Abstract
There is emerging evidence of overlap across cognitive processes. One explanation of this overlap is the presence of a single,
higher-order latent process. In this study we tested for a core process and its ability to account for symptoms of depression
and anxiety. Using Structural Equation Modeling we compared a model where processes (worry, thought suppression and experiential
avoidance) are treated as separate predictors of symptoms (anxiety and depression) against a model where they are represented
by one latent factor. These models were applied in three analyses: a cross-sectional student sample; a longitudinal subset
of this analogue sample; and a cross-sectional sample of individuals with long-term health conditions. Comparison of the models
showed that while the two sets of models provided comparable fits to the data, the single factor models provided a more parsimonious
solution. In addition, the latent factor explained a large proportion of variance in all measured processes, suggesting a
high degree of overlap between them. It also explained more variance in symptoms than the processes separately. A Confirmatory
Factor Analysis further supported a single factor solution, and the item loadings indicated that the core process represented
a perceived inability to control negative thinking.
higher-order latent process. In this study we tested for a core process and its ability to account for symptoms of depression
and anxiety. Using Structural Equation Modeling we compared a model where processes (worry, thought suppression and experiential
avoidance) are treated as separate predictors of symptoms (anxiety and depression) against a model where they are represented
by one latent factor. These models were applied in three analyses: a cross-sectional student sample; a longitudinal subset
of this analogue sample; and a cross-sectional sample of individuals with long-term health conditions. Comparison of the models
showed that while the two sets of models provided comparable fits to the data, the single factor models provided a more parsimonious
solution. In addition, the latent factor explained a large proportion of variance in all measured processes, suggesting a
high degree of overlap between them. It also explained more variance in symptoms than the processes separately. A Confirmatory
Factor Analysis further supported a single factor solution, and the item loadings indicated that the core process represented
a perceived inability to control negative thinking.
- Content Type Journal Article
- Category Original Article
- Pages 1-17
- DOI 10.1007/s10608-012-9475-2
- Authors
- Timothy Bird, School of Psychological Sciences, The University of Manchester, Manchester, Coupland I, Oxford Road, Manchester, M13 9PL UK
- Warren Mansell, School of Psychological Sciences, The University of Manchester, Manchester, Coupland I, Oxford Road, Manchester, M13 9PL UK
- Chris Dickens, Peninsula College of Medicine and Dentistry, Universities of Exeter and Plymouth, Exeter, UK
- Sara Tai, School of Psychological Sciences, The University of Manchester, Manchester, Coupland I, Oxford Road, Manchester, M13 9PL UK
- Journal Cognitive Therapy and Research
- Online ISSN 1573-2819
- Print ISSN 0147-5916