Self-Serving Criteria Bias
People may rate their own abilities differently depending on whose criteria they use. In the described experiment, participants given another person’s criteria lowered their self-ratings, while those using their own criteria showed little change. This demonstrates sensitivity to standards, without by itself establishing which ratings were more accurate.
THE IDEA
Explanation
The self-serving criteria bias refers to the tendency of individuals to use self-serving standards when evaluating their own abilities, leading to inflated self-assessments. In an experiment, participants rated themselves on athletics, talent, and extracurricular activities. They were assigned to one of three conditions: control, own-criteria (using their own personal benchmarks), or yoked-criteria (using criteria matched to another participant). The results showed that only the yoked-criteria group significantly lowered their self-evaluations from pretest to posttest, indicating that when forced to use external criteria, people adjust their self-views downward. The own-criteria and control groups showed little change, suggesting a bias toward maintaining high self-assessment through flexible criteria.
Results
We predicted that self-evaluations would remain the same between pretest and posttest for the control and own criteria groups, but would dramatically decrease in the yoked-criteria condition. In order to test this prediction, we collapsed self-evaluations across the three domains studied and subjected those evaluations to a 3 (criteria condition: own, yoked, control) × 2 (pretest vs. posttest) ANOVA, with the last factor serving as a repeated measure. This analysis revealed the expected Criteria Condition × Test interaction, F (2, 33) = 7.39, p *< .*005. Indeed, similar significant interactions were observed for two of the individual domains we examined: athletics, F (2, 33) = 5.98, p *< .*01, and talent, F (2, 33) = 4.35, p *< .*03.
As can be seen in Table 18.3, subjects did not reliably alter their self-assessments between pretest and posttest in the control and own-criteria conditions. Of eight possible pretest–posttest comparisons, only one revealed a statistically significant change, which was a small rise in ratings of athletic skill in the control condition (from M = 50.1 to 54.8), t(11) = 2.45, p *< .*05. Within the yoked-criteria group, however, subjects consistently lowered their self-evaluations from pretest to posttest. If we look at responses collapsed across domain, we find that percentile estimates went from M of 53.3 on the pretest to 41.3 on the posttest, t(11) = 2.55, p *< .*05. A significant decrease was also seen in the domain of athletics (from M = 51.3 to 34.3), t(11) = 2.33, p *< .*05, and artistic talent, t(11) = 2.99, p *< .*05. When considering activity in extracurricular programs, subjects did lower their assessments an average of 9.2 points, though that difference was not statistically significant. . . .
When to use it
Performance Reviews: Employees might overestimate their performance if they use personal standards; using standardized criteria can provide more accurate evaluations.
Academic Self-Assessment: Students might overestimate their knowledge if they compare to their own criteria; using exam averages or rubrics can adjust their self-perception.
Health and Fitness: People may think they are more fit than they actually are if they use subjective criteria; objective measures like VO2 max or body fat percentage can correct this.
Skill Development: When learning a new skill, using benchmarks from experts or peers can prevent overconfidence and encourage improvement.
FROM THEORY TO EVERYDAY LIFE
In practice
Yoked-Criteria Condition in Athletics
In the experiment, participants in the yoked-criteria condition lowered their self-evaluation of athletic ability from a mean percentile of 51.3 (pretest) to 34.3 (posttest), showing a significant decrease (p < .05).
Yoked-Criteria Condition in Artistic Talent
Artistic-talent ratings also fell significantly in the yoked-criteria group. The means 51.3 and 34.3 belong to the athletics result above, not the artistic-talent result; the excerpt reports the latter’s significance as p < .05 without giving its means.
A MOMENT TO REFLECT
Put it to work.
Recall a situation where you felt confident about your ability in a domain. Were you using your own criteria or an external standard? How might your self-assessment have been biased?
How can you incorporate external, objective criteria into your self-evaluations to reduce the self-serving criteria bias?
In what areas of life (e.g., work, academics, sports) might this bias lead to overconfidence, and what can you do to mitigate it?
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)