The effect of descriptive norms and social identification on performance and exertion during a physical fitness task

Abstract

Research suggests that people's perceptions of the typical physical activity behaviors of others-descriptive norms-shape their own physical activity. However, prior work has primarily used cross-sectional designs, focused on people's physical activity intentions or self-reported overall physical activity, and failed to attend to how the normative referent affects the norm-behavior relationship. In a pre-registered experiment, we manipulated the descriptive norm to explore its effect on perceived exertion during, and objectively assessed performance on, an exercise (running) task. We also measured the strength of participants' social identification as a member of the norm reference group as a potential moderator. Following a baseline trial, participants (N = 78) were either told that their baseline score on the running task was below average (high norm condition) or above average (low norm condition) compared to members of the reference group. Participants in both conditions tended to perform better in Trial 2 than Trial 1. However, participants in the high norm condition improved their performance significantly (2.5 times) more than participants in the low norm condition. Social identification moderated the effect of the norm manipulation on perceived exertion. High identifiers tended to exert themselves less than low identifiers during Trial 2 if they were told they were above average, whereas if told they were below average, high identifiers exerted themselves slightly more than low identifiers. Results provide evidence that descriptive norm messages can improve people's performance on physically demanding tasks, and indicate that descriptive physical activity norms may be more influential when the norm reference group is subjectively meaningful. Opportunities to fruitfully deploy norm messages in applied contexts are discussed.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 11 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.