The Berkeley Initiative for Transparency in the Social Sciences (BITSS) is an international network of researchers and institutions committed to improving the standards of openness and integrity in economics, political science, psychology, and related disciplines.
Central to BITSS efforts is the identification of useful tools and strategies for increasing transparency and reproducibility in research, including the use of study registries, pre-analysis plans, version control, data sharing platforms, disclosure standards, and replications.
The demand for more rigorous and credible research has led to prodigious growth in empirical social science methods. However, problems arising from data mining, specification searching, selective presentation of results, publication bias, and inadequate documentation of hypothesis testing strategies undermine the rigor of this work and call many research findings into question.
BITSS grew out of a December 2012 conference organized at the University of California, Berkeley, in order to discuss solutions to these issues. The convening brought together a range of social science researchers leading independent, but related, transparency initiatives – as well as bio-statisticians and medical researchers familiar with existing efforts in their respective discipline.
At the end of the event, participants expressed interest in continuing and expanding these efforts by creating a formal initiative, so as to rapidly and decisively shift research norms in favor of more transparent practices.
BITSS is a program of the Center for Effective Global Action (CEGA). CEGA designs and tests solutions for the problems of poverty, generating actionable evidence for policy-makers in less developed countries. Using rigorous field trials, behavioral experiments, and tools from data science, we measure and maximize the impacts of economic development programs throughout the world.
BITSS seeks to achieve these objectives by:
- Raising awareness about systematic issues in the practice of empirical social science research, with the goal of reaching a greater consensus among the scientific community.
- Identifying practical resources and tools for increasing openness and reproducibility, in close partnership with data scientists and empirical researchers.
- Promoting the dissemination and adoption of effective approaches, through training courses, grant competitions, and coordination with funders, government agencies, professional societies, and academic journals.