Researchers

Prof Stephen P. Jenkins, Prof Peter Lynn, Emanuela Sala, Annette Jackle, Dr J. A. Rigg
Institute for Social and Economic Research
University of Essex

1 October 2002 - 31 September 2004

Context

Measures of income and employment from surveys underpin much research and policy analysis. However, derivation of accurate and reliable information is notoriously difficult. This project investigates measurement error with the aim of improving both methods of adjusting for its effects and the quality of future data collection.

Aims and Objectives

• To develop and assess methods of collecting validation data for key survey items regarding income and employment;
• To assess the validity of measures of point-in- time variables and changes over time with respect to these items;
• To provide guidance on the design of future survey validation studies and on the analysis of survey measures of income.

Methodological aspects

These include:

• evaluation of various dependent interviewing strategies
• assessment of respondents’ willingness to give consent to have survey data matched with administrative data and to give NI numbers in conjunction with this
• assessment of the feasibility of matching of survey and administrative records
• assessment of the validity of survey responses through comparisons of matched records

Research Design

• Collection of survey data from a UK sub-sample of the European Community Household Panel Survey (ECHP), using interviews in early 2003;
• Use of data from previous interview waves as auxiliary information for the analysis of measurement error;
• Matching responding sample records with administrative records held by the Department for Work and Pensions, and by employers, to assess the validity of survey data;
• Testing and evaluating dependent interviewing as a method for reduction of measurement error.

Outputs

Dissemination will take place through presentations, seminars, and research papers. An article will be written for the GSS newsletter. Working papers will be made available via the ISER web site.

Papers will also be submitted to academic journals.