Greitemeyer, T., & Mügge, D. O. (2014). Video games do affect social outcomes: A meta-analytic review of the effects of violent and prosocial video game play. Personality and Social Psychology Bulletin, 40(5), 578-589. (Original DOI:
For the above-mentioned article, when coding the effect sizes, the authors also coded the study design, the type of outcome, and the research group to test for possible moderation effects. Four coding errors were discovered after the print version of this article was published. The four errors were as follows:
Jerabeck and Ferguson (2013) was given research group = 2; the correct code was research group = 1. Regarding Gabbiadini, Andrighetto, and Volpato (2012), an r of .83 was coded for the dependent measure moral disengagement; the correct correlation coefficient was r = .11. Von Salisch, Vogelsang, Kristen, and Oppl (2011) was entered as experimental; the correct design was longitudinal. Regarding Hasan, Bègue, and Bushman (2012), the outcome of hostile expectation bias was entered as a form of aggressive behavior; the correct outcome was aggressive cognition.
None of the errors changed the results of the meta-analyses or conclusions in a substantive way. Following is a list of the corrected values.
On p. 581, in the second sentence of the first paragraph under "Comparison of Violent Versus Prosocial Video Game Effects," the data for "exposure to violent video games" should read r = .17, 95% CI = [.15, .19], z = 13.91 (instead of r = .18, CI = [0.15, 0.21], z = 12.70), and in the last sentence of the paragraph, the data should read z = 0.22, p = .829 (instead of z = 0.17, p = .862). In Table 1 on p. 582, the data for "the aggressive behavior–violent outcome" should read n = 21,145, K = 45, effect size and 95% CI = .18, [.15, .21], z = 11.65 (instead of n = 21,215, K = 43, effect size and 95% CI = .19 [.16, .22], z = 11.38). In Table 1 on p. 582, the data regarding "the aggressive cognition–violent outcome", the data should read n = 8,038, K = 29, effect size and 95% CI = .24 [.19, .29], z = 8.98 (instead of n = 7,938, K = 28, effect size and 95% CI = .25 [.19. 31], z = 7.89). In Table 2 on p. 582, the data for "the experimental–violent design" should read n = 7,437, K = 56, effect size and 95% CI = .19 [.15, .23], z = 9.36 (instead of n = 7,761, K = 57, effect size and 95% CI = .20 [.15, .25], z = 7.94). In Table 2 on p. 582, the data for "the longitudinal–violent design" should read n = 6,607, K = 10, effect size and 95% CI = .11 [.05, .17], z = 3.57, p = .000 (instead of n = 6,283, K = 9, effect size and 95% CI = .10 [.04. 16], z = 3.09, p = .002). In Table 3 on p. 583, the data for "Anderson/Bushman" should read n = 8,495, K = 19, effect size and 95% CI = .20 [.15, .25], z = 7.46 (instead of n = 8,595, K = 20, effect size and 95% CI = .19 [.14. 24], z = 7.39). In Table 3 on p. 583, the data for "Ferguson" should read n = 2,544, K = 8, effect size and 95% CI = .03 [–.05, .10], Z = 0.67, p = .503 (instead of n = 2,444, K = 7, effect size and 95% CI = .02 [–.05. 10], Z = 0.61, p = .540). In Table 3 on p. 583, the data for "Others" should read effect size and 95% CI = .18 [.16, .21], Z = 13.63 (instead of effect size and 95% CI = .20 [.16, .23], Z = 11.50). On p. 583, in the second paragraph under "Publication Bias," the first sentence should read "For experimental studies (K = 56) of the effects of violent video game exposure, the adjusted effect (r = .19, with 5 trimmed/imputed effects) was essentially unchanged; Rosenthal’s fail-safe N = 3,096; Orwin’s fail-safe N = 42" (instead of "For experimental studies (K = 57) of the effects of violent video game exposure, the adjusted effect (r = .25, with 12 trimmed/imputed effects) was larger than the observed effect (r = .20); Rosenthal’s fail-safe N = 4,082; Orwin’s fail-safe N = 63." On p. 583, in the first sentence of the last paragraph, "the overall effect sizes of violent video game exposure" should read r = .17 (instead of r = .18).
The authors thank Joseph Hilgard for discovering the errors and thank Andreas Kastenmüller for help with the re-analyses. The authors’ codings of all studies can be found in the spreadsheet in the article’s online supplemental materials (available at pspb.sagepub.com/supplemental). The spreadsheet includes the sample size and the effect size, as well as the moderator codings.
