Research Methods Unit
For more than 25 years, Ipsos MORI has been actively involved in experimenting and improving the techniques used in research. In recent years this component of our work has been brought together under the umbrella of the Research Methods Unit (RMU). The RMU is a forum which brings together experienced researchers from across the company in order to:
- ensure that as a company our work is informed by the latest methodological thinking and findings from across the world;
- initiate and manage new methodological work at Ipsos MORI either alone or in collaboration with academic partners;
- share this knowledge and experience with the wider research community.
The RMU is headed by Dr Patten Smith, Ipsos MORI's Director of Research Methods, and Emma Wallace, Research Director, but includes researchers from across the whole company in its membership.
At the broadest level our work in quantitative research in the Research Methods Unit is informed by the "total survey error" perspective. Quantitative research is concerned with measuring the prevalence of attitudes, behaviours or features of a defined population. A survey error perspective recognises the sources of error that detract from the accuracy of findings across all aspects of the research process. There are two main types of survey error:
- Representation issues These are errors relating to the ability of the research to accurately represent its target audience; as well as sampling error, this includes issues of sample frame coverage, bias arising from non-response and adjustment errors causes by inaccurate weighting strategies.
- Measurement issues These are errors relating to the ability of the research to measure the intended constructs. These errors include inaccurate construction of concepts required for measurement, poor quality questionnaire design leading to respondent miscomprehension or misreporting, administration or research context effects, such as problems of social-desirability bias in respondent reporting, and processing errors at the analysis stage.
The RMU has recently initiated a New Working Paper series which includes both reports of methodological studies we have undertaken and good practice guides.Click here to download papers.
Consultancy projects — Members of the RMU team are regularly involved in providing internal and external methodological consultancy. Click here to review recent external consultancy projects
Survey Lifecycle and Survey Error
One advantage of taking such an approach is that it alerts us to consider all forms of error in surveys, and not just the ones that are easy to measure like sampling error. Indeed the harder to measure forms of error — such as that arising from poorly framed questions — can be far more damaging to accuracy than those we conventionally take into account when presenting confidence intervals.
Another advantage of the approach is that it gives us a sense of perspective about particular forms of error. For example, in genuine random digit dialling telephone surveys, how much should researchers be worrying about the 10% of households that are mobile-only when we know that true response rates for these surveys in the UK are generally under 40%?
Working Papers
The RMU has recently initiated a New Working Paper series which includes both reports of methodological studies we undertaken and good practice guides.
Download papers in the series:
Responding to Sensitive Questions in Surveys: A Comparison of Results from Online Panels, Face to Face, & Self-Completion Interviews (pdf, 207KB)
Best practice in questionnaire design
Cognitive testing research with children
Summary of TSE approach
Download earlier RMU papers
Consultancy projects
Members of the RMU team are regularly involved in providing internal and external methodological consultancy, for government social research clients. Please see below to review recent external consultancy projects.
- British Crime Survey: options for extending the coverage to children and people living in communal establishment residents (2007)
- A study carried out for the Home Office jointly with the National Centre for Social research to look at how the survey population for the British Crime Survey might be extended. The work will shortly be published as a Home Office RDS research report.
- Ofsted and DfES Every Child Matters Survey Development (2006-2007)
- Ipsos MORI has recently completed stage 1 of a project to develop a national survey of children to measure the five Every Child Matters Outcomes at a national and local authority level.
- 2007 TellUs2: Children and young people survey (Word doc, 874KB)
- Student survey feasibility study (2006)
- Carried out for the Higher Education Statistics Agency (HESA), involving graduates from higher education institutions, the large scale survey was designed to test the impact of different approaches on the response rate to the survey such as survey branding, the use of incentives etc.
- HM Revenue & Customs (2007-2008): Development work for a survey of customer perceptions of HMRC
- Ipsos MORI carried out development work to inform the design of a new survey examining perceptions of HMRC among the Department's three main customer groups: individuals, agents (accountants and tax advisers) and businesses.
- Research and consultancy work for the Audit Commission to help inform how local area context can be taken into account in CPA assessments (2006)
- Research built on pioneering work previously undertaken by Ipsos MORI about the impact of place on residents' attitudes to local area and council services.
- Arrestee Survey Review (2007-2008)
- A recently completed study funded by the Home Office to review the methodology and uses of data of the Arrestee Survey.