Reproducibility improves exponentially over 63 years of social learning research

Reproducibility is integral to science, but difficult to achieve. We surveyed 560 empirical publications, published between 1955 and 2018 in the social learning literature, a research topic that spans animal behaviour, behavioural ecology, cultural evolution, and evolutionary psychology. Data was recoverable online or through direct data requests for 30% of this sample. Moreover, data recovery declines exponentially with time since publication, halving every 6 years, and up to every 9 years for human experimental data. When data for a publication can be recovered, we estimate a high probability of subsequent data usability (87%), analytical clarity (97%), and agreement of published results with reproduced findings (96%). This corresponds to an overall rate of recovering data and reproducing results of 23%, largely driven by unavailability or incompleteness of data. We thus outline clear measures to improve reproducibility of research on the ecology and evolution of social behaviour.