An instrumentality-expectancy-task-goal model that sought to add specificity to the conceptualization and measurement of key variables by focusing on employees' decision process was tested in two organizations. Support for the major links in the model was found in one of the organizations, while the results from the other organization generally failed to support the hypotheses. Special attention is given to the potential impact of the organizational environment on employee cognitions and motivation. The differences in results between the two organizations are attributed to possible boundary conditions that may affect the relationships among employee cognitions, task goal selection, and performance. Implications of the basic model and its boundary conditions for research and practice are discussed. It is concluded that the model facilitated the observation of systematic and orderly results, thus adding to the understanding of behavior in organizations.In recent years, research and theory con- (Campbell et al., 1970;Graen, 1969; Mitclk cerning work motivation have given increasing ell & Biglan, 1971). However, a closer examiattention to valence-instrumentality-expec-nation of the research reveals (a) a number tancy (VIE) models (see Campbell, Dun-of inconsistent findings, (b) rather modest nette, Lawler, & Weick, 1970;Graen, 1969; relationships between the theoretical compo-Lawler, 1971;Porter & Lawler, 1968; Vroom, nents and performance, and (c) inadequate 1964). Reviewers have generally concluded support for the hypothesized interaction that empirical tests of VIE models have pro-effects of the theoretical components on bevided moderate support and that therefore havior (Heneman & Schwab, 1972; Miner & VIE theory shows promise of providing a sci- Dachler, 1973;Mitchell, 1972).
entifically useful model of work motivationThis state of VIE theory is at least in part attributable to limitations that have charac-1 This research project was made possible by finan--cial support from the General Research Board and various agencies and companies involved. Apprethe Department of Psychology of the University of elation is expressed to these people who are too Maryland, as well as by funds provided by the Office numerous to mention. We owe a great deal to Tove of Economic Opportunity through Community Action Hammer, Gene Hoffman, Joe Schneider, and Kent Program Grant 2653. Computer time was provided Boyd, all of whom spent long hours in helping overby the Computer Science Center of the University come some of the many problems presented by the of Maryland. This paper emerged out of the frame-scope and nature of this study. The helpful comments work for a larger study of organizational behavior of our colleagues, Ed Locke and Benjamin Schneider, and out of the related doctoral dissertation submitted on earlier drafts of this paper are gratefully acknowlby the second author to the Graduate Faculty of the edged. University of Maryland. The authors have been listed2 Requests for reprints should be sent to-H. Peter alphabetica...