‘If they do it, so can I’: a test of a moderated serial mediation model of descriptive norms, self-efficacy, and perceived similarity for predicting physical activity

Abstract

Objective

Though engaging in physical activity (PA) is associated with several health benefits, physical inactivity rates remain high. Previous research revealed that descriptive norms (i.e., perceptions of others’ PA) positively relate to PA intentions/behaviours. This study tested whether self-efficacy mediated the influence of descriptive norms on PA intentions and behaviours, and whether the relationship between descriptive norms and self-efficacy was moderated by perceived similarity with normative referents.

Design

Online questionnaires were administered.

Measures: Baseline measures included descriptive norms, perceived similarity, self-efficacy, and PA intentions. Weekly PA behaviour frequency was assessed at baseline and two weeks post-baseline.

Results

Descriptive norms positively predicted PA behaviour frequency at time 2 indirectly via self-efficacy and PA intentions. These indirect effects were attenuated when baseline PA behaviour frequency was accounted for. There was no support for perceived similarity as a moderator, which may be attributed to a potential ceiling effect or the nature of items used to assess this perception.

Conclusion

The findings support the notion that individuals may feel more efficacious when they perceive that others are successfully engaging in the targeted task/activity. Future interventions may consider providing descriptive norm messages to individuals who lack self-efficacy or previous PA experience to promote their PA.

Keywords:

• Social norms
• vicarious experience
• exercise
• normative influence
• bootstrapping
• conditional indirect effect

Disclosure statement

No funding details or conflict of interest to disclose.

Data availability statement

The data that support the findings of this study are available from the corresponding author, Kim, J., upon reasonable request.

Notes

1 The raw dataset included 297 (time 1) and 190 (time 2) respondents. Responses from 15 (time 1) and 5 (time 2) participants were removed prior to analysis because (1) they self-reported as competitive athletes, or (2) they answered the screening question that indicated ‘Please leave this question blank’, suggesting participant neglect.

2 Baseline comparisons between those who were retained vs. dropped out at time 2 revealed no statistically significant differences in age, self-efficacy, PA behaviours, PA intentions (fixed and open scale), descriptive norms, and perceived similarity (p > .05 for all comparisons).

3 Prior to the analysis, the dataset was examined for missing data. For measures at time 1, no missing responses were noted for the fixed and open intention items, as well as for the mean scores of descriptive norms and perceived similarity across selected referents. For self-efficacy, the number of missing responses (i.e., selecting the ‘Not applicable’ option or no options) ranged from 0 to 29 across the items. One participant did not answer any of the self-efficacy items. For the frequencies of moderate and strenuous PA, four missing responses were noted for each type. At time 2, no missing responses were noted for moderate and strenuous PA. Overall, the data were missing completely at random, χ2 (469) = 494.39, p = .201. Of note, missing responses for descriptive norms and perceived similarity were assessed as the mean scores across the normative referents as opposed to the individual item scores, because they were invariably missing in a systematic manner depending on each participant’s normative referent selections. For instance, if a participant only selected ‘parents’, descriptive norm and perceived similarity items for all other referents were not administered for this participant and thus were missing in the dataset.