Zimmerman, Katie; (2025) Assessing caregiving representations: A systematic review and empirical study of the psychometric properties of measures. Doctoral thesis (D.Psych), UCL (University College London).

Preview

Text Zimmerman_10205913_Thesis.pdf Download (2MB) | Preview

Abstract

Part 1: Context: To identify and evaluate the psychometric properties of semi-structured interviews measuring parental representations Data source: A systematic review of English articles and theses using PsycINFO, Pubmed, Proquest. Additional studies were identified by hand searching and contacting experts. Search items included Parental representations, internal working model, measures, interviews. Study selection: Only articles and theses which used a measure of parental representation and reported quantitative finding on the measure were included. Data synthesis: Psychometric properties in terms of methodology and quality were assessed for each paper identified in this review. An overall rating of Psychometric Quality and Quality of Evidence was given for each psychometric property of each measure of parental representation. Conclusions: Findings were mixed with some measures of parental representation showing promising results. However generally, psychometric testing needed to be carried out to a better methodological standard, using larger sample sizes. Part 2: Background: The assessment of representational risk (ARR) is a new coding system which is applied to the parent development interview (Slade et al., 2004; Sleed, Isosävi, et al., 2021). It is designed to assess for potential risk in the caregiver and child relationship. Thus far, there are few studies that have explored the psychometric properties of the ARR. Objective: The paper aims to firstly, examine a number of psychometric properties of the ARR in a population of parents with social care involvement including, discriminant validity, internal consistency, concurrent validity compared with the Child Abuse Potential Inventory (CAPI; Milner, 1986); and criterion validity. Participants and Setting: The data was gathered from a larger project called the Supporting Parents Project (Sleed, Fearon, et al., 2021). A total of 110 caregiver-child dyads participated, across five local authorities in England. All children were subject to a child in need, child protection plan or planning law outline (PLO). Methods: The data was analysed using SPSS. For factor structure of the ARR, Confirmatory Factor Analysis (CFA) and Principal Components Analysis (PCA) was used. For internal consistency, Cronbach’s alpha was calculated. Correlations were used for concurrent validity. Criterion validity was tested using independent t-tests. Discrimination validity was tested with t-tests for nominal variables and correlations for continuous variables. Results: CFA revealed a poor fit against previous models (CMIN/DF=3.4). The Cronbach’s alpha coefficients for the hostile, idealizing and the total score were good (α=.699 and α=.772 respectively). The reliability for the helplessness and idealising subscale were acceptable (α=.650 and α=.643). Concurrent validity and criterion validity were good. Discriminant validity was good, except in relation to parent’s education. Conclusions: Findings indicate that the ARR has good psychometric properties to measure risk in the parent-child relationship, with the total ARR score being the most robust use of the ARR for both clinical and research purposes.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item