Fig. 5
From: g versus c: comparing individual and collective intelligence across two meta-analyses
Forest plot: Average IQ and Criterion Task Performance. Note. Results are displayed for a meta-analysis using a random effects model for correlations between average IQ scores and group performance on criterion tasks. The average IQ is operationalized using the total IQ scores of each individual member divided by the number of members (= ∑IQ / n). The study by Barlow and Dennis (2016) did not apply the c-factor to the criterion task (the profit maximization task) because the authors found it was not valid. Studies behind the other (five) correlations, however, did apply both the c-factor and Av.IQ to the same criterion task—allowing for relative comparison. The Wonderlic Personnel Test (WPT) was used for all except Woolley et al. (2010) Study 1 where 18 odd items of 36-item Raven’s Advanced Progressive Matrices were used and in Rowe (2019) where the ICAR-16 was used. Box sizes are relative to sample size weights. *Two effects (team learning and team synergy) by Woolley and Aggarwal (under review) from one unique sample (n = 59) were included in this meta-analysis, leading to a total N = 425 when the sample was added on the basis of each unique effect or a total N = 366 when the sample was added on the basis of each unique group