Skip to main content

Table 2 Fit Indices of all Models, and Δχ2 Test Results of Relevant Model Comparisons

From: A latent growth curve model to estimate electronic screen use patterns amongst adolescents aged 10 to 17 years

Variable Model χ2 (df) CFI RMSEA (90% CI) SRMR Δχ2 (df) to previous
Overall (1) Full Accelerated LGCM 286.53 (60), p < .001 .858 .076 (.068, .085) .113 N/A
(2) Multi-Group LGCM 98.20 (48), p < .001 .969 .040 (.029, .051) .048 188.33 (12), p < .01
(3) Sex Added to (2) 121.55 (60), p < .001 .968 .040 (.030, .050) .044 N/A
Social Networking (1) Full Accelerated LGCM Does not converge*     
(2) Multi-Group LGCM 79.30 (48), p = .003 .968 .032 (.019, .044) .045 N/A
(3) Sex Added to (2) 105.36 (60), p < .001 .964 .034 (.023, .045) .042 N/A
Gaming (1) Full Accelerated LGCM 160.40 (60), p < .001 .887 .051 (.041, .060) .098 N/A
(2) Multi-Group LGCM 81.61 (48), p < .001 .962 .033 (.020, .045) .053 78.79 (12), p < .01
(3) Sex Added to (2) 98.44 (60), p = .001 .967 .031 (.020, .042) .049 N/A
Web use (1) Full Accelerated LGCM 151.08 (60), p < .001 .883 .048 (.039, .058) .106 N/A
(2) Multi-Group LGCM 74.51 (48), p = .008 .966 .029 (.015, .042) .052 76.57 (12), p < .01
(3) Sex Added to (2) 94.29 (60), p = .003 .965 .030 (.017, .041) .048 N/A
Passive (1) Full Accelerated LGCM 185.52 (60), p < .001 .872 .057 (.048, .066) .092 N/A
(2) Multi-Group LGCM 75.48 (48), p = .007 .972 .030 (.016, .042) .045 110.04 (12), p < .01
(3) Sex Added to (2) 93.23 (60), p = .004 .972 .029 (.017, .040) .041 N/A
  1. *Covariance matrix (PSI) in one or more groups is not positive definite
  2. Good model fit = CFI values above .95; RMSEA scores of .06 or less; SRMR <.05