An Italian validation of the Psychological Maltreatment Review (PMR)

PURPOSE: The aim of the current study was to evaluate the construct validity and the internal consistence of the Italian version of the Psychological Maltreatment Review (PMR), and to assess the concurrent validity to provide adequate and reliable instruments to measure retrospectively child psychological maltreatment in the Italian population. METHODS: The participants to our study were 209 patients and 217 nonclinical subjects. The first group consisted in 209 adult patients, 106 males and 103 females (mean age of 41.43 years; SD=12.34) consecutively admitted at the psychiatric unit of the L'Aquila San Salvatore Hospital. The second group consisted in 217 non-clinical subjects, 96 males and 121 females (mean age of 36.38; SD=10.38) that completed an online survey, including the self-report version of the PMR and the Risky Families Questionnaire (RFQ). All subjects were invited to answer to Sociodemographic Information Form and to take a self-report battery composed by two instruments: the Psychological Maltreatment Review (PMR) and the RFQ. RESULTS: Internal consistency reliability analyses were performed separately for the two samples, all the scales had very good internal consistency in both samples, with Cronbach's alpha coefficients equal or greater than .88. An EFA was performed, using exploratory principal axis factoring, on the data of individuals from the non-clinical sample, separately for paternal and maternal scales, followed by varimax rotation and scree testing. The CFA was performed on the data of the patients' sample, separately for paternal and maternal scales, supported a three factor model yielding the best fit indexes, both for paternal scales, χ 2 =725, p<.001; df=402, CFI=0.92; TLI=0.91, SRMR=0.053, RMSEA=0.063, 90% CI [0.056, 0.07]; and maternal scales, χ 2 =758, p<.001; df=374, CFI=0.89; TLI=0.88, SRMR=0.064, RMSEA=0.07, 90% CI [0.06, 0.07]. The CFA performed on the patient' sample supported a three factor model yielding the best fit indexes. The convergence of the EFA and the CFA in different samples supported the structural validity of the PMR and replicated its factorial structure, for both maternal and paternal ratings. CONCLUSIONS: This study provides evidence on the appropriateness of the Italian version of the PMR to retrospectively measure childhood psychological maltreatment. The three PMR scales (psychological abuse, psychological neglect and psychological support) demonstrated good internal consistency with average alpha coefficients, equal or greater than .88. The findings provide evidence of the construct validity, according to the literature, suggesting that psychological abuse, psychological neglect and parental support are dimensionally separated constructs, as defined in the literature.