Measuring choice for adults with an intellectual disability – a factor analysis of the adapted daily choice inventory scale. Issue 5 (9th March 2017)

• Record Type: Journal Article
• Authors: O'Donovan, M.‐A.; Byrne, E.; McCallion, P.; McCarron, M.
• Is Part Of: Journal of intellectual disability research. Volume 61:Issue 5(2017)
• Page Start: 471
• Page End: 487
• Publication Date: 2017-03-09
• Subjects: choice -- factor analysis -- intellectual disability -- living arrangement; Mental retardation -- Research -- Periodicals; 362.2
• DOI: 10.1111/jir.12364
• Languages: English
• ISSNs: 0964-2633

Abstract

Abstract: Background: For most people, choice making is an everyday occurrence, but for adults with an intellectual disability (ID), such opportunities are often limited, if not, absent. Defining choice, and related opportunity capacity and supports continue to feature prominently in academic, practice and policy discourse within the field of ID as reflected in the range of measures available. This paper examines the factor analytic properties of an adapted 14‐item choice inventory scale. Method: Presence and type of choice were recorded in wave 1 of the Intellectual Disability Supplement to the Irish Longitudinal Study on Ageing using a choice inventory scale adapted for the Irish context for 753 participants with ID over age 40 years. Analysis included both an exploratory and confirmatory factor analysis. Descriptive statistics on choice by type of living arrangement, type of interview (proxy, self or supported) and level of ID are presented. Results: Exploratory factor analysis indicates good model fit when using both a 3‐item and 4‐item response with the 4‐item version suggesting a two‐factor model. Further exploration of this two‐factor model through confirmatory factor analysis highlighted an improved fit for the 4‐item model. Further improvement in model fit is found when four item pairs are co‐varied within the model. Conclusion: Two broad types of choice were found to exist for adults with ID – everyday decisions and key life decisions. In addition, the factor analysis support for the inclusion of a 'no choice' response may help reduce the potential for missing data.