Relationship of Goal Level to Performance Level

Several previous srudies (Locke, 1966) have yielded a positive, linear relationship between the level or difficulty of the goal assigned to S and quantity of output on "brainstorming" tasks. However, the strength of the relationships may have been attenuated in these studies since Ss assigned easy goals usually exceeded these goals. This was prevented in the present study. Fifty-six male and 49 female paid, undergraduate, volunteers worked for 15 1-rnin. trials, the task being to list objects or things that could be described by a given adjective (e.g., "heavy"). The adjectives were different for each trial. Goals were assigned in terms of the n ~ ~ m b e r of things S was to try and list. The goals were: 4 ( N = l l ) , 8 ( N = 17) , 10(N = 15 ) , 12(N = 4 1 ) , and 13(N = 2 1 ) . A given S had the same goal on all 15 trials. Ss were told to stop working if they reached their goal before 1 min. was up. E checked the protocoIs for repetitions but not for quality of answers. Mean output quantity as a fiinction of goal level, and probability of reaching the goal is shown in Fig. 1. The Pearson r between goal level and output was .88 ( p < .001; N = 105) .