PSYCHOMETRIC ANALYSIS OF THE SLOVAK VERSION OF THE POSTPARTUM BONDING QUESTIONNAIRE FOR NON-CLINICAL SAMPLE

Abstract

"Bonding represents an emotional tie that one experiences towards one´s own child. There are several instruments to measure the level and quality of bonding. Among them we chose and translated the Postpartum Bonding Questionnaire by Brockington et al. (2001) into Slovak language. The aim of this study was to analyse its psychometric qualities. Our non-clinical sample consisted of women (N= 372) 18 and 44 years (M= 29.74; SD= 5.25) who recently gave birth in Slovakian hospitals. Data collection was carried out from September 2015 until March 2018. Participants filled the Postpartum bonding questionnaire by Brockington et al. (2001) and some of them other three tools: Edinburgh Postnatal Depression Scale by Cox, Holgen and Sagovsky (1987), Depression Anxiety Stress scale-42 (Lovibond & Lovibond, 1995) and Parental Stress Scale (Berry & Jones, 1995). The distributions of the items of the bonding were mostly skewed and leptokurtic. Internal consistency is high for the overall Lack of Bonding (?= .897) and varies in factors- ?= .820 for Impaired Bonding, ?= .779 for Rejection and Anger, ?= .506 for Anxiety about Care and ?= .321 for Risk of Abuse. In order to prove convergent validity, we correlated overall Lack of Bonding with depression (Edinburgh Postnatal Depression Scale, rs= .251, Depression Anxiety Stress scale-42 depression rs=.404; n=79), stress and anxiety (Depression Anxiety Stress scale-42, stress rs=.392; anxiety rs=.496; n=79) and parental stress (Parental Stress Scale score; rs= .674, n=99). We did not confirm original factor structure via confirmatory factor analysis using principal axis factoring with oblimine rotation. Then, we used principal component analysis with varimax rotation method to reduce the items. 6 components were extracted. Component 1 was comprised of 15 items that explained 35,6 % of the variance with loadings from .306 to .733. Hence, we proposed new item-structure for the Slovak PBQ. We concluded that the Slovak version of the PBQ proved to have good overall reliability. We found evidences for the convergent validity with parental stress, anxiety, stress, and partly depression, because there were two different results. We also suggest creating a shorter version based on the analysis. Among limits we can see tools used for validity evidence and sample without participants for clinical population. We advise to use the Slovak version of the Postpartum Bonding Questionnaire as a tool to measure bonding in a research context and to use overall summary index (Lack of Bonding) instead of factors."