PROBLEMS WITH GOAL-SETTING RESEARCH
Locke, E. A. (1991). Problems with goal-setting research in sports--and their solution. Journal of Sport and Exercise Psychology, 13, 311-316.
That goal-setting improves task performance is one of the best established findings in management and psychology. However, there are studies in sport and exercise psychology that have had anomalous results. This article highlights the errors in research that have produced these results.
- The failure to manipulate the "no-goal" or "do best" condition so that spontaneous goal-setting does not occur. When Ss are given feedback about performance, they often use it to set goals. It has been found that Ss do spontaneously set goals, particularly when given feedback in laboratory settings. When "do-best" Ss, for some reason, do not set goals or are prevented from doing so, the value in setting goals for improving performance can be observed. Thus, it is important that "do-best" Ss do not receive feedback and/or are prevented from doing so in comparative studies.
- Measure personal/actual goals. It is imperative to know what personal (actual) goal each person sets in response to the external goal that was assigned. Goal theory asserts that assigned goals affect performance through their effects on personal goals. Knowing that a person is not committed to an assigned goal is not very helpful unless the goal which was substituted is known. Ss should be required to record actual goals in writing or on a voice tape to locate the real goals used in studies.
- Make specific goals difficult. Specific goals that are actually easy usually lead to lowered levels of performance. To display effects goals must be difficult. A suggested level of difficulty is that no more than 10% of Ss can reach them.
- Make sure subjects are committed to achieving the goal. Commitment reveals if an S has accepted an assigned goal. A personal goal indicates what new personalized goal has been set. It is necessary to obtain some indication (usually marking a position on a 5-point commitment scale) to indicate if the goal-setting that is being manipulated in the experiment is being used that way.
- Baseline. It is important that experimental groups do not start from different baseline levels. One way of controlling this is to perform single-subject studies.
- Competition. The effects of competition, which is a variation of goal-setting, have not been controlled well. The measurement of this factor is very difficult. If A's goal is to beat B, then A's personal goal becomes B's performance level or better. It is best to use single-subject designs so that goal-aspirations will be reliably set.
- Measure self-efficacy, not subjective difficulty or effort. Subjective difficulty is not a very useful measure because it is confounded. It reflects the level of objective goal-difficulty and the S's perception of his or her ability to achieve the goal. Subjective difficulty is correlated positively with objective difficulty but negatively with self-perceived ability. A better measure is self-efficacy, one's confidence in being able to execute a course of action.
Many of these problems can be alleviated by recording a S's plans and reactions and using single-subject designs. It is not appropriate to assume any path of action with goal-setting research because of the common tendency of Ss to personalize and modify externally imposed goals in challenging settings.
Return to Table of Contents for this issue.