Testing Measurement Invariance of the 7-item Social Physique Anxiety Scale

ABSTRACT

Social physique anxiety is commonly examined in exercise science to understand individuals’ physical activity motivation and behavior. Although a 12-item and 9-item version of the Social Physique Anxiety Scale exists, a 7-item version has become a popular tool to measure this construct despite the need to further examine its measurement properties. The current study examined the factor structure of the 7-item scale and measurement invariance across age, gender, body mass index, typical exercise mode, and education level. Similar to past work, confirmatory factor analysis model modifications were needed to achieve adequate fit in our large sample of physically active men and women (N = 1009; M age = 28.6 ± 10.6 years). Multi-group latent analysis revealed measurement invariance only for the education level grouping variable. Researchers should examine and report the measurement model of the 7-item scale along with measurement invariance before proceeding with main analyses.

KEYWORDS:

• Self-perception
• measurement equivalence
• multi-group analyses

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Martin et al. (1997) removed three items (i.e., items 1, 2, and 5) based on statistical and conceptual arguments.

² Though not the focus of the present study, measurement invariance can also be used to assess a construct across time to ensure the same factor is being measured at multiple time points (see Pacewicz et al., 2022 for a detailed explanation of both approaches).

³ Weight for participants who reported a weight below 50 lbs were treated as missing and BMI was not calculated. Additionally, we treated BMI values below 13.0 as misreported as this value is linked to organ failure and is life threatening. It is highly unlikely the individual would be completing a survey if their health was in such decline. Thus, this signaled misreporting or the desire not to report one’s weight.

⁴ Sáenz-Alvarez et al. (2013) deleted this item and utilized a 6-item scale in a sample of adolescents.

Additional information

Funding

This work was supported by California State University, San Bernardino