Multiple Imputation of University Degree Attainment

Abstract

Historically higher education in the UK has been shaped by a dual system: elite universities on the one hand and polytechnics and other higher education institutions on the other. Despite the for- mal equivalence of both degrees, the two institution types faced different financing, target populations, admission procedures and subjects taught. Nevertheless, in survey data they are often indis- tinguishable. In this paper, we differentiate the institution types among degree-holders using a multiple imputation technique in the UKHLS and BHPS datasets. We examine the validity of inference based on imputed values using Monte Carlo simulations. We also verify that the imputed values are consistent with university graduation rates computed using the universe of undergraduate students in the UK.