Transparent and Reproducible Social Science

Social science fields have each taken their turn in the spotlight with instances of influential research that fell apart when scrutinized. Beyond deliberate fraud, there is growing evidence that much social science research features sloppy yet inadvertent errors, and a sense that many analyses produce statistically “significant” results that are wrong. Due in part to a rising number of highly publicized cases, there is growing demand for solutions. A movement is emerging across disciplines towards greater research transparency, open science, and reproducibility.
Researchers have developed new tools for combating false positives and non-reproducible findings. Miguel and Christensen developed this textbook to describe and synthesize these important new methods. More researchers are conducting meta-analyses, pushing to reform the journal peer review process to focus on good research design rather than on “sexy” results, and posting pre-specified hypothesis documents in public registries, all to curb rampant publication bias. New software tools make it easy to implement version control with dynamic documents that can reproduce an entire research workflow with a single mouse click, and data repositories make it simple to download data, encouraging the replication and extension of previous work.
This textbook is intended to serve as a companion to a semester-long course on transparency and reproducibility in social science research. University of California Press Press will publish the textbook in Spring 2019.