Effect of previously assigned goals on self-set goals and performance.

[Correction Notice: An erratum for this article was reported in Vol 70(2) of Journal of Applied Psychology (see record 2008-10965-001). The third line was left out on p. 696, top right column. The entire section should read as follows: "Figure 1 compares the goals on T-1 with the mean goals chosen on T-2 by the same groups of subjects (i.e., grouped according to T-1 goals). On T-2 the mean goal levels of the seven original groups ranged from 4.8 to 12.9, in contrast with the T-1 range of 2 to 26."] 231 undergraduates were asked to list from 2–26 uses for common objects depending on the level of difficulty of the section (1 of 7) to which they were assigned. Ss were given a 1-min practice trial after task explanations and were then administered 2 experimental trials. Ss were told they were free to change their goals to a higher or lower level for the 2nd trial if they did not like the goal they had been assigned. Results show that Ss chose more difficult goals, if the assigned goals had been easy, and easier goals, if the assigned goals had been difficult. Ss were heavily influenced in their self-set goals by their previously assigned goals. The performance of Ss with impossible goals did not drop on the 2nd trial. A goal–expectancy interaction was found on the 2nd trial that was due to expectancy being positively related to performance at the higher but not the lower goal levels. (8 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)