Possibilities of a Narrative Analysis for Paradata

Possibilities of a Narrative Analysis for Paradata: An Historically Situated Exploration

Background and Aims
Paradata captures the gamut of by-products of the collection of survey data and is of interest in understanding and improving survey quality and costs. The main focus has been on automatically captured macro items, but this is now expanding to include interviewer-generated observations. What can be done with such observational data and how best it can be analysed is an emerging area. The overarching aim of this project was to investigate the possibilities of narrative analysis for paradata, and attendant ethical issues, through working with historically situated archived data.

Specific objectives were to:

  • explore the possibilities of narrative analysis for micro-level marginalia from a sub-set of Townsend's Poverty in the United Kingdom archived survey material;
  • extend understanding of secondary narrative analysis with data not collected for this purpose;
  • place the historically situated narrative analysis in the context of analysis of macro-level data of the archived material;
  • explore ethical tensions for survey interviewers and the implications for contemporary survey fieldwork; and
  • pursue the potential of this project for informing a framework for the collection and understanding of contemporary paradata, with a 'community of interest'.

Study Design
Peter Townsend's influential Poverty in the United Kingdom: A Survey of Household Resources and Standards of Living 1967-69 data set comprises around 3000 survey booklets with information about household resources and levels of deprivation/style of living and is housed at the UK Data Archive at the University of Essex.   Selecting 69 of the survey booklets, based on geographical location, interviewer id, and presence of extensive or noteworthy marginal notes, we developed a typology for marginalia and undertook thematic analysis. We then undertook a narrative analysis of six booklets focused on the conveyance of ideas about identity and behaviour, and positioning of researcher and researched. Ethical analysis focused on the interview relationship and survey research practices of the period.

Findings
The study showed that narrative analysis on paradata could illuminate the historical specificities of the period in which interviews were conducted as well as the interview process and informants' accounts and enhance and develop thematic analyses.

Publications and outputs can be found here