This RAND Europe Working Paper is part of the Measuring Outcomes for Public Service Users (MOPSU) project funded by the UK Treasury under the Invest to Save programme and led by the Office for National Statistics (ONS). The preference study was undertaken as part of the personal social services element of the project, which is led by the Personal Social Services Research Unit (PSSRU), with the objective that the measures of social care outcomes will reflect the relative importance of the domains (e.g. food and nutrition, accommodation) and levels of these domains, rather than an assumption that all domains, and improvements between levels within those domains, are of equal importance. This Working Paper establishes preference weights for the Adult Social Care Outcomes Toolkit (ASCOT) measures of social care outcomes, and examines whether the use of 4-level domains improves the sensitivity of the measure at lower levels of need than 3-level domains.
Table of Contents
Exploratory analysis of best-worst experiments
Modelling the best-worst choice data
Wording of domain levels
Choice elicitation and model estimation procedure in Best-Worst Scaling data