Area under the Curve as an Alternative to Latent Growth Curve Modeling When Assessing the Effects of Predictor Variables on Repeated Measures of a Continuous Dependent Variable

Researchers conducting longitudinal data analysis in psychology and the behavioral sciences have several statistical methods to choose from, most of which either require specialized software conduct or advanced knowledge inform selection correct model options (e.g., correlation structure). One simple alternative conventional is calculate area under curve (AUC) from repeated measures then use this new variable one’s model. The present study assessed relative efficacy two AUC measures: with respect ground (AUC-g) increase (AUC-i) comparison latent growth modeling (LGCM), a popular method. Using ongoing Panel Study Income Dynamics (PSID), we effects four predictor variables on social anxiety, using both LGCM. We used full information maximum likelihood (FIML) method account for missing LGCM multiple imputation calculation measures. Extracting parameter estimates these models, next conducted Monte Carlo simulations assess bias power (two performance) same sample sizes ranging 741 50. results initial models paralleled those LGCM, particularly baseline. With simulations, preformed as well even better than all assessed. These suggest that may be viable especially researchers less access necessary