One way of making social science more open is to move beyond individual or small research teams working in isolation, and use modern technology to scale up and crowdsource. This component of the MZES Open Social Science Conference 2019 promotes and explores the practice of crowdsourcing in social science studies.

For the OSSC19 Crowdsourced Replication Initiative, we seek researchers to participate in a crowdsourced replication project on a high-profile social science
question: How does immigration shape public opinion?

Besides the theoretical and practical importance of this question, the crowdsourced replication project is relevant to social science for two reasons. One is that research on this topic is thus far inconclusive. The other is a need to explore the epistemic potential of crowdsourcing approaches in macro-comparative social research.

Crowdsourcing is a methodologically innovative means for answering research questions using secondary data. Crowdsourcing replication, expansion, deliberation, experimental practices, and meta-analyses may provide substantial improvements in substantive and methodological findings over what any research team could achieve alone. With this initiative, we aim to explore and develop crowdsourcing as a methodology for the social sciences.

Participation in the project helps promote this goal. Specifically, participating researchers will (a) replicate and (b) expand a previously published cross-national quantitative study. We plan to distribute the results of this collective effort through the Open Science Framework (OSF) and prepare them for publication in a high visibility social science journal. All participants who complete the analytical tasks will be co-authors on the final paper, written under the stewardship of Dr. Nate Breznau.

We invite teams of 1-3 researchers to independently analyze the data. Team members should have solid knowledge in and prior experience with quantitative statistical analysis. The data provided for analysis by the conference co-organizers are both individual and country-level; therefore knowledge of or willingness to engage in macro-comparative research, i.e., identifying effects at ‘level-2’, is desirable. Researchers from all disciplines are welcome to participate, at all levels of research, including professors, postdocs, PhD students, lecturers, and analysts working in non-academic professions. The amount of time needed to participate for each team will depend on their methodological skills. The conference co-organizers will provide all necessary data. We estimate the first part of the project, the replication of the original study, will take no more than six to eight hours on average. The second part of the project, the expansion of the study with new data, will depend on the individual goals of the participating researchers in improving upon the original study but is likely to take a few days of work.

Researchers should apply to take part in the OSSC19 Crowdsourced Replication Initiative by filling out the following form no later than 27 July 2018. In August we will distribute data and instructions. Replications should be completed by 23 September 2018 and expansions by 19 November 2018. Each team’s models and results will be made public in the bundle of all results, but will not be identifiable to any other team or the public. Irrespective of the ‘quality’ of research or ‘capabilities’ of researchers, each team’s participation makes a positive contribution to developing crowdsourcing as a specific open social science practice and our specific crowdsourced initiative.

Please contact Dr. Nate Breznau with any inquiries you may have regarding the OSSC19 Crowdsourced Replication Initiative.