Mixed Modes and Measurement Error, 2009

<p>Abstract copyright UK Data Service and data collection copyright owner.</p>The aim of the <I>Mixed Modes and Measurement Error</I> study was to increase our understanding about the causes and consequences of mixing modes in order to improve survey research quality, and to provide practical advice on how to improve portability of questions across modes, in particular to answer the following questions: which mode combinations are likely to produce comparable responses? And which types of questions are more susceptible to mode effects? The project ran from 2007-2011, with data collection taking place in 2009.<br> <br> Increasing pressures of falling response rates and rising costs of survey operations have led many to explore the potential benefits of combining different modes of survey data collection, such as face-to-interviewing, telephone interviewing, postal surveys and web surveys. The drawback of using more than one mode is that the data may not be comparable if people give different answers depending on the mode of data collection. There is a need for practical advice to inform decisions about when and how to mix modes, since survey designers are making these decisions in an ad hoc manner, driven by considerations of costs and response rates, but often ignoring the potential impact on data comparability.<br> <br> <b>Constructing the sample</b><br> The samples for the mixed mode experiment consisted of respondents from two previous surveys who had agreed to be re-contacted:<br> 1. The <I>NatCen Omnibus</I> survey (not currently held at the UK Data Archive; two rounds of data collection administered in July/August 2008 and September/October 2008. The NatCen Omnibus survey is based on a probability sample of adults aged 16 and over in Great Britain, whereby clients are able to buy questionnaire space on topical issues. The survey is administered quarterly to a fresh sample of respondents and 1,600 interviews are administered face-to-face using CAPI (Computer Assisted Personal Interview). <br> 2. The <I>British Household Panel Study</I> (BHPS) (held at the Archive under SN 5151); a sub-sample of Wave 18 respondents (surveyed September-December 2008). The BHPS has become part of the UK Household Longitudinal Survey now known as 'Understanding Society'. It is managed by the Institute for Social and Economic Research at the University of Essex and is funded by the Economic and Social Research Council. Its main objective is to further understanding of social and economic change at the individual and household level in Britain and the UK. It is based on an original probability sample of 5,000 households in Great Britain in 1991. Individuals from these households have continued to be followed annually ever since, and are therefore seasoned panel members. The interviews are conducted face-to-face using CAPI.<br> <br><B>Main Topics</B>:<br>The main purpose of data collection was methodological rather than substantive. It was essential that the survey should cover the full range of question types relevant to the hypotheses being tested (e.g. difficult versus easy questions, long response lists versus short response lists, branching, code all that apply questions, agree/disagree scales as well as other rating scales, fully labelled versus end labelled scales). Wherever possible, existing questions from other surveys were included in the questionnaire covering a range of topics such as management of personal finances, personal health, national identity, views about Great Britain (e.g. democracy, standard of living and economy), and views about their neighbourhood