Replicability, Robustness, and Reproducibility in Behavioral Research

At the end of the module, the students will understand the current research practices and other problems that have contributed to the replication crisis (e.g., p-hacking, HARKing, underpowered studies, publication bias). The students will be able to set up a preregistered study, implement practices of open science (e.g., open data, open analysis code) and know about approaches in data analysis (e.g., Bayesian statistics) that promise greater robustness in statistical inference.

The current replication crisis that has shaken several disciplines in the behavioral sciences raises many important questions about current research and publication practices. In this module, we discuss the history and possible causes of the replication crisis and get to know recent methodological developments and proposals towards a more reliable, robust, and transparent science (e.g., Bayesian data analysis, replication research, preregistration, open data).

There will be presentations in which students present empirical investigations and analyses that have shaped the recent discussion on the replicability of behavioral research. In group discussions, the students will analyze seminal empirical articles and discuss methods for improving the robustness, replicability, and transparency of empirical research. In small-group exercises, students will get hands-on experience with drafting a preregistration document and preparing a repository for making data and analysis code publicly available.