Hulin and Blood (1968) suggested that conflicting findings relating to reactions to job enrichment could be reconciled if alienation from middle-class work norms was considered as a moderator. Wanous (1974), using the pro-Protestant Ethic items of the Blood (1969) scale, found adherence to Protestant Ethic ideals to moderate task characteristic-satisfaction relationships as expected, though the moderating effects were not as strong as those evidenced by higher order need strength. Relationships between job behaviors and task characteristics were not, however, moderated by adherence to Protestant Ethic ideals. Merrens and Garrett (1975) administered the Protestant Ethic Scale developed by Mirels and Garrett (1971) to introductory psychology students and formed subgroups on the basis of those scores. The high Protestant Ethic group was found to spend more time on a subsequent task and to produce greater output.
Given the apparently key role played by work values in influencing employee affective responses to work in general and to specific task characteristics, this study attempted to replicate the Blood (1969) findings using a markedly different sample and different measures of affective responses. Furthermore, while correlations of scores on the Protestant Ethic Scale with such variables as sex guilt, conscience guilt, authoritarianism, locus of control, and social desirability have been reported (Mirels and Garrett, 1971), relatively little is known about the relationships of adherence to Protestant Ethic ideals with task and leader dimensions. Consequently, correlations between work-value indices and employee perceptions of task dimensions and of leader behavior were examined. Information concerning these latter correlations is important if work-value indices are to be viewed as moderators of task-dimension-employee-response or leader-behavior-employee-response relationships. Finally, in an exploratory vein, this study examined whether work values were associated with biographical measures or with particular individual needs, traits, or abilities.
Subjects of the study were hourly employees of a manufacturing firm in a mid-western state. A large number of position titles were represented (for example, crane operator, furnace tender, die caster, material handler, mechanical finisher, heat treater, welder). Questionnaires and a follow-up letter were mailed to all hourly employees. Of 415 questionnaires mailed, 131 were returned by the time of the analysis. Completed questionnaires were sent directly to the researchers and were anonymous.
To gauge work values, the Blood (1969) scale was used. That scale includes four items intended to be in agreement with Protestant Ethic ideals (pro-Protestant Ethic) and four not agreeing with those ideals (non-Protestant Ethic). For the data of the current study, principal components analysis of the eight items with varimax rotation of two components yielded the loadings, which are very similar to those reported by Blood. The two factors accounted for 21.1 per cent and 19.2 per cent of the common variance, respectively.
The questionnaire included a slightly revised version of the Yale Job Inventory used by Hackman and Lawler (1971). This version has been used by Lawler, Hackman, and Kaufman (1973), and by Brief and Aldag (1975), and is reported on by Hackman (1973). Eight affective response measures were taken from the Yale Job Inventory (job involvement, internal work motivation, general satisfaction, hygiene satisfaction, existence satisfaction, supervisory satisfaction, growth satisfaction, and peer satisfaction).
Five core task dimensions were also gauged by the Yale Job Inventory (skill variety, task identity, task significance, autonomy, and feedback from the job itself). Evidence relating to reliability and validity of the task dimension measures is provided by Hackman and Lawler (1971) and by Brief and Aldag (1975).
The leader Behavior Description Questionnaire (LBDQ) was used to gauge leader consideration and initiating structure. Evidence is available to support the reliability and validity of some forms of the LBDQ (cf. Schriesheim and Kerr, 1974; Stogdill, 1969).
Individual abilities, traits, and needs were measured by the Ghiselli Self-Description Inventory, a 64-item forced-choice inventory (Ghiselli, 1971). That scale gauges 13 individual characteristics (for example, initiative, maturity, and need for job security). Also, two measures (Measures A and B) of higher order need strength were taken from the Yale Job Inventory. While Measure A is an absolute gauge of higher order need strength, Measure B is a gauge of higher order need strength relative to lower order need strength.
Employee age, educational level, tenure in organization, time on current job and urbanization of area of socialization and of current home location were also measured. The latter pair of measures were scored as urban, suburban, or rural.
Most correlations of the pro-Protestant Ethic score with affective responses are positive, with two reaching significance. Negative relationships are evident between the non-Protestant Ethic score and all affective response measures, with four of the correlations being significant. Thus, the Blood (1969) findings are generally replicated.
While perceptions of task dimensions are essentially independent of the pro-Protestant Ethic score, the non-Protestant Ethic score is significantly negatively related to all perceptions of task dimensions.
Somewhat similar findings are evident in relation to consideration. That is, perceptions of leader's consideration are significantly negatively related to the non-Protestant Ethic score.
The only significant biographical correlate of the pro-Protestant Ethic score was age (r=.305, p<.001). No relationships of the non-Protestant Ethic score to biographical indices were significant.
Only 2 of 26 correlations of work-value indices with the characteristics measured by the Ghiselli Self-description Inventory were significant. Those two significant correlations were between the non-Protestant Ethic score and masculinity-femininity (r = -.207, p < .05) and maturity (r = -.230, p < .05).
Summary and Discussion
The findings of Blood (1969) regarding relationships of work-value indices to employee affective responses were replicated using a different sample and different affective response measures. Further, numerous, significant relationships of the indices to perceptions of task dimensions and of leader behavior were evident. While correlations of the work-value indices with higher order need strength measures were significant, relationships with other individual difference measures were generally weak.
The significantly positive correlation of the pro-Protestant Ethic score with age is consistent with popular conceptions of that relationship. Similarly, the findings relating to correlations of work values with higher order need strength agree with the arguments and findings of Wanous (1974). However, the suggestion of Hulin and Blood (1968) and of Wanous (1974), that urbanization of area of socialization may be related to adherence to Protestant Ethic ideals, was not supported by this study. It should be noted that while Wanous argued in terms of a developmental sequence from socializing experiences, to work values, to higher order need strength, his own data revealed no significant relationship between the pro-Protestant Ethic score and the degree of urbanization of the area of socialization.
Despite the fact that work-value-affective-response relationships were found in this study to closely approximate those presented by Blood (1969), the overall set of results does little to enhance confidence in the existence of a causal relationship between those variables. In particular, the relationships found between work-value indices and perceptions of task dimensions and of leader behavior are disquieting and potentially important. For example, if an employee's perceptions of the dimensions of his task color his work values and also influence affective responses, spurious work-value-affective-response relationships may result. On the other hand, if as Blood (1969) suggests, "It seems more logical... to assume that work values precede and influence job satisfaction than the opposite", and if those work values also influence perceptions of task dimensions, findings of task-dimension-affective-response relationships may be spurious. Similar potential dangers of spurious results are suggested by the findings relating to consideration.
Thus, while the findings indicate that previously neglected relationships may be relevant, further investigations of causal mechanisms are required. Such investigations must include either longitudinal analysis or objective measures of such variables as task dimensions and leader behaviors. Also, relationships of perceptions of task characteristics and leader behaviors to other work-value measures, such as those of Wollack, Goodale, Wijting, and Smith (1971), should be considered. Finally, since it could be argued that correlations in the current study may have been inflated because of method variance, future studies should be structured so as to minimize such a possibility. Until studies such as those outlined above are completed, adequate determination of the roles of work values is precluded. Consequently, caution should be exercised in the use of such indices and in the interpretation of findings from studies in which work values may be relevant.