A psychometric and factorial investigation of the family environment scale

Abstract

Since traditional psychological assessment instruments focus mainly on the characteristics of individuals, techniques for systematically measuring social environments may help to increase assessment validity. One developed assessment technique is the Family Environment Scale (FES) a 90 item inventory with ten subscales that are designed to assess certain features of the social and interpersonal climate of the family along three major dimensions - Relationship, Personal-Growth and System Maintenance. According to Moos & Moos (1981), "It focuses on the measurement and description of interpersonal relationships among family members, on the directions of personal growth that are emphasised in the family, and on the basic organisational structure of the family". The present study attempted to determine whether the ten subscales of the FES do measure ten independent dimensions (factors) of the family environment. The psychometric characteristics of the questionnaire (internal consistencies, subscale reliabilities and item-subscale correlations) and its factor structure were examined using a sample of New Zealand children and young persons (n=370). Results showed that the subscales had varying levels of reliability with the Independence, Achievement-Orientation, Expressiveness subscales achieving low reliabilities (Cronbach's Alpha). The ten subscales did not emerge as ten factors. A six factor solution (incorporating seven of the subscales) was selected as the best solution in fulfilling the notions of parsimony, psychological meaningfulness and simple structure where each item loaded primarily on only a single factor. Caution should be taken in generalising these results. Sampling and demographic features have been influential in this study as they have been in others, eg: young people as opposed to members of entire families completed the FES, and children and parents may perceive family environment dimensions differently. The implications of these findings are discussed in terms of future research attempting to strengthen the validity and reliability of the FES as a measure of perceived family environment.