Publications by authors named "Eskil Forsell"
Article Synopsis
- The study examines how well black-box statistical models can predict the replication accuracy of experimental results from psychology and economics.
- Using data from four large-scale replication projects and machine learning techniques, the models achieved a prediction accuracy of 70% for binary replication outcomes.
- Key factors influencing replication success include sample sizes and reported effect sizes, suggesting that these models could streamline the evaluation of new findings in research.
Article Synopsis
- - Replicated 21 experimental studies from reputable journals (Nature and Science, 2010-2015) using rigorous methods, including pre-registered analysis plans and larger sample sizes, averaging five times bigger than the originals.
- - Out of the replications, 62% showed effects consistent with the original studies, but the average effect size was about 50% of what was originally reported.
- - The study found that researchers’ beliefs about replicability were linked to actual outcomes, indicating that the scientific community can often predict which results are likely to replicate, suggesting that failures are not purely due to random chance.
The replicability of some scientific findings has recently been called into question. To contribute data about replicability in economics, we replicated 18 studies published in the American Economic Review and the Quarterly Journal of Economics between 2011 and 2014. All of these replications followed predefined analysis plans that were made publicly available beforehand, and they all have a statistical power of at least 90% to detect the original effect size at the 5% significance level.
View Article and Find Full Text PDF