The omnipresence of non-response in longitudinal studies is addressed by assessing the accuracy ofstatistical models constructed to predict different types of non-response. Particular attention is paid tosummary measures derived from receiver operating characteristic curves and logit rank plots as ways ofassessing accuracy. The ideas are applied to data from the first four waves of the UK Millennium CohortStudy and the results suggest that our ability to discriminate and predict non-response is not high.Changes in socio-economic circumstances do predict wave non-response with implications for theunderlying missingness mechanism. Conclusions are drawn in terms of the potential of interventions toprevent non-response and methods of adjusting for it.
Original language | English |
---|
Number of pages | 19 |
---|
Publication status | Published - Dec 2010 |
---|
Name | CCSR Working Papers |
---|
No. | 08 |
---|
- Longitudinal studies; missing data; attrition; propensity scores;