A systematic meta-analysis of the Stroop task in depression

https://doi.org/10.1016/j.cpr.2012.02.005Get rights and content

Abstract

Despite the extensive use of the Emotional Stroop task in depression, only qualitative reviews have been produced to date, and these reviews conclude that Stroop performance in depression is equivocal. The present meta-analysis addressed the need to summarize the data quantitatively. A thorough search of the literature was conducted and 47 published studies and unpublished doctoral dissertations were included in the analyses. The meta-analysis revealed large and robust depression-related Stroop effects (e.g., for clinically depressed versus control participants on negative stimuli, g = .98, and on positive stimuli, g = .87). Although the effects did not reflect a strong emotion-congruent bias, they did distinguish among levels of depressive experience, in that greater levels of depression severity were associated with larger between-groups effect sizes. Moreover, these effects have been obtained without priming procedures, or the presentation of self-relevant or disorder-congruent stimuli. These findings challenge schema-based theories of the Emotional Stroop effect and predictions based on previous qualitative reviews of the literature. The findings also suggest that further comparative behavioural research on the depression-related Stroop effect, at least among clinically depressed populations, is not necessary. Future research should address questions about underlying mechanisms and focus on a more direct measure of depression-related attentional bias.

Highlights

► Large and robust depression-related effects were revealed. ► Greater levels of depression severity were associated with large effects. ► Effects were found without priming, self-relevant or disorder-congruent stimuli. ► The results however do not reflect a strong emotion-congruent bias. ► The results settle the debate about the Stroop task in depression.

Introduction

The Stroop task (Stroop, 1935) has received considerable research attention over the last seventy years. MacLeod (1991) estimated that there were more than 700 Stroop-related articles in the literature, and the numbers have increased since that time. Continued interest in the Stroop may be attributable to the widening breadth of applications of the task, such as research on cognitive models of psychopathology. Meta-analytic reviews of Stroop performance can now be found for such diverse groups as individuals with eating disorders (Dobson & Dozois, 2004), addiction problems (Cox, Fadardi, & Pothos, 2006), difficulty with anxiety (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg, & IJzendoorn, 2007), attention-deficit hyperactivity disorder (Lansbergen, Kenemans, & van Engeland, 2007) and schizophrenia (Szöke et al., 2008). Despite the extensive use of the Stroop task in depression, however, only qualitative reviews have been produced to date (Gotlib et al., 1996, Mogg and Bradley, 2005, Williams et al., 1996). These reviews conclude that Stroop performance in depression is equivocal. Thus, it is timely to conduct a meta-analytic review to ascertain the state of the science.

Section snippets

The Stroop task in depression

The original, or Classic Stroop task (Stroop, 1935) involves the presentation of color words, in incongruously colored ink (e.g., the word ‘red’ printed in blue ink). Participants are asked to name, as quickly as possible, the ink color of each stimulus word, while attempting to ignore the meaning of the word. This attempt to suppress word meaning in order to name ink color has reliably been shown to result in longer response latencies than those that result from color naming congruent stimuli

Evidence for a depression-related Stroop effect

Beck's and Bower's models predict that depression should be associated with an attentional bias for mood-congruent stimuli. With the Stroop task, this prediction suggests that there should be mood-congruent interference effects for depressed populations, compared with controls. On the other hand, several researchers have argued that depression is associated with biases in controlled or effortful processing in processes such as interpretation and memory, and not with early or relatively

Meta-analysis as a statistical tool

Research results are known to be inconsistent (Schwarzer, 1991), at least partly due to variability in methodologies across studies. Meta-analysis is a statistical technique which aggregates the summary statistics from a number of studies, to draw overall conclusions on the data from a broad literature. The statistical value reported in meta-analysis is the effect size (ES), defined as the standardized mean difference between a criterion group and a comparison group on an outcome variable (

Predictions and queries

Given the mixed pattern of results demonstrated using the Stroop task and debate in the literature, the objective of the current meta-analysis was to address the following predictions: 1) Overall, weak effect sizes were expected for depression-related Stroop effects, given the inconsistencies in the literature. 2) Despite uncertainty in the empirical literature, greater Stroop effects were predicted for clinically depressed groups than for dysphoric or mood induced groups. According to Beck's

Literature search

Studies were collected through a search of the online bibliographic search engines PsycINFO and PubMed, using the keywords: ‘depression’, ‘Stroop’, ‘attention’, ‘information processing’, ‘attentional bias’, and ‘cognitive interference’. The search was restricted to English language articles, published electronically or in hard copy, up to and including 15 July 2010. Unpublished doctoral dissertations were also collected. Using a process of tracking back articles, relevant citations from

Description of the data

A search of PsycInfo using the previously described keywords yielded 7151 hits, and a search of PubMed yielded 5570 hits. There was significant overlap across search results. Thus, the most relevant and least overlapping combinations of keywords were further explored both in PsycInfo and PubMed: ‘depression + Stroop’, ‘depression + cognitive interference’, and ‘depression + attentional bias’. This more refined search yielded a total of 930 abstracts and from those abstracts, 229 articles which

Discussion

The mixed pattern of results that have been identified for the Stroop task in depression, combined with debate in the literature, motivated the current meta-analysis in order to quantitatively examine these relationships. The results of the current meta-analysis demonstrated large effect sizes for both the Emotional Stroop task and the Classic Stroop task, when clinically depressed individuals were compared to control groups (see Table 2, Table 3 for a summary of the results). Between-groups

Conclusion

The present meta-analysis demonstrated that there are large and robust depression-related Stroop interference effects. Although these effects do not reflect an emotion-congruent bias, they do distinguish among levels of depressive experience, where greater severity is associated with larger effect sizes. Moreover, these effects do not require priming procedures, longer stimulus exposure, or the presentation of self-relevant or disorder-congruent stimuli, to be obtained. Heterogeneity within the

Sources of Funding

The first author was supported in her doctoral studies by a Doctoral Canada Graduate Scholarship (Social Sciences and Humanities Research Council of Canada) and an Alberta Innovates—Health Solutions Studentship for Mental Health Research. These funding bodies were not involved in any aspect of the study design, in the collection, analysis, and interpretation of data, in the writing of the report or in the decision to submit the paper for publication.

References (0)

Cited by (143)

View all citing articles on Scopus
2

References that were included in the analyses are indicated by an asterisk at the start of the citation.

View full text