Premium
Multilevel regression analyses to investigate the relationship between two variables over time: Examining the longitudinal association between intrusion and avoidance
Author(s) -
Suvak Michael K.,
Walling Sherry M.,
Iverson Katherine M.,
Taft Casey T.,
Resick Patricia A.
Publication year - 2009
Publication title -
journal of traumatic stress
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.259
H-Index - 134
eISSN - 1573-6598
pISSN - 0894-9867
DOI - 10.1002/jts.20476
Subject(s) - multilevel model , association (psychology) , psychology , intrusion , regression analysis , longitudinal data , regression , longitudinal study , sample (material) , human factors and ergonomics , poison control , developmental psychology , clinical psychology , statistics , computer science , data mining , medicine , mathematics , medical emergency , psychotherapist , chemistry , geochemistry , chromatography , geology , psychoanalysis
Abstract Multilevel modeling is a powerful and flexible framework for analyzing nested data structures (e.g., repeated measures or longitudinal designs). The authors illustrate a series of multilevel regression procedures that can be used to elucidate the nature of the relationship between two variables across time. The goal is to help trauma researchers become more aware of the utility of multilevel modeling as a tool for increasing the field's understanding of posttraumatic adaptation. These procedures are demonstrated by examining the relationship between two posttraumatic symptoms, intrusion and avoidance, across five assessment points in a sample of rape and robbery survivors ( n = 286). Results revealed that changes in intrusion were highly correlated with changes in avoidance over the 18‐month posttrauma period.