Abstract
Experiencing personal expressiveness (PE) and hedonic enjoyment (HE) during physical activity are associated with positive psychological and behavioural outcomes. Nevertheless, the validity evidence for scores from the questionnaire used to assess PE and HE is based on estimates of internal consistency and relationships with other variables only. The purpose of this investigation was to examine the factor structure of the Personally Expressive Activities Questionnaire (PEAQ; Waterman et al., 2003) contextualized to physical activity. Participants (N = 200; Mage = 32.56 years, SD = 15.69 years) completed the PEAQ. A confirmatory factor analysis model with two correlated subscales yielded a high correlation between PE and HE (r = .96) and demonstrated undesirable data-model fit. A confirmatory bi-factor model provided the best fit according to goodness-of-fit statistics (CFI=.95, RMSEA=.10. BIC=6493.97), yet parameters estimates were problematic (i.e., significant negative factor loadings).Conversely, the exploratory structural equation model had interpretable parameter estimates and demonstrated good fit (CFI=.93, RMSEA=.12. BIC=6516.93, rPE.HE=.79, p<.05). All target items loaded on their target latent factors (λ's .24-.92, p < .05) save for one PE item ("physical activity gives me my greatest feeling of really being alive"). Numerous cross-loadings were present (λ's .24-.71, p<.05) suggesting that some items tapped both PE and HE. Overall, researchers examining PE and HE using the PEAQ should account for cross-loadings to ensure parameter estimates are not biased. Further research and validation for scores of the PEAQ is needed before the instrument is widely adopted in physical activity contexts.
