This thesis is concerned with the statistical adjustment of survey-based indicators to account for unobserved and observed sources of heterogeneity. Recent years have seen a growth in the use of survey-based indicators to measure performance, but questions have been raised over their legitimacy due to high levels of nonresponse, particularly among certain groups, and the influence of factors unrelated to organisational performance, which complicate their interpretation. In light of this, this thesis uses a range of methods that go beyond those ordinarily applied to performance assessment, to explore the role that nonresponse and factors unrelated to performance, i.e. case-mix, have on indicators. The empirical analysis focuses on the Adult Social Care Outcomes Framework (ASCOF) indicators drawn from the English Adult Social Care Survey. The core concerns of this thesis are whether (i) nonresponse and (ii) adjusting for factors beyond the control of organisations affects the interpretation of indicator scores. Nonresponse has a limited effect on inferences about performance, but conclusions depend on the method used to explore the effects of nonresponse, the level of nonresponse, the importance of unobserved factors and the value placed on accuracy over intelligibility of indicators. Adjustment for case-mix has an important effect on the interpretation of indicators, but the adjustment method used was less critical for inference, at least where the aim is to compare organisations. This thesis suggests that the accuracy of some of the ASCOF indicators would be improved by adjusting for case-mix and, possibly, for nonresponse. It is important for future studies to explore the effect of nonresponse on indicators. Policymakers may also wish to consider amending the survey design to improve its representativeness of the adult social care population. Future studies of survey-based performance indicators would benefit from using a wider range of methods similar to those applied here.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:734005 |
| Date | January 2017 |
| Creators | Malley, Juliette Nicola |
| Publisher | London School of Economics and Political Science (University of London) |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://etheses.lse.ac.uk/3638/ |
Page generated in 0.0019 seconds