m... . . . . I. - . . . .. m....,......., ...x. .n. . _ . . . ... . _ _ . . . ......e: ..H.§._r.-;r . ... ... 4.... .. .....o. . ...4. ....... . _ . . ....f ..H.. . .. .... .... .2.............q.., ...... . _ . . . W . _ _. . m . .4 . A. _. . . . . .2 ...»...J. . M B . m no . . .. . V .110 .N M S S . . ... 2w. m I 0 (I: R N . . . ... sf“. .. . , A e. O . . z .. . . em gum :.af} . N I. m WI; maul 0 R W 7 . .. .. m ..J M N D . .. _. . 0 D f. M N _ . . .H N ...». I 0 .. .. .. . R m M M . .. . - . R . . .... . .- . H . n. ... . 'P uI . ... I... . . .. . . . . .. ....... .o ,. .. .... 0‘... 0.... o . . .. .. ... . . o ....... - .I..1.. f ,.a . ... . ... ._ r. I .o J 1.. u .. . . . . I . ~ ..I ... o o § . ......o... ......u.. \ . . . .c . .59... ...4...) ...v‘:¢.‘r. } . A . ... 0440. 31"... \ . c ... , . ..p.._..... .....u..o....r9.9 - . o . . . O . . | . . n . . . p . . .. . . r o. v . . . o. . . . .. _. 2... . .. .... 4.x . x . ... -..-......t. 2..... 1.1.... .. ...:.......o......ub...s.o. a.._.~_.........¢.-.~A at... «35.. 31%;... ..efl.fiv \. .. o o . . l] iln‘ n1 i LIBRARY 1' Michigan State Univcmity I ABSTRACT INTERRELATIONS AMONG SEX, JOB LEVEL, AND JOB ORIENTATION By Raymond Howard Johnson This study investigated the following hypotheses: l. Males and females differ significantly on the amount of importance they each place on Herzberg's dichotomous job factors when the effects of age, education, and job level are not controlled. Males and females do not differ significantly on the amount of importance they each place on Herzberg's dichotomous job factors when the effects of age, educa- tion, and job level are controlled. Supervisors place greater importance on intrinsic job factors than do rank-and-file employees. Rank-and-file employees place greater importance on extrinsic job factors than do supervisors. Males and females do not differ on the amount of impor- tance they place on certain specific job factors when the effects of age, education, and job level are controlled. Males and females differ on the amount of importance they each place on certain specific job factors when the effects of age, education, and job level are not controlled. The control variables age and education are related to intrinsic and extrinsic job orientation, with job level controlled. The subject sample consisted of 2755 employees of six medium sized industrial organizations who responded to a questionnaire. Job orientation was measured using a normative scale. All hypotheses were tested using analysis of variance techniques. Raymond Howard Johnson The results strongly support hypotheses 1 and 2. Hypothesis 3 was partially supported, i.e., job levels only differed on extrinsic job orientation, in the predicted direction. Hypothesis 4 was largely supported, with sex differences emerging on only 2 of the 16 job factors. Similarly, hypothesis 5 was largely supported, with sex differences emerging on 9 of the 16 job factors. Hypothesis 6 was partially supported. In the rank-and-file group, only education was related to extrinsic job orientation. In the supervisory group, age and education were both related to extrinsic job orientation. Only age was related to intrinsic job orientation. Job level was thus found to moderate the relationships among age, education, and job orientation. Thesis Committee: Dr. Carl Frost, Chairman Approved: Date: Dr. Robert Ruh Dr. Frank Schmidt INTERRELATIONS AMONG SEX, JOB LEVEL, AND JOB ORIENTATION By Raymond Howard Johnson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1971 067054 ACKNOWLEDGMENTS The name on this thesis is mine, but the contributions to it from other people are many. Dr. Robert Ruh stimulated many of the hypotheses under investi- gation by introducing the writer to relevant research articles. In addition, his empathy with my computer problems was greatly appreciated. Dr. Frank Schmidt was a constant source of information on quanti- tative matters. When the problems arose, he was there with the solutions. Dr. Carl Frost, my committee chairman, has been a spring of encouragement, both to the completion of this thesis, and to the pursuit of my graduate education.i His confidence and support were invaluable, as well as his helpful comments which smoothed the roughness from the first rough draft of the thesis. Finally, my wife deserves most of the credit for this thesis. She has been a source of consolation when the computer didn't compute, a patient proofreader, and a tolerant listener to her husband's ravings on statistics and other mysterious topics. But most impor- tant, she has been a loving and understanding wife. ii TABLE OF CONTENTS LIST OF TABLES O O C O O O O O O O O O O O O O O O O O O O 0 REVIEW OF THE LITERATURE O 0 O O O O O O O O O O O O O O O 0 man S O O O O O C C O O O O O O O O O O O O O O O O O O O O HYPOTHESES AND ANALYSES . . . . . . . . . . . . . . . . . . . RESULTS 0 O O O O O O O O O O O O 0 0 O O O O O O O O O O O O DISCUSS ION O O 0 O O O O O O O O O O O O O O O O O O O O O O BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . APPENDIX A SORTING INSTRUCTIONS AND RESULTS OF SORTING . . . . . . APPENDIX B JOB ATTITUDE ITEMS AND INSTRUCTIONS TO SUBJECTS . . . . APPENDIX C SUMMARY OF ANALYSES OF VARIANCE FOR THE RELATIONSHIP OF SEX (A) TO EACH OF THE 16 JOB ORIENTATION ITEMS, IN ORIGINAL AND CROSS-VALIDATION SAMPLES, WITH AND WITHOUT JOB LEVEL CONTROLLED . . . . . . . . . . . . . . APPENDIX D CORRELATIONS AMONG HERZBERG'S 16 JOB ORIENTATION ITEMS . APPENDIX E CORRELATIONS AMONG AGE, EDUCATION, JOB LEVEL, INTRINSIC ORIENTATION, EXTRINSIC ORIENTATION, AND SEX . . . . . . APPENDIX P SAMPLE COMPOSITIONS AND T VALUES FOR SEx DIFFERENCES 0N AGE, EDUCATION, AND JOB LEVEL . . . . . . . . . . . . iii Page iv 15 33 42 44 47 50 60 61 62 LIST OF TABLES Table Page 1. Summary of analysis of variance for the relationship between sex (A) and extrinsic job orientation (job level not controlled). . . . l6 2. Summary of analysis of variance for the relationship between sex (A) and intrinsic job orientation (job level not controlled). . . . l6 3. Mean job orientation scores by sex (job level not controlled) . . . . . . . . . . . . . . l7 4. t values and means for sex differences on age, education, and job level in samples in which job level was not controlled . . . . . . . . . . 63 S. t values and means for sex differences on age, education, and job level in samples in which job level was controlled . . . . . . . . . . . . 65 6. Summary of analysis of variance for the relation- ship of sex (A) and job level (3) to extrinsic job orientation (job level controlled). . . . . . l9 7. Summary of analysis of variance for the relation- ship of sex (A) and job level (B) to intrinsic job orientation (job level controlled). . . . . . l9 8. Mean extrinsic job orientation for sex by job level (job level controlled) . . . . . . . . . . 20 9. Mean intrinsic job orientation for sex by job level (job level controlled) . . . . . . . . . . 20 10. Component and compound probabilities of mean differences of 16 job orientation items by sex (job level controlled) . . . . . . . . . . . 22 11. Job orientation item means by sex in two samples (job level controlled) . . . . . . . . . 23 iv Table 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Component and compound probabilities of mean differences of 16 job orientation items by sex (job level not controlled). . . . . . . . . Job orientation item means by sex in two samples (job level not controlled). . . . . . . Summary of analysis of variance for the relation- ship of age (A) and education (B) to intrinsic job orientation within rank-and-file. . . . . Mean intrinsic job orientation by age and education within rank-and-file. . . . . . . . Summary of analysis of variance for the relation- ship of age (A) and education (B) to extrinsic job orientation within rank-and-file. . . . . Mean extrinsic job orientation by age and education within rank-and-file. . . . . . . . Summary of analysis of variance for the relation- ship of age (A) and education (B) to intrinsic job orientation within supervisory. . . . . . ‘Mean intrinsic job orientation by age and education within supervisory . . . . . . . . Summary of analysis of variance for the relation- ship of age (A) and education (B) to extrinsic job orientation within supervisory. . . . . . Mean extrinsic job orientation by age and education within supervisory . . . . . . . . Summary of analyses of variance for the relation- ship of sex (A) and job level (B) to each of the 16 job orientation items for sample 1 (job level controlled) . . . . . . . . . . . Summary of analyses of variance for the relation- ship of sex (A) and job level (B) to each of the 16 job orientation items for sample 2 (job level controlled) . . . . . . . . . . . Summary of analyses of variance for the relation- ship of sex (A) to each of the 16 job orienta- tion items for sample 1 (job level not contrOIIed) O I O O O O O O O O O O O O O O O O Page 25 26 29 29 30 3O 31 31 32 32 SO 53 56 Table Page 25. Summary of analyses of variance for the relation- ship of sex (A) to each of the 16 job orienta- tion items for sample 2 (job level not controlled) . . . . . . . . . . . . . . . . . . . 58 26. Correlations among Herzberg's 16 job orientation items . O C C C C C C C O O O O O C C O C O O O C 60 27. Correlations among age, education, job level, intrinsic orientation, extrinsic orientation, and sex 0 O C O O O O O O O O O O O O O O O O O O 61 vi REVIEW OF THE LITERATURE Herzberg's well-known study, The Motivation To Ng;k_(l959), resulted in a dichotomization of work attitude factors, i.e., elements of work orientation. Job content factors ("motivators") were delineated as achievement, responsibility, recognition, advance- ment, and nature of the work. Job context factors ("hygiene") were delineated as working conditions, salary, interpersonal relations with supervisor, subordinates, and peers, technical supervision, company policy and administration, status, job security, and personal life factors. This dichotomy has also been expressed as intrinsic and. extrinsic job orientation, respectively (Saleh and Lalljee, 1969). Not many studies have investigated the relationship of these job factors to the sex variable. Herzberg, et al. (1957), reviewing the literature on job attitudes, concluded: "In general, intrinsic aspects of the job appear to be more important to men than to women" (p. 72). Jurgenson (1947) compared males and females on job orientation and found similar differences, but there was no control over occupational differences. However, more recent studies do not support this conclu- sion. Burke (1966a, 1966b) found no significant differences between sexes when male and female college students were asked to rank order ten job factors (five extrinsic, five intrinsic) in order of importance. Both sexes preferred intrinsic factors. These differing results cannot be attributed to differences in subject samples (working vs. student population) because Centers and Bugental (1966) also found no signi- ficant differences between the sexes using a sample of the working population. Mere specifically, Centers and Bugental studied "692 employed adults representing a cross section of a major urban area." Each subject was given a card listing six job factors--three intrinsic and three extrinsic--and was asked to rank order the three most impor- tant. Although some sex differences were found on particular items ("women placed a higher value on interpersonal relations than did men, and men laid greater importance on self-expression than did women"), no sex differences were found on the dichotomous variable, intrinsic vs. extrinsic job orientation. These specific sex differences are consistent with previous research findings. Kulen (1963) found that occupational concerns are psychologically more central to males than females. This suggests why men place higher value on self-expression-- using one's skill--than females. Hardin, Reif, and Heneman (1951) found that women place more emphasis on social factors on the job than do men. This suggests why women place higher value on having "good co-workers" than men. Saleh and Lalljee (1969) account for the differences in the results of Centers and Bugental's study and those reviewed by Herzberg by suggesting that the latter studies were contaminated by non-sex variables, namely, education, job level, and age. To test the hypo- thesis that there are no significant differences in job orientation, they used the "Job Attitude Scale," which is a forced-choice measure consisting of 60 pairs-~intrinsic vs. extrinsic job factors--developed by Saleh (1964). They administered the scale to three subject samples-- college students, teachers, and service industry employees. In each of these three groups, where age, education, and job level were con- trolled, there were no significant differences in job orientation between the sexes. In a fourth industry group, for which there were no controls for age, education, and job level (education and job level were significantly different for males and females) males emerged more intrinsically oriented than females. Combining sexes, and comparing a supervisor sample with a clerk sample (age and education controlled), the supervisor group was found to be more intrinsically oriented than the clerk group. This last finding supports the hypothesis that the higher the job level, the higher the level of intrinsic job orientation regardless of sex (Centers, 1948; Centers and Bugental, 1966; Gurin, et al., 1960). It should be noted that the study by Saleh and Lalljee (1969) did not probe into the sex differences on specific types of job orientation, nor did it systematically investigate the exact relation- ship of the age and education control variables to the two types of job orientation. It is also important to note that the measurement of job orienta- tion was ipsative in nature in many of the studies relating sex to job orientation. This is especially true of the Saleh and Lalljee study (1969) in which the "Job Attitude Scale" was a forced-choice measure. It was thus impossible for a given respondent to score high on intrinsic and extrinsic orientation. In this study, the method of measurement was normative in nature. It is thus possible for a respondent to consider both types of job orientation highly important. The purpose of this current study is to elaborate and clarify the research findings cited thus far by: (l) examining the relationship between sex and job orientation with and without controls for age, education, and job level; (2) examining the relationship between job level and job orientation; (3) examining sex differences on particular job orientation items, with and without controls for age, education, and job level; (4) examining the exact relationship of age and education to job orientation in each of two job levels. These relationships will be examined using normative measures of job orientation, a more appropriate mode of analysis, and a different subject sample from that used in previous research. METHODS Subjects The subjects were 2755 employees of six medium-sized manufac- turing companies in the Midwest. These employees completed the 1968 Michigan State University "You and Your Job" attitude survey, which represents the source of data for this study. Of 4162 questionnaires given out, 2755 usable questionnaires were returned, providing a return rate of .66. Of these 2755 subjects, 1646 subjects responded to every relevant item. All analyses were conducted using only these complete data respondents. Su ve Demo ra hic Items In addition to the attitudinal items on the survey, respondents were asked to provide data regarding their personal background. The demographic variables of interest to this study were: age, sex, number of years of education, and hierarchical level. Survey Attitude Items The section in the survey on "individual goals" provided the items for analysis. In this section respondents were asked to evaluate various aspects of their jobs and surrounding conditions in terms of their relative importance. Of the 24 items in the section, 16 items representing Herzberg's job factors were selected. There were seven extrinsic and nine intrinsic items. Five psychology graduate students were asked to sort the 24 items, each on a three by five card, into one of three categories--intrinsic, extrinsic, or neither. All five graduate judges were given the same written instruc- tions defining the intrinsic and extrinsic categories in Herzberg's words (Herzberg, 1959). Only those items which were placed in the same category by at least four of the five judges were used. See Appendix A for sorting instructions and results of sorting by the five independent graduate student judges. The response measure was a five point Likert-type scale. Coding of the responses was as follows: 1 - not at all important 2 - slightly important 3 - somewhat important 4 - very important 5 - extremely important For instructions to the respondents concerning the 24 individual goal items, and exactly which items were selected for analysis as representing extrinsic or intrinsic orientation, see Appendix B. The alpha estimates of internal consistency reliability for the intrinsic and extrinsic job orientation scales were .80 and .70, respectively. The correlation between the scales was .41. These correlations were based on a sample size of 1646 respondents with complete data on these variables. Medias The age variable was trichotomized into the following categories: Low - 30 years or less Medium - 31 to 50 years High - 51 years or greater The education variable was trichotomized into these groups: Low - less than high school graduate Medium - high school graduate High - beyond high school These trichotomies were for purposes of a least-square analysis of variance for cells with unequal frequencies. It is assumed that the unequal cell frequencies reflect the population frequencies. The above levels were used to analyze the relationship of age and educa- tion to job orientation for rank-and-file employees. Because of technical difficulties on the computer, the age variable had to be trichotomized in a slightly different manner in order to examine the relationship of age and education to job orienta- tion for the supervisory group. This adjusted trichotomization of age was the following: Low - 31 years or less Medium . 32 to 49 years High - 50 years or greater In the analysis which compared males and females on job level and education, the coding was as follows: Job Level: 9 - rank-and-file 8 - lead-man 7 - first-line supervisor 6 - department supervisor Education: 10 grad school 9 college 8 - some college ‘1 I 12 years 0‘ I 11 years U! I 10 years 4 - 9 years 3 = 8 years 2 - 7 years 1 = 6 years or less HYPOTHESES AND ANALYSES H othesis l Hypothesis 1 predicts that males and females differ significantly on the amount of importance they each place on Herzberg's dichotomous job factors, i.e., intrinsic and extrinsic job orientation, when the effects of age, education, and job level have not been controlled. An "interaction" effect is predicted. Specifically, males place greater importance on intrinsic orientation factors than females. Conversely, males place less importance on extrinsic orientation factors than females. This hypothesis is derived from Saleh and Lalljee (1969), and is essentially a replication of a part of their study. However, a non- ipsative measurement technique, a different population, and a more appropriate mode of analysis will be applied. Analysis: This interactive hypothesis was tested using a one-way, least-squares analysis of variance design. The independent variable was sex with its two levels. The two dependent variables were the respondents' mean response to the extrinsic scale items, and mean response to the intrinsic scale items. To indicate that the extraneous variables age, education, and job level were operating, t test analyses were performed on the analysis of variance sample. That is, three t test analyses compared the means of males and females on age, education, and job level. 9 10 Hypotheses 2 and_3 Hypothesis 2 predicts that males and females do not differ sig- nificantly on the amount of importance they each place on Herzberg's dichotomous job factors (intrinsic and extrinsic orientation) when age, education, and job level have been controlled. No significant interactive effect is predicted. That is, there should be no sex differences on either of the two dependent variables. This hypothesis is also derived from Saleh and Lalljee (1969). It is a replication with different scale measurements (non-ipsative), a more powerful mode of analysis, and a different population. Hypothesis 3 predicts that the higher the job level (supervisory), the greater the importance of intrinsic job factors; the lower the job level (rank-and-file), the greater the importance of extrinsic job factors. Again, this is essentially a replication hypothesis, with the unique aspects being (1) non-ipsative measurement scales, (2) more appropriate analysis technique, and (3) different population. Analyses for hypotheses 2 and 3: To test these hypotheses, the same 1646 respondents with complete data that were used in testing hypothesis 1 were employed. A least-squares analysis of variance was applied in a sex by job level design. Age and education were already controlled in that t tests indicated no sex differences on these variables. The groups of males and females were not deliberately matched on age and education. The above design was run twice, once with extrinsic job orientation as the dependent variable and once with intrinsic job orientation as the dependent variable. 11 To support hypothesis 3, there should be a significant job level effect in the appropriate direction for each of the dependent variables. Hypothesis 4 It is predicted that with the effects of age, education, and job level controlled, males and females do not differ significantly on the amount of importance they place on the following job factors: (1) interpersonal relations (items E77, E84, E75); (2) self-expression or using one's skill (items I74, I92, 194). This aspect of the hypothesis is designed to replicate item sex differences found by Centers and Bugental (1966), Kulen (1963), and Hardin, Reif, and Heneman (1951). However, Centers and Bugental did not control for job level when they looked at specific job orientation differences between the sexes. In addition to hypothesis 4, other possible item sex differences on the remaining 10 items will be explored. Analysis: To test for sex differences on particular job orientation items, a two-way analysis of variance design was used. This design was run 16 times, once for each item dependent variable, with sex and job level as factors, each with two levels. These 16 sex-by-item comparisons were run in two samples. The total group of 1646 respondents with complete data were randomly assigned to two groups, with the stipulation that each group has exactly equal numbers of male and female rank-and-file, and exactly equal numbers of male and female supervisors. Because of the increased probability with multiple t (and F) tests showing significant sex difference on items due to chance alone, 12 the following criteria for reliable sex differences on items were used: the t value (or F) must be significant at least to the .20 level in both original and cross validation samples, and the compound probability must be £2.05. Hypothesis 5 This global hypothesis consists of two parts. First, it predicts that when job level is not controlled (essentially the condition in Centers and Bugental's study), males place greater importance on job orientation items dealing with self-expression or using one's skill (items I74, I92, I94). Females place greater importance on job orien- tation items dealing with interpersonal relations (items E77, E84, E75). The second aspect of this global hypothesis predicts that the pattern of sex differences on the 16 job orientation items will change from those found, under the conditions of hypothesis 4, i.e., when age, education, and job level are controlled. Analysis: Sixteen least-squares, analyses of variance were run, one for each item dependent variable. The design was sex by job level. The same criteria for reliable sex differences on items (as in hypothesis 4) were applied. However, since the initial conditions of this hypothesis resembled Centers and Bugental's study, one-tailed t tests were used to test the directional differences on the six items (items I74, I92, I94, E77, E84, E75) on which previous research has found significant sex differences. These six items were tested using a signed t derived from the analysis of variance (t - VF). 13 Hypothesis 6 This exploratory hypothesis is directed at the relationship of the control variables age and education to job orientation. It is predicted that age and education are significantly related to both intrinsic and extrinsic orientation, with job level controlled. The age portion of this hypothesis is partly derived from Saleh (1964), who found that pre-retirement managers demonstrated less con- cern with intrinsic job factors than when they were younger. The hypothesis in general was stimulated by the fact that most of the studies that have been cited in this thesis were either criti- cized for lack of control of age and education, or confidently con- trolled for it, without mention of exactly how lack of control would have affected the dependent variable. This is the question to be answered. Agglygigz This hypothesis was tested using a least-squares, two-way analysis of variance. The design was age by education, and the depen- dent variables were intrinsic and extrinsic job orientation. The sample consisted of 1200 complete data rank-and-file_employees, and 76 supervisors to control for job level. Within each job level, there were equal numbers of males and females. In addition to controlling the effects of different job levels, separate analyses done within each of the two job levels may demon- strate a moderating effect of job level on the relationship of age and education to job orientation. The benefit of this analysis of variance approach over the corre- lational approach is that interactive effects on the dependent variable 14 can be detected, and possible non-linear relationships between the independent variable and dependent variable may emerge. RESULTS Hypothesis 1 Tables 1-4 contain the information necessary for the testing of hypothesis 1. Table 1 indicates a highly significant difference in extrinsic job orientation, with females being more extrinsically oriented than males. Similarly in Table 2, a highly significant difference occurs, indicating that males are more intrinsically job oriented than females. The direction of the sex differences in job orientation is seen from inspection of the means in Table 3. Because the differences in job orientation by sex were both significant in the predicted direction, an interaction effect can be inferred, even though it is not possible to directly test it statis- tically with this design. The reader will recognize the nature of this inferred interaction from inspection of Table 3 means. Hypothesis 1 is thus supported by the analysis. It should be noted from Table 4 in Appendix F that males and females in this sample of 1646 respondents did not differ significantly in mean age or mean education, so that differential effects of these variables on the dependent variables were probably not operating. Males and females did differ significantly on job level, with males, on the average, occupying jobs higher in the hierarchical structure of the organization. See Appendix F for exact composition of the sample used in this analysis, i.e., the distribution of sexes by job level. 15 16 Table 1. Summary of analysis of variance for the relationship between sex (A) and extrinsic job orientation (job level not controlled). Source df MS F A l 6.77 22.63* Error 1644 .30 Total 1645 *p (.0005 Table 2. Summary of analysis of variance for the relationship between sex (A) and intrinsic job orientation (job level not controlled). Source df MS F A l 14.05 31.52* Error 1644 .45 Total 1645 *p (.0005 17 Table 3. Mean job orientation scores by sex (job level not controlled). Sex Job Orientation Male Female Intrinsic 3.89 3.71 Extrinsic 4.15 4.28 18 fiypotpegis 2 Data necessary to test hypothesis 2 are seen in Tables 5-9. The analysis of variance Table 6 indicates no significant re- lationship between sex and extrinsic job orientation. The interaction was also non-significant. It is seen from inspection of Table 5 that males and females in this sample did not differ significantly from one another on age and education. The variable on which the sexes did differ significantly, job level, was controlled as a factor in the design. See Appendix F for the distribution of sexes by job level, such that the effect of different job levels is the same for males and females. Similarly, the analysis of variance Table 7 indicates no signi- ficant relationship between sex and intrinsic job orientation, with age, education, and job level controlled. The interaction was also non-significant. Thus hypothesis 2 was supported. No inferred interactive effect emerges. When the effects of age, education, and job level are con- trolled, males and females do not differ in job orientation. Hypothesig 3 This hypothesis was tested using information found in Tables 6-9. Table 6 indicates a highly significant relationship between job level and extrinsic job orientation in the predicted direction. From Table 8, it is clear that supervisors were less extrinsically oriented than rank-and-file. However, looking at Table 7, no significant relation- ship between job level and intrinsic job orientation emerges. Super- visory and rank-and-file people did not differ in the importance they each placed on intrinsic job factors. Thus, hypothesis 3 was not 19 Table 6. Summary of analysis of variance for the relationship of sex (A) and job level (B) to extrinsic job orientation (job level controlled). Source df MS F A l .58 2.16 r B 1 5.37 20.03* A x B l .03 .10 ' Error 1272 .27 g *P < .0005 2* Table 7. Summary of analysis of variance for the relationship of sex (A) and job level (B) to intrinsic job orientation (job level controlled). Source df MS F A 1 .65 1.44 B l .21 .47 A x B l .31 .69 Error 1272 .45 20 Table 8. Mean extrinsic job orientation for sex by job level (job level controlled). Sex Job Level Male Female Total Supervisory 3.94 4.04 3.99 Rank and File 4.23 4.30 4.26 Total 4.08 4.17 Table 9. Mean intrinsic jbb orientation for sex by job level (job level controlled). ._____$£§.___ Job Level Male Female Total Supervisory 3.87 3.84 3.85 Rank and File 3.88 3.72 3.80 Total 3.87 3.78 21 entirely supported. It should be noted that the supervisory group had significantly more education than the rank-and-file (r - -.14, p‘(.OOl). The supervisors were also significantly older than the rank-and-file group (r - -.12, p (.001). Refer to Appendix E for the source of the above correlations. Hypothesis 4 The compound probabilities found in Table 10 are relevant to this hypothesis. NOte that sex differences emerge on only 2 of the items, controlling for job level, age, and education. Both of the discriminating items dealt with intrinsic job orientation. Item I85 concerned the importance of having good chances for promotion. Item 192 dealt with the importance of being able to decide how to do your job. On both of these items, males placed greater importance than females. Table 11 contains the job orientation item by sex means in both original and cross-validation samples. Only one of the sex differences found by Centers and Bugental emerged: (I92) self expression in deciding how to do the job. No sex differences on the three items involving interpersonal relations (E77, E84, E75) emerged as significant. Thus, since only one of the six item differences emerged, this replication aspect of hypothesis 4 was mainly supported. It should be noted that t tests were performed in both job control samples 1 and 2 to check on the random assignment of subjects to the samples. As was the case with the total sample of 1276 sub- jects, both samples showed no sex differences on age, education, or u..r:w".v.‘-rx--r _ - an; m 5.. ,. s 1.3 an. -- 22 Table 10. Component and compound probabilities of mean differences of 16 job orientation items by sex (job level controlled). Item P in Sample 1 P in Sample 2 Compound P E73 .28 (M (F) .01 (MF) .56 (M< F) N.S. E81 .55 (M 7F) .02 (MF) .18 (M >F) N.S. I85 . 17 (M >F) .04 (M>F) (.05 I87 .30 (M< F) .61 (MF) .64 (M >F) N.S. I91 .95 (M' F) .08 (MF) .004 (M>F) (.01 I93 .50 (M) F) .88 (M< F) N.S. I94 .13 (M>F) .37 (M>F) N.S. I95 .51 (MF or MF) N.S. E77 .01 (M< F) .10 (M< F) (.01 E78 .05 (MF) N.S. E83 (.0005 (MF) (.00025 (M 7F) (.00025 185 (.0005 (M >F) <. 0005 (M >F) < .0005 I87 .41 (M4 F) .40 (M71?) N.S. I89 .02 (M >F) .001 (M7F) (.01 I91 .07 (M< F) .32 (M7F) N.S . 192 (.0005 (M? F) (.00025 (M>F) (.0005 I93 .37 (M7F) .08 (M>F) N.S. I94 .03 (M) P) 4.00025 (M >F) (.01 I95 .91 (M-F) .17 (M7F) N.S. Criterion: P1£.20, P2 $.20, Compound £3.05. MF indicates direction of mean difference on a particular item. Nete: Samples 1 and 2 can be considered original and cross-validation samples, respectively. 26 Table 13. Job orientation item means by sex in two samples (job level not controlled). Sample 1 SampleRI (N - 842) (N - 804) Item Male Female ‘Male Female E73 4.36 4.54 4.38 4.63 E75 4.10 4.10 4.20 4.12 E77 4.21 4.35 4.24 4.32 E78 4.27 4.38 4.27 4.43 E81 4.34 4.40 4.38 4.37 E83 3.85 4.16 3.85 4.11 E84 3.77 4.01 3.84 3.94 I74 3.94 3.61 3.98 3.64 175 3.74 3.44 3.80 3.34 187 3.90 3.96 3.86 3.80 189 3.86 3.69 3.84 3.57 I91 4.28 4.38 4.32 4.26 192 3.73 3.44 3.71 3.25 193 3.82 3.75 3.80 3.67 I94 3.78 3.63 3.90 3.59 195 3.84 3.84 3.98 3.88 a ll T‘lnmwm—r wan—1 7 Note: Samples 1 and 2 can be considered original and cross-validation samples, respectively. 27 has found significantly related to sex, also emerged as significant discriminators in this study, in the predicted direction. To summarize the pattern of change of item by sex differences from the uncontrolled job level to the controlled job level situation, (age and education controlled throughout) note the following: With job level controlled, males and females differed significantly (compound probability) on only 2 of the 16 items (12.51). With job level uncontrolled, males and females differed on 9 of the 16 items (562). Both the items on which the sexes differed with job level controlled were also significant in the uncontrolled situation. They both decreased their compound probabilities from controlled to uncon- trolled job level situations. That is, the probability of getting a sex item difference due to chance was greatest when job level was controlled. Hypothesis 6 Tables 14-21 contain information from the analyses of variance for the relationship of age and education to job orientation (intrinsic and extrinsic), and the corresponding means. Within the rank-and-file sample, Tables 14 and 16 indicate that the only significant relationship to emerge was that between education and extrinsic job orientation. No significant interactions occurred. Table 17 presents the extrinsic job orientation means by age and education within rank-and-file. The marginal means for education indicate a curvilinear relationship, with the medium education group placing greater importance on extrinsic factors than either the low or high education groups. 28 Within the supervisory sample, several significant relationships emerged. Table 18 shows that age is significantly related to intrinsic job orientation. Table 19 indicates that this significant relationship is linear and positive. That is, the older supervisory group tended to place greater importance on intrinsic job factors than did the younger groups of supervisors. Looking at Table 20, it is clear that both age and education are significantly related to extrinsic job orientation. The marginal means in Table 21 indicates that education is related to extrinsic job orientation in a linear, negative fashion. That is, the lower the educational level, the greater the importance placed on extrinsic job factors. Similarly, age is related in a linear, negative way to extrinsic job orientation. The younger the supervisory group, the greater the importance placed on extrinsic job factors. 29 Table 14. Summary of analysis of variance for the relationship of age (A) and education (B) to intrinsic job orientation within rank-and-file. Source df MS F A 2 .48 1.03 B 2 .86 1.82 A x B 4 .67 1.43 Error 1191 .47 Table 15. Mean intrinsic job orientation by age and education within rank-and-file.* Education Low Medium High Low 3.84 3.84 3.84 3.84 JMHL Medium 3.69 3.67 3.98 3.78 High 3.83 3.75 3.88 3.82 3.79 3.76 3.90 *See data coding section for meaning of levels. Table 16. Summary of analysis of variance for the relationship of age (A) and education (B) to extrinsic job orientation within rank-and-file. Source df MS F A 2 .12 .46 B 2 .84 3.18* A x B 4 .18 .67 Error 1191 .26 *P‘(.OS Table 17. Mean extrinsic job orientation by age and education within rank-and-file.* (Egpggtion. Low Medium High Low 4.20 4.30 4.28 4.26 Agg_ Medium 4.22 4.25 4.28 4.25 High 4.32 4.34 4.15 4.27 4.25 4.30 4.24 *See data coding section for meaning of levels. 31 Table 18. Summary of analysis of variance for the relationship of age (A) and education (B) to intrinsic job orientation within supervisory. Source df MS F A 2 1.23 5.82* B 2 .32 1.50 A x B 4 .47 2.21 Error 67 .21 *p-.005 - Table 19. Mean intrinsic job orientation by age and education within supervisory.* Egpcatiop Low Medium High Low 4.21 3.34 3.88 3.81 App Medium 3.38 3.91 4.16 3.82 High 4.12 4.01 4.12 4.09 3.90 3.76 4.06 *See data coding section for meaning of levels. 32 Table 20. Summary of analysis of variance for the relationship of age (A) and education (B) to extrinsic job orientation within supervisory. Source df MS F A 2 1.04 3.49* B 2 2.02 6.80** ‘ A x B 4 .22 .73 Error 67 .30 i *p (.05 Table 21. Mean extrinsic job orientation by age and education within supervisory.* Education Low Medium High Low 4.37 3.91 3.71 4.00 4.11 4.05 3.78 *See data coding section for meaning of levels. DISCUSSION flypotheses l and 2 In testing hypotheses l and 2, males were higher than females on ,1; intrinsic job orientation. Females were higher on extrinsic job { orientation than males. Age and education were controlled; job level was not controlled. When job level, age, and education were controlled, males and females did not differ significantly on either type of job orientation. These findings are consistent with previous research by Saleh and Lalljee (1969). It is recalled that part of their study dealt with a sample of a "service-oriented organization," in which it was demon- strated that males typically had more education and higher jobs in the organization. No analysis of age differences was offered. They found males to be higher on intrinsic job orientation than females, using ipsative scale measurements, and concluded that education and job level were contaminating variables. It would appear from this study's data that job level, by itself, is sufficient to contaminate sex differences in job orientations for the reason that sex differences in job orientation emerged when job level was the only uncontrolled variable. Sexes were similar in age and education. One approach to explaining why males and females do not differ in job orientation, with extraneous variables like age, education, and 33 34 job level controlled, is that of social role. Given the assumption that social role effects job orientation, role differentiation between men and women in the world of work may be decreasing. Hence, males and females may view their jobs from the perspective of the same social role. Hypothesis 3 In testing hypothesis 3, supervisors were lower on extrinsic job orientation than rank-and-file. Supervisors and rank-and-file did not differ significantly on intrinsic job orientation. These results are not entirely consistent with previous research in the area (Saleh and Lalljee, 1969; Centers, 1948; Centers and Bugental, 1966; Gurin, et al., 1960). To explain this discrepancy with previous research findings, the method of measurement should be examined. Saleh and Lalljee (1969), and Centers and Bugental (1966) both used ipsative measurement scales, in that the intrinsic and extrinsic scales were dependent, i.e., a high score on one necessitates a low score on the other. That is to say, Saleh and Lalljee (1969) only reported an intrinsic score, perhaps realizing that an extrinsic score would have been dependent and therefore meaningless for interindividual comparisons. The higher supervisory group mean in their study could have resulted from a lower supervisory concern for extrinsic job factors, and not necessarily from an absolutely greater concern for intrinsic job factors. The nature of the ipsative measurement technique they used could not provide the necessary information to evaluate precisely the importance of intrinsic and extrinsic job factors. Normative measurement techniques, used in the present study, provided 35 such information. Thus, the difference in results may be due to the greater clarity of information provided by normative measurement scales over the less informative ipsative scales. Hicks (1968, p. 168) makes the following relevant points on ip- sative versus non-ipsative measurements: In summary, one may state that scores originally obtained as ipsative measures may legitimately be employed only for purposes of intraindividual comparisons. Normative measures may be employed for either interindividual or intraindividual comparisons. Another very plausible explanation of why the two job levels did not differ significantly on intrinsic job orientation may involve the nature of the subject sample. Specifically, all subjects used in the analysis were from companies operating under the Scanlon Plan, which maintains a philosophy of participative management. One of the central tenets of such a system of participative management is that workers are intelligent people, and that these people know better than anyone else the best way to do their jobs. To quote Fred Lesieur (1958, p. 39), a leading advocate of the Scanlon Plan: . This plan doesn't mean giving people a 'sense of participation'; workers don't want that. This plan means giving them real participation. You will discover that we have no set formula, that the plan has worked in these situations because of the desire of both management and labor to get together and mutually solve problems that will help them be more competitive in this industrial society of ours. It means working with your brain instead of your back. Recognizing the possible influence of the Scanlon Plan in the failure to detect differences in intrinsic job orientations by job level does not invalidate an explanation in terms of activation theory (Scott, 1966). Saleh and Lalljee (1969) invoked this theory 36 to explain job level differences in intrinsic job orientation. Specifically, the intrinsic factors in higher level jobs may provide more stimulation than in lower level jobs. In Scott's words (1966, p. 357): From activation theory it may be anticipated that as more variation is introduced into a repetitive task, the result would be a reduction in habitua- tion and a sustained activation closer to that . required for optimal behavioral efficiency. If r the increase in variation results in an activation level near the individual's characteristic norm or perhaps just above that level so that responding to the task results in consistent shifts back to the norm, increases in performance and positive affect could be expected. From what was discussed earlier regarding the possible influence of the Scanlon Plan, the job levels under study may be offering about the same level of activation. Regarding the difference between job levels in extrinsic job orientation that emerged, it may be useful to understand this phenomenon through the notion of role theory. It may be that as the employee moves up the organization's hierarchy, he acquires a different set of role prescriptions and expectations, set by other supervisors, as well as by superiors. These role prescriptions and expectations, once adopted, may be reflected in differences in job orientation. One possible criticism of these results and interpretations of job level differences is that they may be due to greater age and education levels in the supervisory groups. And there is evidence from the correlational matrix in Table 23 that the job levels do in fact differ significantly on age and education in the stated direction. The corresponding t tests for job level differences on age and 37 education are 6.15 (p‘<5001) and 8.71 (P <3001), respectively. How- ever, to meet this criticism, the effects of age and education were partialed from the correlations between job level and extrinsic job orientation. The relationship was still significant in the same direction (r - .11, p<.001, N - 1276). Similarly, when age and education were partialed from the rela- tionship between job level and intrinsic job orientation, the relation- ship remained non-significant (r - .02, n.s.). See Appendix E for inter-correlations among age, education, job level, sex, intrinsic and extrinsic job orientation. Thus, the job level difference cannot be simply attributed to age and education differences. Of course, future research may wish to control the effects of other variables on which the job levels may differ (e.g., tenure, number of jobs held, etc.) and then look at whether or not the job levels still differ significantly in job orien- tation. Hypotheses 4 and 5 In testing hypotheses 4 and 5, it will be recalled that when job level was not controlled, males and females differed on nine job orientation items, in a consistent direction. Males were higher on intrinsic items, and females were higher on extrinsic items. But when Job level was controlled, the sexes differed on only two intrinsic items; the direction was still consistent. The reader should be reminded at this time that due to the corre- lational design of the present study, the nature and direction of causation cannot be pinpointed. But it may be fruitful for future research in this area if some speculation on the influence of the 38 job level factor is presented. It would seem that a great deal of the varianCe in job orienta- tion is accounted for by the job level factor. The reader is reminded, however, that the samples of men and women at the supervisory job level may not be representative of males and females in general in the work situation. That is, the female supervisors may differ markedly from the female rank-and-file on variables other than job level (e.g., aggressiveness, need for independence, etc.). Support for the notion of the atypical nature of the supervisory females may be inferred from the fact that very few women, relative to their numbers in the organizations studied, are found at higher level posi- tions. Of those 2755 people who responded to the "You and Your Job" questionnaire, 51 females occupied supervisory or higher level jobs as compared to 332 males. Further inspection of the overall sex by job level distributions for all 2755 indicated that most of the 51 females occupied the lowest supervisory level jobs. In addition, the percentage of questionnaire respondents who were female was about 501, indicating that males and females at each job level are not proportion- ate to their total numbers in the organizations. All this serves to bring out the point that the barriers to female employment above the rank-and-file level may be stronger and more difficult to overcome than they are for males. Hence, females who do manage to get to supervisory level jobs may differ more markedly from the rank-and-file females on other variables than corresponding supervisory males differ from rank-and-file males. As a consequence, it may be argued that job level as a factor is not what is accounting for the variance in job orientation, but that other traits may be 39 producing the variance (e.g., aggressiveness, or even physiological differences). It is beyond the scope of this thesis to pinpoint precisely the causal relationship involved. A possible explanation of why males placed greater importance on "having good chances for promotion" (I85), and "being able to decide how to do your job" (I92) (with age, education, and job level controlled) is that of differing expectations. The females may expect, perhaps from a knowledge of organizational reality which says females are not frequently found in higher level jobs, that a job as supervisor is the best they can hope for in the organization. In ‘4' other words, if promotions for females are rare and limited, the females may be reducing cognitive dissonance by placing less importance on promotions. The same line of thought may be used to explain why females placed less importance on "being able to decide how to do your job." Of course, the theory that biological bases exist for males being "task-oriented" and females being "social need" oriented can be neither supported nor refuted on the basis of the data in this study. Nor can an environmental-learning approach to sex differences be supported or refuted by the correlational data presented here. Unfortunately, the design of this study was not an experimental one. Mhny variables were uncontrolled, and causation impossible to establish definitely. It is difficult, indeed, to assign people randomly to sex and job level teatments. Thus, as is usually the case with data on complex, human variables, it is wise to acknowledge mul- tiple determinants of behavior, both biological and social learning factors. 40 Hypothesis 6 In exploring hypothesis 6, it was found that education.was significantly related to extrinsic job orientation within the rank- and-file sample. Within the supervisory sample, age was related significantly to both intrinsic and extrinsic job orientation. Edu- cation was related significantly to only extrinsic job orientation. Why these particular relationships emerged is not known. But the fact that the pattern of relationships changed from rank-and-file to supervisory job levels indicates that in future research, the moderating effects of job level should be more thoroughly investigated. These moderating effects should especially be taken into account when dealing with relationships involving job orientation. anegal Digguggiog In this section, certain statements of limitation will be made which relate to the whole study, and which should be taken into con- sideration before attempting to generalize to other populations. ’Pirst, it should be remembered that in all the analyses, only respondents who had complete data were used. It can be argued, or at least suggested, that those people who did not fully complete the relevant questionnaire items differed systematically from those who did complete the questionnaire. Whether and how these two groups of respondents differ is not known. It was reasoned at the beginning of this study that using only complete data respondents would yield more reliable and stable results. Second, the type of respondent involved in this study was some- what unique for several reasons. They were all employed at medium size manufacturing plants in close geographic proximity in the Midwest, L—mm : Eli 41 in relatively small cities. Also, each of the six companies was operating under the Scanlon Plan, which maintains a philosophy of par- ticipative management. It has already been suggested how the Scanlon Plan may have affected the relationship between job level and job orientation. How it may have affected the other relationships under study is not known, and is a matter of speculation. Finally, it should be acknowledged that many of the significant relationships that emerged involved relatively small mean differences, which may have little value in terms of accurate prediction. However, the aim of this thesis was not to develop practical instruments for 'T‘m‘ unu- max 1 .1 prediction, but rather to investigate relationships observed in previous research, to elaborate upon those variables, and observe new, more complicated relationships. BIBLIOGRAPHY BIBLIOGRAPHY Baker, P.C. Combining tests of significance in cross-validation. Edppatioppl apd ngghological Mgaguyegept, 1952, 1;, 300-306. Burke, R. Differences in perception of desired job characteristics of the opposite sex. e o l o e i P 010 , 1966, 122, 27-37. (a) Burke, R. Differences in perception of desired job characteristics of the opposite sex. The Jourppl of Qgpetic Psyghology, 1966, 102, 37-46. (b) Centers, R. Motivational aspects of occupational stratification. Jouypgl 2f §ocia1 Psychology, 1948, 28, 187-217. Garai, J. B., and Scheinfeld, A. Sex differences in mental and behavioral traits- WM. 1968. 21. 169-299. I "H‘ALS-Aun”. .1...- . one- u? Guilford, J. R. ngghometrig Methodg. New York: MeGraw-Hill, 1954. Hardin, B., Reif, H., and Heneman, H. Stability of job preferences of different segments of the working population. Jopypp; of Applied Ppychology, 1951,‘;2, 256-259. Herzberg, F., Mausner, B., and Snyderman, B. The Motivation To WOrk. John Wiley and Sons, Inc., 1959. Hicks, L. E. Some properties of ipsative, normative, and forced-choice normative measures. ngghplpgy_fipllg£ip, 1970, 15, 167-184. Jurgenson, C. Selected factors which influence job preferences. Joprppl pf Applied ngghplogy, 1947, 1;, 553-564. Kulen, R. Needs, perceived need satisfaction Opportunities, and satisfaction with occupation. Jopyppl pf Applied ngcholggy, 1963, 41, 56-64. Lesieur, F. 0. (ed.) The Spaplon Flap; A [poppier ip Laboy-ngnagemepg goopeyatiop. Cambridge, Mass.: The M.I.T. Press, 1958. Saleh, S. A study of attitude change in the pre-retirement period. Jopyppl of Applied Ppyghology, 1964, 48, 310-312 42 43 Saleh, S., and Lalljee, M. Sex and job orientation. Peypopnel Psyghology, 1969, 2;, 465-471. Scott, W. B., Jr. Activation theory and task design. a ati fighayipr app Hagen Perfoygppge, 1966, 1, 3-30. In Cummings, L. L., and W. E. Scott, Reading; ip gyganigatioppl Behayior agd Hyman Perfoymapp . Homewood, 111.: Dorsey, 1969. Winer, B. J. Statiptipal gyippipleg in Expegigepgal Depigp. New York: McGraw-Hill, 1962. APPENDICES APPENDIX A SORTING INSTRUCTIONS AND RESUETS OF SORTINC Item Sortipg Ipstrpgtiops to Five Gradpate Stpdept Judge; You are asked to sort the accompany deck of 3"x5" cards into three categories. The first category will be for those cards containing items that relate to Herzberg's intrinsic job factors (or "motivators," "satisfiers"), which he defines as "factors responsible for good feelings about the job," and which "relate to the doing of the job itself or to the intrinsic content of the job rather than to the context in which the job is done" (e.g., "factors of achievement, responsibility, work itself, advancement, and recog- nition"). The second category of cards will contain items relating to Herzberg's extrinsic factors (or "hygiene factors," "dissatisfiers"), which he defines as factors "that define the job context" (e.g., "company policy and administration, technical supervision, inter- personal relations--with peers, subordinates, and superiors working conditions, and salary"). Finally, the third category will contain cards with items that relate to peitgey of the above two categories. 44 Res ta 0 13.9.1.1 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 9O 91 ive APPENDIX A - continued a ate latrines; 11111 1111 11111 11111 11111 45 de t d e e Egtripgig 11111 11111 11111 11111 111 11111 11111 1111 Neither 1111 1111 1111 1111 11111 Item f 92 93 94 95 96 46 APPENDIX A - concluded 1mm 11111 1111 11111 11111 1 Extginpig 11 Neithe; 11 IFMJm‘l- APPENDIX B JOB ATTITUDE ITEMS AND INSTRUCTIONS TO SUBJECTS A set of JOB CONDITIONS or GOALS which people can try to - achieve on their jobs are listed below. Briefly look over this I list and then continue reading these instructions. It is obvious that different people will disagree about the 1.04; am- A _ relative importance of these goals even though all of them'will be in some sense important to just about everyone. We are inter- ested in finding out the RELATIVE importance of each of these goals FOR YOU. We do pp£_want to know what your job is like now. Rather, we would like to know what you ypp£,it to be like. In order to indicate their relative importance for you, would you please choose the alternative from the list below which best describes how important each goal is for you. Remember, we are interested in the diffigyeppes between the goals in importance to you. It is very important therefore that you do not use any one alternative too many times. In order to help us tell the differences between the goals, you should try not to use any one alternative more than five times. In other words, you should not rate any more than five goals in the same corresponding letter to the right of each statement. a. extremely important b. very important c. somewhat important d. slightly important e. not at all important 47 73. 74. 75. 76. 77. 78. 79. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 48 APPENDIX B - continued Hm IMPORTANT TO YOU 18: having a supervisor who really knows his job? having a challenging job? having relations of trust and confidence between superiors and subordinates? helping the company make high profits? being trusted by the people you work with? receiving fair pay? everyone in the company, no matter how low his status level, having a say in the decisions affecting his job? working for a socially responsible company (one that helps make your community a better place to live)? steady work and steady wages? helping the company be prepared for change (technological change, product change and improvement)? working under good (safe, clean, pleasant) conditions? being liked by the people you work with? having good chances for promotion? helping the company grow and expand? being recognized and appreciated for doing good work? helping the company build and maintain customer confidence and good will? having responsibility on your job? doing work which helps the welfare of others? turning out high quality work? I 92. I 93. I 94. I 95. 96. 49 APPENDIX B - concluded being able to decide how to do your job? a feeling that the work you are doing is important? being able to learn new skills and gain experience on your job? a sense of achievement in the work you are doing? not having to work too hard? E - Extrinsic Job Orientation Item I - Intrinsic Job Orientation Item ‘- "N. L.._._.— l *4‘. APPENDIX C SUMMARY OF ANALYSES OF VARIANCE FOR THE RELATIONSHIP OF SEX (A) TO EACH OF THE 16 JOB ORIENTATION ITEMS, IN ORIGINAL AND CROSS-VALIDATION SAMPLES, WITH AND WITHOUT JOB LEVEL CONTROLLED Table 22. Summary of analyses of variance for the relationship of sex (A) and job level (B) to each of the 16 job orientation items for sample 1 (job level controlled). '9._‘-ale..a-|Qu_" - W "" influx; Item Source df MS F 9 E73 A 1 .85 1.16 .28 B l 2.51 3.43 .06 A x B 1 .00 .00 .93 Error 634 .73 E75 A 1 .09 .11 .74 B 1 .36 .45 .51 A x B l .42 .53 .47 Error 634 .80 E77 A 1 .51 .66 .42 B 1 1.79 2.33 .12 A x B l .01 .01 .89 Error 634 .77 E78 A 1 .06 .09 .76 B 1 4.74 7.55 .01 A x B l .16 .25 .62 Error 634 .63 E81 A 1 .27 .37 .55 B 1 3.43 4.60 .03 A x B l .01 .02 .87 Error 634 .75 50 51 APPENDIX C - continued Item Source df MS F p E83 A .04 .04 .82 B 11.07 11.75 .001 A x B .27 .28 .60 Error .94 E84 A .90 .85 .36 B 3.78 3.58 .06 A x B .10 .09 .76 Error 1.06 174 A .03 .03 .83 B 2.90 2.94 .08 A x B 3.07 3.12 .07 Error .98 I85 A 2.84 1.87 .17 B .96 .63 .43 A x B .18 .12 .73 Error 1.52 187 A 1.26 1.10 .30 B 2.22 1.93 .16 A x B .58 .51 .48 Error 1.15 189 A .09 .08 ' .77 B 9.81 8.76 .003 A x B .09 .08 .77 Error 1.12 I91 A .00 .00 .95 B 2.16 2.89 .09 A x B .10 .13 .71 Error .75 IFK'M A ”‘1 — ... 52 APPENDIX C - continued Item. Source df MS F p 192 A 1 2.71 2.00 .15 B l 10.80 7.95 .005 A x B l .50 .36 .55 Error 634 1.36 I93 A l .59 .48 .50 B 1 1.12 .90 .35 A x B 1 .03 .02 .85 Error 634 1.25 194 A l 3.12 2.23 .13 B 1 .71 .50 .48 A x B 1 .56 .40 .53 Error 634 1.40 I95 A 1 .53 .45 .51 B 1 .74 .63 .44 A x B l .72 .61 .44 Error 634 1.18 . «f. r. .. 53 APPENDIX C - continued Table 23. Summary of analyses of variance for the relationship of sex (A) and job level (B) to each of 16 job orientation items for sample 2 (job level controlled). Item. Source df MS F 9 E73 A 3.90 6.67 .01 B 1.31 2.24 .13 A x B .74 1.26 .26 Error .58 £75 A 3.02 3.64 .05 B .35 .42 .52 A x B 4.15 4.99 .02 Error .83 E77 A .00 .00 .91 B .05 .07 .78 A x B .13 .18 .67 Error .74 E78 A .25 .35 .56 B 3.38 4.82 .03 A x B .03 .05 .81 Error .70 E81 A 3.89 5.78 .02 B 2.25 3.34 .06 A x B 5.91 8.77 .003 Error .67 E83 A 3.23 3.39 .06 B 8.41 8.82 .003 A x B .52 .54 .47 Error .95 54 APPENDIX C - continued Item Source df MS F p 384 A .44 .44 .52 B 3.67 3.67 .05 A x B .08 .08 .77 Error 1.00 I74 A 1.54 1.78 .18 B 2.14 2.49 .11 A x B .09 .10 .75 Error .86 I85 A 5.21 4.02 .04 B .30 .23 .63 A x B .30 .23 .64 Error 1.29 I87 A .26 .26 .61 B .01 .01 .09 A x B 1.14 1.16 .28 Error .98 I89 A .26 .22 .64 B .34 .29 .60 A x B .68 .58 .45 Error 1.17 I91 A 2.09 2.99 .08 B .87 1.24 .27 A x B 1.92 2.75 .09 Error .70 192 A 11.46 8.18 .004 B .05 .04 .82 A x B 1.78 1.27 .26 Error 1.40 55 APPENDIX C - continued Item Source df MS F p 193 A 1 .01 .01 .88 B 1 .06 .05 .80 A x B 1 .26 .22 .64 Error 634 1.18 I94 A 1 1.13 .81 .37 B 1 1.07 .76 .39 A x B l .19 .13 .71 Error 634 1.40 ‘ 195 A 1 .42 .41 .53 B l .01 .01 .88 A x B 1 .86 .84 .36 Error 634 1.02 56 APPENDIX C - continued Table 24. Summary of analyses of variance for the relationship of sex (A) to each of the 16 job orientation items for sample 1 (job level not controlled). ,1 Item Source df MS F p E73 A l 6.66 9.34 .002 Error 840 .71 L. E75 A l .01 .01 .89 ‘ Error 840 .88 E77 A l 4.27 5.78 .02 Error 840 .74 E78 A 1 2.65 3.65 .05 Error 840 .72 E81 A l .82 1.01 .32 Error 840 .82 E83 A 1 19.71 19.10 .(.0005 Error 840 1.03 E84 A l 11.87 10.61 .001 Error 840 1.12 174 A l 23.27 24.08 (.0005 Error 840 .97 185 A l 18.42 13. 11 < .0005 Error 840 1.41 187 A 1 .70 .71 .41 Error 840 .99 APPENDIX C - continued Item Source df MS F p :1 189 A 1 6.16 5.32 .02 Error 840 1.16 191 A l 2.21 3.25 .07 Error 840 .68 192 A l 18.12 12.47 .001 L Error 840 1.45 193 A 1 1.01 .81 .37 Error 840 1.24 194 A 1 5.28 3.66 .05 Error 840 1.44 195 A 1 .004 .003 .91 Error 840 1.17 58 APPENDIX C - continued Table 25. Summary of analyses of variance for the relationship of sex (A) to each of the 16 job orientation items for sample 2 (job level not controlled). 1 Item Source df MS F p . E73 A l 11.72 17.63 (.0005 Error 802 .67 2 3 E75 A 1 1.28 1.79 .18 (N.S.) 1 Error 802 .72 E77 A l 1.26 1.56 .21 Error 802 .81 E78 A 1 5.15 7.26 .007 Error 802 .71 E81 A 1 .02 .02 .85 Error 802 .78 E83 A l 13.56 13. 15 ( .0005 Error 802 1.03 E84 A 1 2.03 1.99 .16 Error 802 1.02 174 A 1 23.87 26.51 .(.0005 Error 802 .90 185 A 1 42.66 31.06 ( .0005 Error 802 1.37 187 A l .85 .73 .40 Error 802 1.16 59 APPENDIX C - concluded Item Source df MS , F p 189 A 1 13.99 12.11 .001 Error 802 1.16 191 A 1 .78 1.00 .32 Error 802 .77 192 A l 42.34 31.43 .(.0005 Error 802 1.35 193 A l 3.54 2.93 .08 Error 802 1.21 194 A l 20.41 15.31 (.0005 Error 802 1.33 I95 A l 1.97 1.85 .17 Error 802 1.06 Table.26. Item E73 E75 E77 E78 E81 E83 E84 I74 185 187 189 I91 I92 I93 I94 I95 N I 1647 28 22 26 23 3O 16 O4 13 02 14 O9 O9 11 O4 31 14 14 21 23 17 16 14 14 16 21 16 18 15 17 20 3O 43 12 18 16 18 19 25 22 22 18 37 32 18 05 20 21 00 O6 16 O7 11 11 It] I - Intrinsic Item E - Extrinsic Item 31 19 -Ol 10 16 11 17 11 O7 07 It] APPENDIX D 34 05 21 28 13 17 25 23 18 16 Ill 10 28 27 27 16 28 27 24 23 ill 60 33 14 33 13 26 28 32 31 sf (s H 33 15 33 32 41 28 If) H 28 24 32 33 26 30 h H 27 37 39 37 31 G H 24 26 22 29 F! O\ H 46 41 32 H bu Mun- I93 U “H 194 Correlations among Herzberg's 16 job orientation items. 195 p- ...,. .11-7.: mink} TIV APPENDIX E Table 27. Correlations among age, education, job level, intrinsic orientation, extrinsic orientation, and sex. .. fiWI‘ .1 (1) Age 1.0 (2) Education -.30 1.0 (3) Job Level -.12 -.14 1.0 (4) Intrinsic O. -.07 .09 -.02 1.0 (5) Extrinsic O. -.03 -.09 .12 -.43 1.0 (6) Sex .01 -.03 .00 .11 -.07 1.0 (l) (2) (3) (4) (5) (6) u - 1276, a"; - .029 NOte: Male - 1, Female - -1 Intrinsic 0. - intrinsic orientation Extrinsic 0. - extrinsic orientation 61 APPENDIX F SAMPLE COMPOSITIONS AND T VALUES FOR SEX DIFFERENCES ON AGE, EDUCATION, AND JOB LEVEL Sample Composition for Analyses with Job Level not Controlled Tots Sa 1e I '| F‘Ll“"~v“ll ml' uflrmflfi ' 3 Supervisory males I 210 Supervisory females I 38 Level 6 I 73 Level 6 I 5 Level 7 I 92 Level 7 I 20 Level 8 I 45 Level 8 I 13 Rank-and-file males I 616 Rank-and-file females I 782 Sa e 1 I 2 Supervisory males I 96 Supervisory females I 19 Level 6 I 28 Level 6 I 2 Level 7 I 48 Level 7 I 10 Level 8 I 20 Level 8 I 7 Rank-and-file males I 316 Rank-and-file females I 411 a la I Supervisory males I 114 Supervisory females I 19 Level 6 I 45 Level 6 I 3 Level 7 I 44 Level 7 I 10 Level 8 I 15 Level 8 I 6 Rank-and-file males I 300 Rank-and-file females I 371 62 63 APPENDIX F - continued Table 4. t values and means for sex differences on age, education, and job level in samples in which job level was not controlled. Sex Ma le Female t (Ii-826) (NI820) Age 'iI33.99 sIll.89 ‘RI33.26 sI12.55 1.22 Total Sample Education 'i- 6.43 sI 1.84 ‘!I 6.35 sI 1.52 .90 (NI1646) Job Level RI 8.46 sI 1.00 XI 8.92 sI .40 12.20* (NI412) (“'430) Age xI34.13 sIll.82 xI33.05 sI12.29 1.30 Sample 1 (NI842) Education '!I 6.38 sI 1.89 'ii 6.24 sI 1.60 1.13 Job Level ‘EI 8.51 sI .95 'EI 8.92 sI .38 8.29* 01-411.) (II-390) Age 'fiI33.86 sIll.96 'i333.49 sI12.84 .42 Sample 2 (NI804) Education ‘53 6.47 sI 1.80 '53 6.47 sI 1.43 .01 Job Level ‘ ‘!I 8.40 sI 1.05 '15 8.91 sI .42 8.90* *P4(.0005 64 APPENDIX F - continued Sample Composition for Job Level Control Analyses Total Sagple (N I lgzgj Supervisory males I 38 Supervisory females I 38 Level 6 I 5 Level 6 I 5 Level 7 I 20 Level 7 I 20 Level 8 I 13 Level 8 I 13 Rank-and-file males I 600 Rank-and-file females I 600 a e I 6 Supervisory males I 19 Supervisory females I 19 Level 6 I 2 Level 6 I 2 Level 7 I 10 Level 7 I 10 Level 8 I 7 Level 8 I 7 Rank-and-file males I 300 Rank-and-file females I 300 a e I 8 Supervisory males I 19 Supervisory females I 19 Level 6 I 3 Level 6 I 3 Level 7 I 10 Level 7 I 10 Level 8 I 6 Level 8 I 6 Rank-and-file I 300 Rank-and-file females I 300 65 APPENDIX F - concluded Table 5. t values and means for sex differences on age, education, and job level in samples in which job level was controlled. Sex Male Female t (NI638) (NI638) Age YI33.05 sI12.06 1333.21 sI12.35 .23 Total Sample Education 11'- 6.22 sI 1.81 Y- 6.32 sI 1.53 .00 (NI1276) Job Level Tc'I 8.89 sI .45 II 8.89 sI .45 .00 (ta-319) (N-319) Age 31331.87 sIll.88 7.32.20 sIll.52 .36 Sample 1 (NI638) Education 51'- 6.33 sI 1.79 SEI 6.34 sI 1.57 .05 Job Level 3?- 8.90 sI .44 11'- 8.90 sI .44 .OO (N-319) (NI319) Age YI34.24 sI12.15 51334.22 sI13.06 .03 Sample 2 (NI638) Education YI 6.12 sI 1.83 YI 6.30 sI 1.48 1.38 Job Level i- 8.89 sI .47 ‘x"I 8.89 sI .47 .00 ....... HICHIGQN STRTE UNIV. LIBRRRIES 31293102179730