Validation of the Measures of the Transtheoretical Model for Exercise in an Adult African-American Sample

Abstract

Purpose. African-Americans have high rates of physical inactivity-related morbidity and mortality, thus effective interventions to increase exercise are necessary. Tailored interventions show promise, but measures need validation in this population. This study validated transtheoretical model measures for exercise in an African-American sample. Design. Cross-sectional measure development. Setting. Telephone survey of individuals in North Carolina. Subjects. 521 African-American adults. Measures. Stages of change, decisional balance (pros and cons), self-efficacy and processes of change (POC) for regular exercise. Analysis. Confirmatory factor analyses tested measurement models. Multivariate analyses examined relationships between each construct and stages of change. Results. For decisional balance, the two-factor uncorrelated model was the most parsimonious good-fitting model (χ235 = 158.76; comparative fit index [CFI], .92; average absolute standardized residual [AASR], .04), and alphas were good (pros α = .85 and cons α = .74). The one-factor model for self-efficacy (α = .80) revealed an excellent fit (χ29 = 45.51; CFI, .96; AASR, .03). For the POC subscales with good alphas (α = .62–.91), a 10-factor fully correlated model fit best (χ2[360] = 786.75; CFI, .91; AASR, .04). Multivariate analyses by stage of change replicated expected patterns for the pros, self-efficacy, and POC measures with medium-sized effects (η2 = .05–.13). Results by stage of change did not replicate for the cons scale. Conclusions. The structures of these measures replicated with good internal and external validity, except for the cons scale, which requires additional development. Results support the use of these measures in tailored interventions to increase exercise among African-Americans.