ESRC Research Methods Programme

Linkage of survey responses with administrative data records is increasingly being done in the UK. It can potentially provide a route to reducing respondent burden, collecting additional information and reducing measurement error. A number of issues need to be addressed in the data linkage, e.g. what should the matching variable(s) be? Since linkage typically requires explicit consent from respondents, are those who give consent a representative sample?

The aims of this workshop are:

to bring together the various groups with interests in linkage of survey responses with administrative data records, e.g. survey producers, data linkers, researchers and research users, to share experiences and discuss issues;

to report on recent research findings by ISER, a University of Essex team funded under the ESRC Research Methods programme, about matching methods and modelling propensities to consent;

to provide a forum where other research teams with relevant findings may also report on them.

Copies of the presentations given at the workshop are made available below, by courtesy of the presenters. With the exception of the Jenkins et al. paper, please do not quote any presentation without first contacting the presenter (email addresses are provided in the presentations).

Programme

10.30	Coffee and Introduction
11.00 - 12.00	Matching methods (e.g. comparing NINO matches with other matches) and modelling propensities to provide consent to match Stephen Jenkins and the University of Essex 'ISMIE' project team
12.00 - 13.00	Report on the experience of having created a national linked file of hospital episodes and deaths Leicester Gill, University of Oxford and University of Warwick
13.00	Lunch
13.45 - 14.25	The experience so far matching ELSA respondents and administrative records Carli Lessof, National Centre for Social Research, for the ELSA team
14.25 - 14.35	Administrative record linkages experiences and plans regarding the MCS Lisa Calderwood, Millenium Cohort Study, Institute of Education
14.35 - 15.15	Linkage of labour market survey data with administrative records Keith Brook, Office for National Statistics This presentation will cover the recent pilot project which we have undertaken to match the LFS to DWP benefit data. The presentation will include a discussion about: the background and objectives of the matching management of the project to ensure confidentiality subject to ONS protocols matching procedures outcome of the pilot matching exercise potential improvements relating to data quality and matching procedures
15.15 - 15.55	Mapping and tracking income deprivation: creating a longitudinal data set from 100% extracts of IS/JSA(IB) data David McLennan, Social Disadvantage Research Centre, University of Oxford The presentation will cover: aims and objectives of the project information management and security data preparation: cleaning, matching, linking presentation of results current and future developments
15.55	General Discussion
4.30	End

index