Selecting relevant items to measure individuals' wellbeing is a crucial first step in constructing a summary wellbeing index. However, this selection does not address how to aggregate (score) these items or if this aggregation should be identical across individuals. Understanding whether a tailored measure better captures subjective wellbeing is critical for policymaking—especially considering the diversity within the disability community and the cognitive burden of adding importance-rating questions. Using responses from 1,881 Australian National Disability Insurance Scheme (NDIS) participants regarding the 14 items of the Disability Wellbeing Index and their reported importance to overall wellbeing, we explored how incorporating importance ratings into a summary measure may improve correlation with established Global Life Satisfaction measures. We considered five aggregation methods, varying in whether they incorporated importance ratings, whether items marked as “not important” were included, and how importance ratings were used (individual, societal, age-cohort specific or disability-group specific). Our results show that, despite some items being reported as “not important”, correlations are consistently higher when they are included with their importance rating. Second, including individual importance ratings generally leads to a slightly higher correlation, whether using compensatory or non-compensatory approaches. Third, among the arithmetic (compensatory), geometric and harmonic means, the harmonic mean based on individual ratings performed best. While incorporating responses to 14 importance-rating questions for each individual slightly increases correlations, the additional cognitive burden must be carefully weighed against these benefits and using societal importance weighting could be considered instead.
Badji et al. (Sun,) studied this question.