Causal Inference in LCA: Does Attending College Lead to Later Drinking Problems?
College is often perceived as a risky environment for problem drinking, but recent studies indicate that individuals who attend college engage in this behavior in adulthood at rates equal to or lower than those who do not attend college; that is, college may actually protect individuals from substance use behaviors in adulthood. These studies, however, often fail to account for selection bias: the fact that the people who attend college are different in many ways than people who do not attend college. In the article, "Causal Inference in Latent Class Analysis," to appear in Structural Equation Modeling, Methodology Center researchers Stephanie Lanza and Donna Coffman implement two propensity score techniques for causal inference in latent class analysis (LCA) to determine whether college enrollment is protective or harmful for substance use later in life.
Email firstname.lastname@example.org for an advance copy of the article.
Methodology Center Annual Report 2011-12
The Methodology Center's first annual report is now available online. The report provides context for our work and an overview of our research. It includes a letter from our director, summaries of our largest research projects and recent grants, and a review of our productivity during the last year. If you want to understand how we collaborate to address pressing public health issues, read the two-page spread about our work on developing time-varying effect models and applying those models to smoking-cessation research.
Open the annual report.
New Grant to Optimize a Technology-Based Obesity Intervention
Methodology Center Director Linda Collins is working on a new grant with Principal Investigator Bonnie Spring of the University of Illinois at Chicago. In this project, the researchers will employ the multiphase optimization strategy (MOST) to develop a resource-efficient, Internet-and-phone based, weight-loss intervention. Studies have shown that in-person intensive lifestyle interventions (INLIs) are the most effective weight loss treatments, but INLIs are too costly and burdensome to be used on a broad scale. By delivering the intervention via Internet and phone, scientists will be able to bring the intervention to scale at much lower cost. By applying the MOST framework, researchers will ensure that all components of the intervention actively contribute to participants' weight loss. The final goal of the project is to develop an effective weight-loss intervention that can be implemented for $500 or less.
Read more about MOST.
New Software: Updated SAS Macro for LCA with a Distal Outcome
We are pleased to announce the release of version 2 of the LCA Distal SAS macro. Latent class analysis allows researchers to divide subjects into underlying subgroups that cannot be directly observed. The %LCA_Distal SAS macro was created to allow researchers to estimate the effect of membership in a latent class on an outcome at a later time. The new version of the macro can accommodate distal outcomes for multiple groups.
Download the macro or read more.
Videos and New Book: Analyzing Longitudinal Dyadic Data
Our colleagues Niall Bolger of Columbia University and Jean-Philippe Laurenceau of The University of Delaware, recently published a book, Intensive Longitudinal Methods: An Introduction to Diary and Experience Sampling Research, available from Guilford Press. This book is a practical guide to planning and executing an intensive longitudinal study.
In 2010, Niall and Jean-Phillipe presented a workshop on the analysis of longitudinal dyadic data at the 15th Summer Institute on Longitudinal Methods, sponsored by Penn State and the National Institute on Drug Abuse. Fifteen different videos of the workshop are available for viewing on our website.
See the videos from the workshop.