Abstract
PURPOSE: Executive function has been identified as a potential area of vulnerability in individuals who stutter. The present study identified and analyzed the data across empirical studies of the executive function skills of adults who do (AWS) and do not stutter (AWNS). METHOD: Electronic databases, literature reviews, and reference sections of articles and dissertations were searched to identify candidate studies that examined behavioral measures of working memory, inhibition, and/or cognitive flexibility. A total of 39 studies met the eligibility criteria for this meta-analysis. A random-effects model was applied to estimate the pooled effect sizes (Hedges' g ) and 95% confidence intervals. RESULTS: AWS were significantly less accurate than AWNS on measures of working memory (Hedges' g = - 0.41, p < .001), including nonword repetition (Hedges' g = -.57, p < .001), forward digit span (Hedges' g = -0.23, p = .02), backward digit span (Hedges' g = -.38, p = .004) and operation span tasks (Hedges' g = - .37, p = .018). AWS performed comparably to AWNS on inhibition measures (Hedges' g = -0.10, p = .37). An insufficient number of published studies were available to conduct a meaningful analysis of cognitive flexibility. CONCLUSIONS: Present findings suggest that AWS, as a group, exhibit weaknesses in one component of executive function - working memory - compared to AWNS. Additional research is necessary to determine potential differences in inhibition and cognitive flexibility in AWS.