Abstract
Gerotranscendence is a rapidly growing theory of psychological development in older adulthood that is implicated in well-being and life satisfaction during later life. The purpose of this study was to assess the structural validity of Tornstam’s 10-item gerotranscendence scale in a population of unmarried older adults. A total of 227 older adults (n = 65 never-married, n = 93 widowed, and n = 69 divorced) from the Iowa Unmarried Survivors Study participated (71% female, 92% Caucasian), ranging in age from 65 to 94 years old. A confirmatory factor analysis (CFA) using Mplus was computed with the gerotranscendence scale’s traditional three factor structure; however, this resulted in no convergence of the data. An exploratory factor analysis (EFA) was completed to identify a better fitting model, which indicated a four factor structure fit the data more optimally. However, due to insufficient factor loadings, one factor was dropped from further analysis. The adjusted gerotranscendence scale CFA was computed with the modified factor structure. Results indicated an acceptable model fit to the data, χ2 (df = 11) = 25.53, p = .01; CFI = .95; RMSEA = .08, using the modified gerotranscendence scale. Factor loadings were also acceptable, ranging from .61 to .93, with one item loading at .41. Results suggest that a modified version of Tornstam’s gerotranscendence scale with revised dimension factors may be more appropriate for assessing gerotranscendence in a population of unmarried older adults.