Category: Research Methods and Statistics

Symposium

Symposium 117 - Cutting-Edge Longitudinal Models for CBT Research

Sunday, November 19
8:30 AM - 10:00 AM
Location: Aqua 300 A & B, Level 3, Aqua Level

Keywords: Research Methods | Statistics | Longitudinal
Presentation Type: Symposium

In order to establish causal mechanisms of behavioral interventions, longitudinal studies with three or more timepoints are necessary (MacKinnon, Lockhart, Baraldi, & Gelfan, 2013). However, longitudinal CBT investigations have commonly evaluated only two timepoints, at pre- and post-treatment. This practice impedes the evaluation of treatment mechanisms, discards information about how each session affects the entire course of CBT, and reduces statistical power to detect effects (Moerbeek & Teerenstra, 2016). As a result, conventional longitudinal designs limit the ability to evaluate hypotheses of interest and force researchers to recruit more participants. Furthermore, even when using newer longitudinal models, common errors made when implementing these models in behavioral research can actually make them perform more poorly than traditional ANOVA/regression methods (Kelley & Rausch, 2011).


In this symposium, we will discuss cutting-edge methods that address many problems that investigators face with longitudinal models. First, Eiko Fried will address longitudinal applications of network models. These models are growing in influence in CBT research given their ability to resolve common problems found in latent variable models, but information on their use in longitudinal settings has not been well disseminated. Second, Nicholas Jacobson will present how the differential time-varying effect model (DTVEM) can address linear and nonlinear change in the context of ecological momentary assessment. In this era of “big data,” conventional methods are insufficient for intensive longitudinal data often produced by technological input (e.g., hundreds of datapoints per participant obtained from smartphones). Given the nonlinear possibilities with such intensive datasets, DTVEM allows for maximal utilization of the additional data.Third, Hani Zainal will show how structural equation modeling can be used to differentiate direct and indirect longitudinal effects in behavioral research. Finally, Alessandro De Nadai will address common myths and lesser-known features of linear and nonlinear multilevel models. In particular, he will identify how to address common errors made when using these models and how to maximize their strengths. Following these presentations, Eric Storch will highlight current and future applications of these methods based on his extensive experience with applied longitudinal CBT research.


Through these presentations, audience members will learn how to implement a number of innovative models in real-world CBT research. They will also learn how to avoid common pitfalls that affect conventional and modern longitudinal methods. Applied examples and specific syntax will be provided.

Learning Objectives:

Alessandro S. De Nadai

PhD candidate
University of South Florida, University of Mississippi Medical Center, Texas State University

Presentation(s):

Send Email for Alessandro De Nadai

Lance M. Rappaport

Postdoctoral Fellow
Virginia Commonwealth University

Presentation(s):

    Send Email for Lance Rappaport

    Eiko I. Fried

    Postdoctoral Fellow
    University of Amsterdam

    Presentation(s):

    Send Email for Eiko Fried

    Nicholas C. Jacobson

    Graduate student
    The Pennsylvania State University

    Presentation(s):

    Send Email for Nicholas Jacobson

    Hani Zainal

    Doctoral Student
    The Pennsylvania State University

    Presentation(s):

    Send Email for Hani Zainal

    Alessandro S. De Nadai

    PhD candidate
    University of South Florida, University of Mississippi Medical Center, Texas State University

    Presentation(s):

    Send Email for Alessandro De Nadai


    Assets

    Symposium 117 - Cutting-Edge Longitudinal Models for CBT Research



    Attendees who have favorited this

    Please enter your access key

    The asset you are trying to access is locked. Please enter your access key to unlock.

    Send Email for Cutting-Edge Longitudinal Models for CBT Research