The purpose of the proposed research project is to make the use of statistically sophisticated techniques for the modeling of longitudinal data more accessible to applied researchers in the drug abuse prevention and intervention area. To this end, this project seeks to demonstrate the use and usefulness of latent growth models (LGMs) in the analysis of longitudinal substance use data.
Three aims are proposed: (1) To contrast the use of latent growth models with the use of more traditionally used techniques for longitudinal, experimental design data in the analysis of two substance use prevention datasets; (2) To design and execute a simulation study to evaluate the statistical power and the Type I error rate for 2 different LGM approaches and 2 different ANOVA approaches in the analysis of longitudinal experimental design data in regards to 6 commonly occurring intervention effect patterns; and (3) To provide guidance to applied researchers in the use of these techniques by providing relevant, explicit examples. Jointly, the proposed project will provide further rationale and guidance for efforts to more closely integrate advances of applied statistics and behavioral research with the method of data analysis employed by substantive researchers.
| Hoeppner, Bettina B; Goodwin, Matthew S; Velicer, Wayne F et al. (2008) Detecting longitudinal patterns of daily smoking following drastic cigarette reduction. Addict Behav 33:623-39 |
