**Undertaking Transparent and Reproducible Longitudinal Data Analysis**
2nd July 2019,
Understanding Society Conference,
University of Essex
Across a wide range of academic disciplines there is increasing concern that empirical results cannot be reproduced because of a lack of transparency in the research process. These concerns have led to numerous calls for social researchers to undertake data analyses more transparently and for results to be reproducible.
This workshop is a short practical introduction to undertaking transparent and reproducible research with longitudinal data. Using illustrations from Understanding Society and the British Household Panel Survey the workshop will introduce participants to the fundamental concepts of undertaking transparent social research that is reproducible. The focus of the workshop will be developing and executing a transparent workflow and generating reproducible results. The workshop will draw on insights from data science and e-research that can inform the social science longitudinal data analyses, such as the concept of ‘literate programming’. The workshop will also introduce participants to tools that assist in developing and executing an efficient, transparent and reproducible workflow in longitudinal data analysis (e.g., using Jupyter notebooks and git repositories).
Programme:
15:30 – 1545 Roxanne Connelly: Transparency and Reproducibility in the Social Sciences. What is the problem? What are the solutions?
1545 – 1615 Vernon Gayle: Reproducible Workflows
1615 – 1630 Break
1630 – 1700 Roxanne Connelly: Publishing Reproducible Research
1700 – 1730 Chris Playford: The Challenge of Reproducible Research with EUL and Secure Data
