We provide a comprehensive review of simple and advanced statistical analyses using an intuitive visual approach explicitly modeling Latent Variables (LV). This method can better illuminate what is assumed in each analytical method and what is actually estimated, by translating the causal relationships embedded in the graphical models in equation form. We recommend the graphical display rooted in the century old path analysis, that details all parameters of each statistical model, and suggest labeling that clarifies what is given vs.
View Article and Find Full Text PDFThe advantages of modeling the unreliability of outcomes when evaluating the comparative effectiveness of health interventions is illustrated. Adding an action-research intervention component to a regular summer job program for youth was expected to help in preventing risk behaviors. A series of simple two-group alternative structural equation models are compared to test the effect of the intervention on one key attitudinal outcome in terms of model fit and statistical power with Monte Carlo simulations.
View Article and Find Full Text PDFThis study investigates the impact of exposure to information about a relatively new prevention method, the female condom (FC), on actual FC use in a community of adults at risk of HIV/STI. A community-wide survey from a mid-size US city is used to estimate unbiased effects of information about the FC on FC use among sexually active men and women. To control for potential confounders we use propensity score matching (PSM) which matches the group exposed to FC information to participants who were not exposed, achieving a statistical quasi-randomization in terms of ten measured confounders.
View Article and Find Full Text PDFObjective: To compare the reliabilities and convergent validities of five CES-D (Center for Epidemiologic Studies Depression) composite scores in two ethnic/racial groups of community-dwelling older adults.
Design: CES-D measurement equivalence was tested with Structural Equation Modeling in Puerto-Rican (PR) and African American (AA) older adults, then reliabilities of five composite scores and their convergent validities were compared.
Findings: Bayesian CES-D scores had the highest reliabilities, followed by software estimated factor scores, the unit weight, and the weighted scores.