From February 2017, information about the work of the MRC/CSO Social and Public Health Sciences Unit, University of Glasgow is available and updated on the University of Glasgow website.

Missing Data and Attrition

In June 2006 we hosted a workshop, funded by the MRC’s Population Health Sciences Research Network, entitled “Maximising Return from Cohort Studies: Prevention of attrition and efficient analysis”. A report on the workshop, along with several presentations, can be found here. In October 2009 we held a second workshop; the presentations from that workshop can be found here. We are also joint grantholders for two cross-unit appointments considering attrition in MRC-funded cohort studies, again funded by PHRSN.

Missing data are problematic in epidemiology since failure to deal with them adequately leads to inappropriate inferences. Multiple imputation (MI) is widely recognised as an appropriate method for dealing with missing data. We are exploring the importance of MI for large routine data sets, analysing the effect of a missing covariate (maternal height) in the relationship between social class and birthweight from national birth data. See also perinatal health in Scotland

External Collaborators


  • Covariate A variable possibly predictive of the outcome under study.
  • MI Myocardial Infarction
  • Multiple imputation A statistical technique for the analysis of incomplete data sets involving the creation of several possible data sets (based on observed data and observed relationships between variables) and the pooling of results obtained during analysis.
  • Occupational social class The Registrar General's classification of social status based upon an individual's occupation.
  • Perinatal The period just before and just after birth.
View all glossary entries