Figure 3

(A) Cumulative Leisure Reading across Different Categories and Self-Reported Prosocial Behavior. Numbers represent standardized coefficients with standard errors in parentheses. Coefficients for models that include control variables (migration background, socioeconomic status, nonverbal intelligence) are printed in bold. Coefficients for the reduced sample of students nested in schools are in italics. Model fit: χ² = 416.480, df = 70, RMSEA = .031, SRMR = .024, CFI = .978, TLI = .972. Model fit with control variables: χ² = 503.060, df = 94, RMSEA = .029, SRMR = .036, CFI = .976, TLI = .966. Nested in schools: Model fit: χ² = 199.042, df = 70, RMSEA = .023, SRMR = .022, CFI = .984, TLI = .979. Model fit with control variables: χ² = 268.861, df = 94, RMSEA = .024, SRMR = .037, CFI = .979, TLI = .970. * p < .05 (two-tailed). (B) Cumulative Leisure Reading across Different Categories and Self-Reported Peer Problems. Numbers represent standardized coefficients with standard errors in parentheses. Coefficients for models that include control variables (migration background, socioeconomic status, nonverbal intelligence) are printed in bold. Coefficients for the reduced sample of students nested in schools are in italics. Model fit: χ² = 656.876, df = 69, RMSEA = .041, SRMR = .036, CFI = .938, TLI = .919. Model fit with control variables: χ² = 809.302, df = 93, RMSEA = .039, SRMR = .042, CFI = .928, TLI = .898. Nested in schools: Model fit: χ² = 327.287, df = 69, RMSEA = .033, SRMR = .035, CFI = .943, TLI = .925. Model fit with control variables: χ² = 451.301, df = 93, RMSEA = .034, SRMR = .043, CFI = .921, TLI = .887. * p < .05 (two-tailed).