Register

About the event

This webinar will consider the elements of mixed mode data collection in the Centre for Longitudinal Studies (CLS) cohorts and provide frameworks and relevant empirical evidence to help researchers think about the possible consequences of mode effects in their own analyses.

What’s covered in the event?

Surveys are increasingly moving to mixed mode data collection – for instance, carrying out interviews via face-to-face, telephone, video and/or web. The potential advantages of mixed mode data collection are lower costs, increased efficiency, and higher participation rates.

However, participants’ responses may differ systematically depending on the survey mode used – termed “mode effects”. For instance, the presentation of a survey item either aurally or visually can influence responses, sensitive information may be reported more accurately when given anonymously and complex information may be reported more accurately when an interviewer is present.

Unaccounted for, mode effects may lead to bias in analyses.

This webinar will consider the elements of mixed mode data collection in the CLS cohorts and provide frameworks and relevant empirical evidence to help researchers think about the possible consequences of mode effects in their own analyses. We will describe methods for handling mode effects, discussing their strengths and limitations, and illustrate their application through worked examples. We will conclude with a Q&A session.

The seminar is appropriate for anyone who is analysing CLS cohort data (or other data) which were collected across multiple modes.

Who should attend?

Anyone who is analysing CLS cohort data which involves data collected across multiple modes (with modes differing either within or between cohort members) – this is the case for all recent sweeps of the CLS core cohorts.
This applies across sector, discipline and career stage.

Why take part?

• Find out about the elements of mixed mode data collection in the CLS cohorts.
• Learn about frameworks and relevant empirical evidence to help you think about the possible consequences of mode effects in your own analyses.
• Hear about methods for handling mode effects, including their strengths and limitations.
• Discover where to go for more guidance on handling mode effects.

Webinar presenters

• Richard Silverwood is Associate Professor of Statistics and CLS Chief Statistician. His applied research is mainly within the context of health, and his methodological interests include approaches for handling missing data, the analysis of linked survey and administrative data, and making causal inferences from observational data.
• Liam Wright is Lecturer in Statistics and Survey Methodology at CLS. His research spans survey methodology (mode effects and survey weighting) and the genetics of obesity.

Draft programme

– Introduction (5 mins)
– Elements of mixed mode data collection in the CLS cohorts (10 mins)
– Why do survey items exhibit mode effects? (10 mins)
– Frameworks for thinking about the consequences of mode effects (15 mins)
– Empirical evidence on mode effects and mode selection (10 mins)
– Methods for accounting for mode effects (20 mins)
– Worked example(s) of accounting for mode effects (10 mins)
– Q&A (10 mins)