Reproducible Social Research

Presenter(s): Vernon Gayle

decorative image with NCRM logo

There is increasing concern across a wide range of academic disciplines that empirical results cannot be reproduced because of a lack of transparency in the research process. Over the last decade there has been increasing anxiety that it is impossible to verify the results presented in many research papers.

There is a growing interest in the need for researchers to provide additional materials alongside traditional publications to enable other researchers to understand, evaluate and build upon previous research work. The purpose of these materials is to provide sufficient information for a third party, that is unconnected with the original work, to reproduce results without any additional information being provided by the original authors.

Transparency is a central tenet in reproducible research, because without it research cannot feasibly be reproduced. Increasingly transparency in statistically orientated social science research is intrinsically attractive for a number of reasons.

Greater transparency will

  1. Increase the capacity to understand how the research was conducted
  2. Help other scholars evaluate the analyses undertaken
  3. Aid the detection of errors and inconsistencies
  4. Facilitate the incremental development of work
  5. Contribute to limiting negative research practices
  6. Provide extra safeguards against nefarious practices
  7. Improve confidence in results within and beyond the academic community

This resources introduces the concepts of transparent and reproducible social research. It includes a 25-minute video, the associated PowerPoint slides and a list of further reading.

Reproducible Social Research

This 25-minute video introduces the concepts of transparent and reproducible social research. The focus of this video is social science research that employs statistical techniques to analyse observational data (e.g. social surveys). Many of the issues associated with undertaking transparent and reproducible data analysis pervade other forms of social science research (e.g. qualitative data analysis), despite the different nature of the data and the analytical techniques that are used.

   Download transcript    |   Download slides [ 416 Views ]

About the author

My work involves the statistical analysis of large-scale and complex social science datasets. These datasets include both social surveys and administrative data resources. The analysis of longitudinal (i.e. repeated contacts) data is an area in which I specialize. The main substantive focus of my work is social stratification. I have particular interests in the sociology of youth and youth transitions, education and sport. I also have interests in demography, with a focus on migration, and to a lesser extent fertility. I have also undertaken work in the area of digital social research. My methodological research focuses on a range of challenges, which include topics such as quasi-variance estimation, missing data and multiple imputation, and the graphical representation of data. I am attempting to promote the 'Public Awareness of Social Statistics'.

Primary author profile page