Applying Bayesian statistics to research on psychological trauma: Introduction to the special section.

Abstract

Use of Bayesian statistics is increasingly common in research in general and in psychological research in particular. This is especially the case for research on psychological trauma, in which the strengths of Bayesian statistics are particularly pronounced. The goals of this introduction to the special section on applying Bayesian statistics to research on psychological trauma are two-fold: first, to overview and discuss the benefits of Bayesian statistics and, second, to introduce the articles in this special section. Clinical Impact StatementThis article introduces a collection of articles applying a specific kind of statistical approach to research on psychological trauma. These articles make use of this statistical approach as well as contribute to basic and applied trauma research.

Influence of trauma and personality on posttraumatic cognitions in military veterans.

“…We examined the influence of trauma and personality on posttraumatic cognitions using a Bayesian approach to multiple regression. Bayesian statistics is becoming increasingly popular for psychological trauma research because of its ability to manage nonnormality and small sample sizes, both of which are frequent issues in trauma research (van de Schoot et al, 2017;Yalch, 2016Yalch, , 2023. We estimated three regression models with trauma (combat exposure and other potentially traumatic military experiences) and personality traits as the predictors and each posttraumatic cognition as the outcome, using the rstanarm package in R (Goodrich et al, 2020).…”

The influence of betrayal trauma on complex posttraumatic stress disorder symptoms.

“…We tested our hypotheses using three multiple regression models (one for the whole sample, one for women, and one for men) with complex PTSD symptoms as the outcome and high, medium, and low betrayal trauma as predictors. We estimated our models within a Bayesian framework, which provides a more accurate analysis of nonnormal distributions of data common in trauma research (Yalch, 2016(Yalch, , 2022 and is thus increasingly common in research on psychological trauma (van de Schoot et al, 2017). We implemented models using Hamiltonian Markov chain Monte Carlo (HMC; see Betancourt, 2017) via the rstanarm package in R (Goodrich et al, 2018).…”

Influence of betrayal trauma on obsessive-compulsive personality pathology.

“…We tested our hypotheses using a Bayesian approach to structural equation modeling (BSEM; Depaoli, 2021), which provides more accurate estimates of effects relative to conventional approaches to data analysis and is more robust to the analysis of nonnormal distributions of data that are common in trauma research and are therefore increasingly common in trauma research (van de Schoot et al, 2017;Yalch, 2016Yalch, , 2023. In this study, we implemented BSEM using the blavaan package in R (Merkle & Rosseel, 2018).…”