Constraints | Construct(s) | ∆χ^{2} | ∆df | CFI | RMSEA | |
---|---|---|---|---|---|---|

Configural model A | None | All | 28 332.77^{a} | 2520^{a} | 0.905 | 0.025 |

Construct testing (single construct fixed; remaining constructs vary across time) | Metric invariance: factor loadings fixed equal across time | CON | 36.87^{b} | 6^{b} | 0.906^{b} | 0.025^{b} |

HYP | 151.47^{b} | 8^{b} | 0.905^{b} | 0.025^{b} | ||

EMO | 77.95^{b} | 10^{b} | 0.908^{b} | 0.024^{b} | ||

PEER | 96.32^{b} | 10^{b} | 0.909^{b} | 0.024^{b} | ||

PRO | 137.51^{b} | 10^{b} | 0.907^{b} | 0.025^{b} | ||

Scalar: factor thresholds | CON | 149.06^{b} | 10^{b} | 0.905^{b} | 0.025^{b} | |

HYP | 933.41 | 14 | 0.903 | 0.025 | ||

EMO | 2226.11 | 18 | 0.900 | 0.025 | ||

PEER | 2549.87 | 18 | 0.899 | 0.025 | ||

PRO | 297.15^{b} | 18^{b} | 0.904^{b} | 0.025^{b} | ||

Strong: factor loadings and thresholds | CON | 108.83^{b} | 16^{b} | 0.906^{b} | 0.025^{b} | |

HYP | 551.46 | 22 | 0.903 | 0.025 | ||

EMO | 734.84 | 28 | 0.903 | 0.025 | ||

PEER | 826.81 | 28 | 0.903 | 0.025 | ||

PRO | 338.95^{b} | 28^{b} | 0.906^{b} | 0.025^{b} | ||

Model B | All factor loadings; CON and PRO thresholds | All | 699.18^{b} | 72^{b} | 0.912^{b} | 0.023^{b} |

Model C^{c} | All factor loadings; CON, PRO, and HYP thresholds | All | 940.66^{b} | 86^{b} | 0.911^{b} | 0.024^{b} |

Model D | All factor loadings; all thresholds | All | 22876.52 | 142 | 0.771 | 0.037 |

χ

^{2}Estimates were obtained by using DIFFTEST (Mplus, Muthén & Muthén, Los Angeles, CA).*N*= 16 659. CFA, confirmatory factor analysis; CON, conduct problems; EMO, emotional problems; HYP, hyperactivity; PEER, peer problems; PRO, prosocial.↵

^{a}Configural (baseline) model.↵

^{b}Acceptable change in fit indices (compared with configural model).↵

^{c}Indicates preferred model (model C).