Factor structure and criterion validity of an enlarged version of the parental bonding instrument

Factorial structure and criterion validity of an enlarged version of the Parental Bonding Instrument (PBI-E) were evaluated in a community sample of young adults. This enlarged version was obtained by adding parental favouritism (FAV) and put-down/shaming (PUT_D) to the original care and overprotection (OV) scales as recalled by the offspring. Factor analysis suggested a five factor model as the best solution, identifying CARE, FAV and PUT_D and splitting the overprotection items into two factors, denial of psychological autonomy (DPA) and discouragement of behavioural freedom, (DBF) with Cronbach's alphas ranging from .77 to .92. These five scales were correlated with depression and anxiety of the offspring, measured by BDI and STAI. Both of them correlated negatively with care and positively with the other parental scales, as expected by Parker's theory on the role of affectionless control for the psychopathological vulnerability of the children. A series of hierarchical regression analyses, including CARE, DPA and DBF at the first step and FAV and PUT_D at the second step, showed that the latter enhanced the predictive power of the instrument. Overall these findings: (1) suggest a five factor structure for the PBI-E and (2) confirm the criterion validity of the PBI scales in respect to children's depression and anxiety, providing also compelling evidence for the incremental validity of Gilbert's scales.