RESUMEN
This systematic review and reliability generalization meta-analysis synthesized psychometric literature on instrumentation assessing child-to-parent violence published through September 2023 across four databases. In the screening, we identified studies reporting Cronbach's alpha internal consistency estimates for the child-to-parent violence scales. The eligible reliability coefficients ranged from 0.610 to 0.930, mostly exceeding the minimum threshold of 0.700. Random-effects models calculated pooled Cronbach's alphas separately for global, father-specific, and mother-specific subscales. The results demonstrated cumulative values of 0.83 (global: standard error = 0.0129), 0.800 (fathers: standard error = 0.0203), and 0.81 (mothers: standard error = 0.0179), denoting largely adequate reliability. However, significant between-study heterogeneity was observed. While the mean alpha levels seem acceptable for most tools, substantial variability coupled with the possibility of some studies violating reliability assumptions indicates that a conservative interpretation is warranted. Ongoing scale refinement and additional psychometric evaluations will strengthen the rigor methodology in this developing research domain. However, these results should be interpreted with caution, as there is a high level of heterogeneity, and it is possible that some studies have not verified the assumptions underlying Cronbach's alpha.
