An SEM Approach for the Evaluation of Intervention Effects Using Pre-Post-Post Designs

Area of Science:

• Evaluation research
• Quantitative psychology
• Statistical modeling

Background:

• Latent curve models (LCMs) are increasingly used for analyzing change over time.
• Previous work established LCMs for pre-post-post designs.
• Further refinement is needed for comprehensive intervention effect analysis.

Purpose of the Study:

• To extend the application of LCMs for pre-post-post designs.
• To demonstrate improved methods for testing intervention effects.
• To illustrate modeling individual differences in change and residuals.

Main Methods:

• Analysis of latent change scores using latent curve models (LCMs).
• Application to pre-post-post research designs.
• Exploration of different LCM parameterizations for intervention effect testing.

Main Results:

• Intervention effects can be more effectively tested using specific LCM parameterizations.
• The study demonstrates testing the mean of a latent variable at any time point.
• Modeling of individual differences in outcomes and residuals is illustrated.

Conclusions:

• LCMs provide a powerful and flexible approach to analyzing change in evaluation research.
• This method effectively handles measurement error and individual differences in treatment response.
• LCMs avoid unrealistic assumptions compared to other change analysis techniques.