Exploratory and Confirmatory Examination of the Factor Structure of the Parental Discipline Questionnaire with Adult Violent Offenders

Background: Intimate partner violence places children at risk for maltreatment. It is important for researchers and clinicians to assess their co-occurrence given the high rates at which these take place. The Parental Discipline Questionnaire is a measurement instrument capable of assessing risk factors associated with child maltreatment. This study examined the factor structure of the Parental Discipline Questionnaire (PDQ) with a sample men and women charged with intimate partner violence.

Methods: The PDQ scale was administered to individuals charged with assault against an intimate partner (N=1056).  The sample was randomly split into two halves and the psychometric properties of the scale were assessed using an exploratory factor analysis (Sample 1) and confirmatory factor analysis (Sample 2).  For Sample 1 (N=535), the mean age was 33.06 years (SD= 9.71); 87.8% were male and 12.2% were female; the majority were White, non-Hispanic (48.8%), followed by Black (23.2%), Hispanic (21.7%), and other (4.9%). For Sample 2 (N=529), the mean age was 32.72 (SD=8.94); 85.1% were male and 14.9% were female; the majority of the participants were White, non-Hispanic (45.2%), followed by Black (21.6%), Hispanic (28.5%), and other (4.7%).

Results: A principal-axis factor analysis, using direct oblimin rotation, was used to explore the factor structure of the PDQ. The Kaiser-Meyer-Olkin measure of sampling adequacy was .88, indicating that the items shared a sufficient amount of common variance for factor analysis. Furthermore, Bartlett’s test of spherecity was significant, indicating that the items relate to each other enough to warrant factor analysis, χ² (136, N=535) = 2217.58, p<.001. A visual scree test was used to determine how many factors to extract (Cattell, 1966). Three factors were extracted with eigenvalues of 4.87, 2.11, and 1.29 respectively. As indicated on the pattern matrix, each of the 17 items loaded at least .30 on one and only one factor. Factor 1 contains 11 items and is concerned with parents’ approval of spanking as a punishment. Factor 2 contains 3 items and is concerned with the rights/freedom of children. Finally, Factor 3 contains 3 items and is concerned with children obeying parental authority. A confirmatory factor analysis using Full Information Maximum Likelihood estimation was employed to examine the 3-factor solution observed above. The 3-factor model demonstrated acceptable fit, χ² (116, N=529) = 523.64, p<.001, χ²/df = 4.51, CFI = .79, RMSEA = .08 (90% CI = .07, .09), SRMR = .07.

Implications: The findings offer preliminary insight towards the applicability of the PDQ as a 3-factor assessment tool and indicates the potential use of the PDQ in both practice and research settings with adult offenders. Future studies should further explore the use of the PDQ as one tool for evaluating the outcomes of specific parenting programs designed with this population in mind. Likewise, the PDQ might be used in cross-training professionals around their knowledge and understanding of this co-occurrence.