A variety of statistical models have been developed to examine longitudinal stability and change of individual differences. Kenny and Zautra’s (1995) trait-state-error (TSE) model models stability as a function of stable variance that does not change (trait), moderately stable variance that changes over time (state), and error variance. Applications of this model have been limited to panel studies with repeated observations of the same individual. The present study developed a non-linear regression model to apply the TSE model to retest correlations from different samples. This model was used to compare the stability of measures of personality traits, affective well-being (AWB), cognitive well-being (CWB), and self-esteem. After correcting for differences in reliability, age, gender, and scale length, the amount of trait (vs. state) variance was similar across constructs. Stability of the state component was highest for AWB and CWB, suggesting that situational influences have most enduring effects on these two constructs.

A variety of statistical models have been developed to examine longitudinal stability and change of individual differences. Kenny and Zautra’s (1995) trait-state-error (TSE) model models stability as a function of stable variance that does not change (trait), moderately stable variance that changes over time (state), and error variance. Applications of this model have been limited to panel studies with repeated observations of the same individual. The present study developed a non-linear regression model to apply the TSE model to retest correlations from different samples. This model was used to compare the stability of measures of personality traits, affective well-being (AWB), cognitive well-being (CWB), and self-esteem. After correcting for differences in reliability, age, gender, and scale length, the amount of trait (vs. state) variance was similar across constructs. Stability of the state component was highest for AWB and CWB, suggesting that situational influences have most enduring effects on these two constructs.