The “Rolling Metaketa” Approach to Knowledge Accumulation

• In Event: The Problem of Knowledge Cumulation

Abstract

In recent years, social scientists have grown increasingly concerned about challenges that make it difficult to learn across studies—including heterogeneity in study design, effect heterogeneity, and selective reporting—and about challenges for establishing the external validity of empirical findings. These are features of the so-called “replication crisis.” At the same time, developments in “data fusion,” causal modeling, and extrapolation methods offer ways to formally assess strategies for accumulation. We present a new approach to cross-study collaboration and integration. A starting point for our efforts has been the “Metaketa” model, pioneered by the Evidence in Governance and Politics (EGAP) network. This model involves the simultaneous coordination of a set of field experiments with harmonized designs in multiple sites that feed into a meta-analysis. Our follow-up initiative offers researchers the opportunity to collaborate in asynchronous ways. We build a platform which serves as a hub that summarizes existing knowledge within a research program. Accessing this hub, researchers are able to design their studies in ways that maximize learning and to feed their results into the existing knowledge base. An overarching theoretical structure aids the aggregation of research findings. We demonstrate our approach by applying it to research on intergroup contact and prejudice reduction.

Authors

• Macartan Humphreys, Columbia University

• Cyrus Samii, New York University

• Alexandra Scacco, WZB Berlin Social Science Center

• Anna Maria Wilke, Columbia University

• Julio Saul Solís Arce, WZB Berlin Social Science Center