Date of Award
Master of Arts (MA)
James M. Thomas
A cognitive mediation model of goal setting (Garland, 1985; Meyer & Gellatly, 1988) is explored from a perspective that combines the social information processing, expectancy-valence, and control systems literatures. Assigned goals, performance norms, and task experience are viewed as information sources that influence goal choice, and therefore effort and performance, through performance expectancy (expectancy of success) and performance valence (anticipated satisfaction with any given performance level). Subjects were 100 introductory psychology and sociology students who participated in exchange for extra credit. Subjects performed a card sorting task and were assigned to one of five treatment conditions. Each experimental subject was assigned either a difficult or easy goal in conjunction with a high or low performance norm. Control subjects did not receive assigned goals, nor were they provided with normative information. Following a baseline period used to assess ability, all subjects participated in four trials which served as repeated measures of the task experience effect. Contrary to hypotheses, there were no goal or norm main effects on any of the five dependent variables contained in the cognitive mediation model (performance, effort, personal goal, performance expectancy, or performance valence). Goal x experience and norm x experience interactions were anticipated for each of the five dependent variables. However, there were no interactions of norms with experience, and only two of the goal x experience interactions emerged: Relative to easy goals, difficult goals initially were associated with higher personal goals and lower valences, but the differences rapidly diminished. A strong effect of experience on performance was found, with performance increasing in later trials. However, the term "practice effect" may not be applicable because there were strong experience effects on all four antecedent variables: As subjects gained experience, they reported higher performance expectancies, lower valences, higher personal goals, and greater effort. These findings are consistent with a past performance interpretation in which individuals ignore social cues such as assigned goals and performance norms, and set personal goals according to their own previous performance levels. The findings offer limited support for the proposed cognitive mediation model. A revised model in which past performance is substituted for goals and norms would better fit the data.
The notion that challenging goals enhance performance has become more of an axiom than a research question, and researchers have turned to the antecedents of goal acceptance and commitment. But few goal setting studies have addressed the role of experience. A possible explanation o f the present findings is that subjects with experience overlook assigned goals and other social cues in favor of their own previous performance levels. They then form expectancy-valence attitudes which, as in the original model, directly influence personal goals. In turn, personal goals influence effort, which determines performance. The regression results offer some support for this past performance argument and for portions of the proposed cognitive mediation model. However, an essential element of the model was not supported: There was no evidence for the path between personal goal and performance. Without this path, there is no mechanism by which expectancies and valences may influence performance.
Conclusions are at best speculative due to certain limitations of the study. For example, in real work settings there exist strong incentives for accepting challenging assigned goals; the study lacked these incentives. The experimental setting also lacked the social pressures that exist in actual work situations. Suggestions for future research are discussed.
Shanahan, Jon, "Task experience, assigned goals, and performance norms as determinants of goal choice and performance" (1994). Student Work. 174.