the-job outcomes. We find out that ratings of job characteristics by respondents show a moderate level of convergence for most of the job dimensions. Correlations among the job dimensions are moderately positively inter-correlated, as has been found previously (Hackman & Lawler, 1971). As expected, the job dimensions are positively related to the job satisfaction.

29 Characteristics of University Graduates in Hong Kong The profile of respondents is shown in Table 1. The sample has approximately equal number of female and male respondents while skewing a little towards the male sex. An average respondent was graduated from the Chinese University of Hong Kong in 1989. The sample indicates almost equal split of number of graduates from the Arts, Science, Social Science and Business Administration while with just 2.7% from the Medicine field. The respondents have an average age of 25 to 26 with an average of 1 1/2 to 2 years of work experience in their present jobs. Their average pay level lies in the range of HK$10001 to HK$12500 per month. They do not change job frequently with an average number of job changes at 1 with low absenteeism. Limitations of Study 1. The questionnaire has been well-tested in Western countries but not in Oriental or Chinese societies. Even very frequently used instruments tested in foreign countries when used in Hong Kong, the instrument may miss some important points relevant (in this culture) to the subject under test. We run

30 the risk of imposing concepts developed in one culture to "another culture with different mind set or set of concepts. 2. Respondents may not be familiar with the research instrument. This may result in significantly different results. (Tirandis, 1972). Misunderstanding and unfamiliarity with the scale ratings will also influence the results obtained. Respondents may also find it difficult to give ratings for statement type questions. 3. No prior research has been done to collect Chinese specific information, such as their opinion on how important each of the job characteristic is, to compile the questionnaire. For instance, Chinese may not view skill variety as one of the core job characteristics. Therefore, comparing the results to previous studies conducted in Western societies may not be valid. 4. The respondents profile indicate that all respondents have at least university education and one year working experience. The fact that they have social experience and more highly educated may blur the Chinese-specific mediocracy mind set. As they have been trained or brought up to be pillars of society and therefore more willing to express

31 unpopular opinions. 5. General language problem is overcome with all respondents having university education. However, in case of "common English mistakes", respondents still may mis-interpret the questions. Implications for Future Study 1. In designing research instruments, it is better first to recruit a group of Chinese respondents. Then in the form of either open discussion, interviews, focus group discussion could be used to compile questions for research instruments. Afterwards, this local developed research instrument can be compared with Western well-tested instruments and then modified to suit Chinese culture. 2. Another approach is to use existing and very frequently used instruments so as to compare results with foreign studies. Then items could be modified and items could be added in order to incorporate the cultural specific definitions or way of expressing ideas. 3. When English is the medium of communication and is

32 not the respondents' mother tongue, we need to consider using short sentences, active voice, repeating nouns instead of pronouns, avoid metaphors and colloquialisms, avoid subjunctives, adding sentences to provide context for key ideas, avoid adverbs and prepositions telling "where" or "when", avoid possessive forms, use specific terms. (Lonner & Berry, 1986) 4. More extensive and vigorous pre-test prior research may be needed. It may bring out some problematic questions before actual data collection. To conclude, future research should be conducted to investigate empirically the potential bias of odd-number response scale for measures other than those tested in this study. _!Ill n

33 APPENDIX Figure 1 JOB ' CHARACTERISTICS MODEL CORE JOB DIMENSIONS... CRITICAL " PSYCHOLOGICAL STATES PERSONAL AND, ~ WORK OUTCOMES Skill variety Task identity Task significance Experienced meaningfulness of the work High internal work 'motivation High quality work performance Autonomy----------~>, Experienced responsibility for outcomes of the work High satisfaction with the work Feedback ----------~> Knowledge of actual results of the work activities Low absenteeism and turnover EMPLOYEE GROWTH NEED STRENGTH

34 Table 1 Sample Characteristics 5-point 4-point scale scale ------- ------- SEX Male 52.9% 51.5% Female 45.9% 47.8% AGE Below 23 years old 3.0% 3.7% 23-24 23.8% 31.3% 25-26 25. 8% 25.9% 27-28 28.8% 24.1% 29-30 9.2% 8.4% above 30 8. 2% 6.4% TENURE Below 1/2 year 14.4% 14.0% 1/2-1 14.6% 18.0% 1-1 1/2 20.6% 21.9% 1 1/2-2 10.4% 8.4% 2-2 1/2 13.2% 14.0% 2 1/2-3 4.5% 2.7% 3-3 1/2 8.7% 8.1% 3 1/2-4 2.5% 2.2% 4-4 1/2 11.2% 10.6% 4 1/2-5 0.0% 0.0%

35 Table 1 cont. PAY LEVEL Below HK$5000 HK$5000 - HK$7500 HK$7501 - HK$10000 HK$10001 - HK$12500 HK$12501 - HK$15000 HK$15001 - HK$17500 HK$17501 - HK$20000 Above HK$20000 5-point scale ------- 0.2% 5.0% 14.4% 22.3% 21.1% 14.1% 7.2% 14.4% 4-point scale ------- 0.2% 4.7% 17.7% 22.2% 22.9% 10.3% 6.2% 15.0% NO. OF JOB CHANGES o 1 2 3 4 5 Above 5 NO. OF DAYS ABSENT o 1 2 3 4 or more days GRADUATION YEAR FROM CUHK 1990 1989 1988 1987 1986 1985 MAJOR FACULTY IN CUHK Arts Science Social Science Business Administration Medicine 37.2% 32.3% 13.6% 10.7% 2.5% 1.7% 0.5% 65.8% 24.8% 3.2% 2.0% 1.5% 22.1% 16.9% 21.8% 18.1% 17.4% 0.2% 16.1% 26.1% 25.6% 26.3% 2.5% 40.6% 31.3% 13.1% 9.4% 2.7% 0.5% 1.7% 61.3% 26.4% 3.7% 2.0% 4.9% 24.4% 23.6% ' 19.5% 16.7% 13.8% 0.0% 18.7% 25.1% 25.1% 25.4% 3.7%

36 Table 2 Coefficient Alphas for All Measures No. of 5-point 4-point Items scale scale ------ ------- ------- Sample Size 403 ' 406 1 Skill Variety 3 0.7091 0.7317 2 Task Identity 3 0.8103 0.8021 3 Task Significance 3 0.7346 0.7880 4 Autonomy 3 0.7200 0.7062 5 Feedback from Job 3 0.3910 0.3292 6 Intention to Leave 3 0.8318 0.7951 7 Perceived Performance 5 0.5782 0.5991 8 Need for Achievement 5 0.6911 0.7358 9 Job Satisfaction 20 0.8646 0.8852 10 Motivating Potential Score 1

37 Table 3 Means and Standard Deviations 5-point Transformed 4-point scale scale* scale 1 Skill Variety 3.54 (0.74) 2 Task Identity 3.60 (0.83) 3 Task Significance 3.62 (0.73) 4 Autonomy 3.61 (0.77) 5 Feedback from Job 3.29 (0.62) 6 Intention to Leave 2.98 (0.97) 7 Perceived Performance 3.32 (0.53) 8 Need for Achievement 3.74 (0.51) 9 Job Satisfaction 3.42 (0.51) 10 Motivating Potential Score 44.94 (20.30) 2.91 2.8 (0.56) (0.61) 2.95 2.82 (0.62) (0.64) 2.97 2.87 (0.55) (0.61) 2.96 2.85 (0.58) (0.59) 2.72 2.67 (0.47) (0.44) 2.82 2.49 (0.73) (0.73) 2.95 2.66 (0.40) (0.46) 3.05 3.00 (0.38) (0.43) 2.81 2.75 (0.38) (0.40) 24.73 22.72 (10.23) (10.42) * The raw data of the five point scale sample was transformed to a four point scale by recoding 1 as 1, 2 as 1.75, 3 as 2.50, 4 as 3.25, and 5 as 4.

4-point scale set Transformed 4-point scale set Means S.D. n1 Means S.D. n2 Pooled S.D. variety 2.80 0.61 384 2.91 0.56 393 0.5852 Identity 2.82 0.64 384 2.95 0.62 392 0.6300 Significance 2.87 0.61 366 2.97 0.55 382 0.5801 Autonomy 2.85 0.59 396 2.96 0.58 395 0.5850 Feedback 2.67 0.44 373 2.72 0.47 382 0.4554 Performance 2.66 0.46 328 2.95 0.40 380 0.4288 Achievement 3.00 0.43 398 3.05 0.38 398 0.4058 Leave 2.49 0.73 387 2.82 0.73 397 0.7300 Satisfaction 2.75 0.40 335 2.81 0.38 357 0.3898 MPS 22.72 10.42 398 24.73 10.23 398 10.3254 t-values 0.1880 0.2064 * 0.1724 0.1880 0.1098 0.6763 0.1232 0.4521 0.1539 0.1947 t-3 PJ tj!---j CD.Po t-3 I < PJ!---J ~ CD rn H1 o ti 3: CD PJ ~ rn w (» * represents the difference between two means is significant t-value > 1.960, p < 0.05

39 Table 5 F-values for Standard Deviations Variety Identity Significance Autonomy Feedback Performance Achievement Leave Satisfaction MPS F-value 1.1865 1.0656 1.2301 1.0348 0.8764 1.3225 * 1.2805 1.0000 1.1080 1.0375 * represents the difference between two S.D. is significant ( 1.31 < F-value < 0.76 )

~ 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Skill 0.4351 0.5289 0.4317 0.4161-0.2007 0.2226 0.3424 0.4458 0.6278-0.1492 0.1006 0.0259 0.2186-0.0331-0.0206 0.0450 Variety (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.001) (0.022) (0.302) (0.000) (0.254) (0.341) (0.186)..-3 PJ 2 Task 0.4593 0.6808 0.4912-0.1315 0.1050 0.2484 0.4852 0.7496 0.0502 0.0604 0.0431 0.0843-0.0621 0.0155 0.0038 0' j---j ill Identity (0.000) (0.000) (0.000) (0.004) (0.017) (0.000) (0.000) (0.000) (0.157) (0.113) (0.194) (0.046) (0.107) (0.379) (0.470) 0"1 3 Task Signi~ ~ 0.4672 0.4434-0.2100 0.1006 0.2452 0.4238 0.6595-0.0369 0.0475-0.0542 0.1554 0.0113-0.0894 0.0526 Cl 0 ficance (0.000) (0.000) (0.000) (0.022) (0.000) (0.000) (0.000) (0.230) (0.170 ) (0.138) (0.001) (0.410) (0.037) (0.147) I"i I"i 4 Autonomy 0.4281-0.2688 0.1017 0.2287 0.5729 0.8316 0.0125 0.1390 0.1372 0.1423-0.0568-0.0364 0.1077 ill j---j PJ (0.000) (0.000) (0.020) (0.000) (0.000) (0.000) (0.401) (0.003) (0.003) (0.002) (0.128) (0.234) (0.016) rt 1-'- 5 Feedback -0.2090 0.0285 0.2725 0.4267 0.7601 0.0141 0.0003-0.0453 0.1471 0.0093 0.0019 0.0301 0 ~ from Job (0.000) (0.284) (0.000) (0.000) (0.000) (0.389) (0.498) (0.182) (0.002) (0.426) (0.485) (0.275) PJ 6 Intention 0.0294-0.0481-0.5657-0.2766 0.0410-0.1279-0.0671-0.2643 0.0396 0.1191-0.1020 0.. m to Leave (0.278) (0.168) (0.000) (0.000) (0.206) (0.005) (0.089) (0.000) (0.215) (0.009) (0.021) 1-'- lq ~ 7 Perceived 0 0.4817 0.0739 0.1638-0.0988 0.1336 0.0952 0.0945 0.1031-0.0284 0.1811 1-'- Performance (0.000) (0.069) (0.001) (0.024) (0.004) (0.028) (0.029) (0.019) (0.286) (0.000) HI 1-'- 8 Need for 0.1447 0.3599-0.0312 0.1108 0.1184 0.0608-0.0262-0.0210 0.1149 PJ ~ Achievement (0.002) (0.000) (0.266) (0.013) (0.009) 0 (0.112) (0.300) (0.338) (0.011) ill 9 Job Satis- 0.5930-0.0091 0.0742 0.0517 0.2737-0.0917-0.0773 0.0316 faction (0.000) (0.428) (0.068) (0.149) (0.000) (0.062) (0.062) (0.264) ~ to 0 1-'- ~ rt m 0 PJ j---j ill

0.3686 (0.000) -0.0312 0.0032 (0.268) (0.474) -0.0144 0.1483 (0.386) (0.002) 0.1455-0.0743 (0.002) (0.069) -0.1679-0.0138 (0.000) (0.391) -0.0414-0.1096 (0.204) (0.014) -0.0456 (0.182) 0.0953 (0.030) -0.2725 (0.000) 0.5957 (0.000) 0.4741 (0.000) 0.4179 (0.000) 0.3424 (0.000) -0.1160 (0.011) 1-:3 PJ tr t--j CD 0'\ 0 0 ~ rt H:=a r-' 10 Motivating Potential Score 11 Sex* 12 Age -0.0062 0.1005 0.0533 0.1867 (0.451) (0.022) (0.143) (0.000) -0.2861-0.1452-0.2517 (0.000) (0.002) (0.000) 0.3739 0.4451 (0.000) (0.000) 13 Tenure 14 Pay Level 15 Change 16 Absent 17 Year * Male is coded as 1 while female as 2

~ 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Skill 0.4305 0.5001 0.4674 0.393-0.2055 0.3035 0.325 0.4292 0.6397-0.102 0.0752 0.0394 0.116 0.0567 0.0325 0.0958 Variety (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.021) (0.067) (0.215) (0.010) (0.130) (0.261) (0.030) 1-3 AJ 2 Task 0.3037 0.72 0.4678-0.1336 0.2456 0.2294 0.431 0.7507-0.0421 0.0712 0.0219-0.0383 0.0656 0.0528 0.0546 tr ~ CD Identity (0.000) (0.000) (0.000) (0.004) (0.000) (0.000) (0.000) (0.000) (0.201) (0.078) (0.330) (0.223) (0.096) (0.148) (0.141) (fj 0 AJ ~ CD ~ ~ 3 Task Signi- f-j- ~ 0 0.3908 0.2834-0.233 0.2586 0.1645 0.3288 0.5394-0.0233 0.0971-0.0011 0.1114 0.014-0.0441 0.0646 () ficance (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.322) (0.027) (0.491) (0.013) (0.391) (0.192) (0.102) 0 11 11 4 Autonomy 0.4086-0.18 0.2826 0.238 0.5198 0.8214-0.0138 0.0686 0.0169 0.0186 0.0666 0.0083 0.1047 CD ~ (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.392) (0.086) (0.368) (0.356) (0.093) (0.435) (0.019) AJ rt 5 Feedback -0.2023 0.1474 0.1942 0.3489 0.7604-0.08-0.0173-0.0457-0.012 0.0953-0.0001-0.0024 0 ~ from Job (0.000) (0.002) (0.000) (0.000) (0.000) (0.056) (0.365) (0.180) (0.406) (0.029) (0.499) (Q.481) AJ 6 Intention 0.011-0.0182-0.5048-0.2275 0.0695-0.0821 0.0133-0.2094-0.0117 0.1661-0.0698 ~ to Leave (0.413) (0.358) (0.000) (0.000) (0.084) (0.051) (0.396) (0.000) (0.408) (0.000) (0.085) (fj f-jto 7 Perceived 0.4857 0.1096 0.3056-0.1473 0.1102 0.0766 0.1415 0.017-0.0123 0.1397 ~ f-jrn Performance (0.000) (0.014) (0.000) (0.002) (0.014) (0.063) (0.002) (0.368) (0.404) (0.003) f-j- 8 Need for 0.1585 0.3045-0.0731 0.0437 0.0649 0.0639-0.0255-0.016 0.0229 AJ ~ Achievement (0.001) (0.000) (0.073) (0.193) (0.097) (0.102) (0.307) (0.377) (0.327) 0 CD 9 Job Satis- 0.5274 0.0032 0.0663-0.0104 0.1613-0.0468-0.0129 0.017 faction (0.000) (0.474) (0.093) (0.418) (0.001) (0.176) (0.399) (0.369) ljl '"d 0 f-j- ~ rt

0.3297 (0.000) 0.0986 0.0088 (0.025) (0.431) -0.0437 0.0326 (0.193) (0.260) 0.1472-0.01 (0.002) (0.422) -0.2579 0.0089 (0.000) (0.431) -0.0217-0.0612 (0.333) (0.113) -0.0123 (0.405) 0.0916 (0.036) -0.2064 (0.000) 0.5548 (0.000) 0.4806 (0.000) 0.4467 (9.000) 0.3392 (0.000) -0.1154 (0.012) t-3 Pl ~ t--j Cl) -...J (1 0 ~ rt ~ W 10 Motivating Potential Score 11 Sex* 12 Age -0.0612 0.0575-0.0085 0.0428 (0.112) (0.126) (0.433) (0.198) -0.3017-0.1405-0.2269 (0.000) (0.002) (0.000) 0.354 0.4494 (0.000) (0.000) 13 Tenure 14 Pay Level 15 Change 16 Absent 17 Year * Male is coded as 1 while female as 2

44 Questions used in Questionnaire are listed as follows: Section I: Part 1: Job Characteristics Skill Variety (1) My job requires me to do many different things, using a number of different skills and talents. (2) The job requires me to use a number of complex or high-level skills. (3) There is so much variety in my job. Task Identity (4) I find that my values and the organization's values are very similar. (5) I am proud to tell others that I am part of this organization. (6) Often, I find it easy to agree with this organization's policies on important matters relating to its employees. Task Significance (7) The outcomes of my work can affect other people in very important ways. (8) This job is one where a lot of other people can be affected by how well the work gets done. (9) The job itself is very significant or important in the broader scheme of things.

45 Autonomy (10) My job gives me almost complete responsibility for deciding how and when the work is done. (11) The job provides me the chance to use my personal initiative or judgement in carrying out the work. (12) The job gives me considerable opportunity for independence and freedom how I do the work. Feedback (13) The job is set up so that I get almost constant "feedback" as I would about how well I am doing. (14) Doing the job itself provides many chances for me to figure out how well I am doing. (15) The job itself provides quite a few clues about whether or not I am performing well. Part 2: Individual Characteristics Intention to Leave (16) I often think about quitting. (17) I will probably look for a new job in the next year. (18) If I were completely free to choose, I would prefer continue working in this organization.

Self-perceived Performance 46 (19) My job performance is better than other people of the same position. (20) My job performance is excellent for the present position. (21) I am an average worker. (22) There are a lot of rooms to improve my job performance. (23) My job performance is so good that I deserve a large raise in salary than other employees in this organization. Need for Achievement (24) I always try very hard to improve my past performance at work. (25) I always try to get added responsibilities on my job. (26) I always try to perform better than my co-workers. Section II: Job Satisfaction (27) Being able to keep busy all the time (28) The chance to work alone on the job. (29 ) The chance to different things from time to time. (30) The chance to be "somebody" in the community. (31 ) The way my boss handles his men. (32) The competence of my supervisor in making decision. (33) Being able to do thins that don't go against my

47 conscience. (34 ) The way my job provides for steady employment. (35 ) The chance to thing for other people. (36) The chance to tell people what to do. (37) The chance to do something that makes use of my abilities. (38) The way company policies are put into practice. (39) My pay and the amount of work I do. (40) The chance for advancement on this job. (41) The freedom to use my own judgement. (42) The chance to try my own methods of doing the job. (43) The working conditions. (44) The way my co-workers get along with each other. (45) The praise I get for doing a good job. (46) The feeling of accomplishment get from the job.

1I1er!ilWl!LM Ji!lll! 48 REFERENCES References in English: Cammam, c., Fichman, M., Jenkins, D., & Klesh, J. (1979). The Michigan organizational assessment questionnaire. Unpublished manuscript, University of Michigan, Ann Arbor, Michigan. Collins, B.E. (1974). Four components of the Rotter internal-external scale: Belief in a difficult world, a just world, a predictable world, and a politically responsive world. Journal of Personality and Social Psychology, 29, 381-391. Edwards, A. L. (1959). Manual of Edwards personal preference schedule. The Psychological Corporation. N. Y. Hackman, J.R., & Oldham, G. R. (1975). Motivation through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16, 250-279. Hoadley, J. S. (1970). "Hong Kong is the lifeboat": Notes on political culture and socialization. Journal of Oriental Studies, 8, 206-218. Podmore, D., Chaney, D., & Golder, P. (1975). "Don't know" responses among young adults in Hong Kong. Journal of Social Psychology, 96, 307-308. Porter, L. W., Steers, R. M., Mowday, R. T., & Boulian, P. V. (1974). Organization commitment, job satisfaction, and turnover among psychiatric

49 technicians. Journal of Applied Psychology, 5, 603-609. Weiss, D. J., Dawis, R.' V., England, G. W., & Lofquist, L. H. (1967). Manual for the Minnesota Satisfaction Questionnaire: Minnesota studies in vocational rehabilitation. Minneapolis: Industrial Relations Center, University of Minnesota. References in Chinese: Chiu, H. Y. & Yang K. S. (1974). Institute of Ethnology, Academic Sinica, Monograph Series B, No. 4, 381-410. Nankang, Taipei, Republic of China. Yang, K. S. (1982). Institute of Ethnology, Academic Sinica, Mongraph Series B, No. 10, 153-188. Nankang, Taipei, Republic of China.

CUHK Libraries 111111111111111111111111111111111111111111111111111111I 000348111