Correlations between various worker attitudes and job motivation, performance, and absenteeism were examined for 290 skilled and semiskilled papermakers. The hypothesis that significant relationships occur more frequently for those employees least job involved was supported, inasmuch as they accounted for 84% of the significant correlations. Further analyses disclosed twice as many associations for skilled as for semiskilled employees. Thus, highly involved employees, more intrinsically oriented toward their job, did not manifest satisfactions commensurate with company evaluations of performance; they depended more on intrinsic rewards. Those employees more detached from the job itself were more extrinsic in orientation and experienced gratifications more in line with company performance assessments due to their greater dependence on extrinsic rewards.It has been surprisingly difficult to demonstrate appreciable relationships between work performances and the satisfactions supposedly inherent in them (Ronan, 1970). Despite frequent disappointments, research in this area continues. These efforts seem to be based on the unacceptable premise that although performance and attitude are often unlike, they should not be. Even when similarities are found, they are often barely significant (Vroom, 1964).This study examines the correlational issue further by directing attention to the job incumbent and his work orientation. The theoretical emphasis on intrinsic-extrinsic factors of job satisfaction (Herzberg, Mausner, & Snyderman, 1959) and the proposition that intrinsic and extrinsic rewards serve as satisfaction-performance moderators (Lawler & Porter, 1967) have prompted much needed research on some vital issues. However, the conceptual confusion that results from the finding that job satsfaction can originate from either job-related (intrinsic) or situational (extrinsic) features (Burke, 1966) and the inability to operationalize intrinsic rewards 1 The author would like to express his appreciation to Dennis W. Organ and H. J. Reitz for their insightful comments on an earlier version of this article.

The general canonical correlation index (/?».«*) proposed by Stewart and Love is examined. The index does not fulfill either claim made of it by Stewart and Love: (a) It does not represent a proportion of overlapping, or redundant, variance since the index includes an improper term for the variance of a canonical variate and (6) it is not equal to the average of the squared multiple correlations between a linear composite of one set of variables and the elements of a second set, nor is it equal to the average of the squared multiple correlations between a linear composite of one variable set and the canonical variates of a second set.

No abstract

Since canonical correlation is being increasingly applied in the behavioral sciences, a comprehensive appraisal of its merits is warranted. A survey of the literature employing this technique indicates incomplete use of the method and confusion in canonical terminology. This paper reviews and integrates four analytical procedures that provide interpretive assistance for canonical solutions: (a) testing for significance, (b) estimating stability, (c) naming dimensions, and (d) predicting within dimensions identified. These procedures are used to analyze 309 skilled factory worker job value and perceived job characteristic responses to 35 work referent items. Results show that: (a) only three of the twelve highly significant (p < .01) canonical correlations indicate meaningful underlying constructs, (b) stability estimates are necessary to identify these constructs, (c) only a selected subset of work referent items is crucial to dimensional naming, and (d) prediction is bi-directionally significant (p < .01) within each dimension interpreted. Future canonical applications can provide substantive contributions to the behavioral sciences only with a full appreciation and implementation of these analytical strategies.CANONICAL correlation is a relatively new multivariate tool in the behavioral sciences. Although the mathematical and conceptual developments have been inventive and made readily available through computer programming, the heuristic value of canonical methodology in assessing behavioral data is not yet complete. Some researchers

No abstract