| Popis: |
This note pre-registers an exact procedure (and code) for producing out-of-sample replication rate predictions for the MTurk Replication Project (Camerer et al. 2022). This procedure will be used to generate replication rate predictions for the 26 studies chosen for replication among the 41 studies selected as eligible. The prediction procedure and code was published on the 19 February 2019 at OSF, prior to replication outcomes being announced. When the replication outcomes are released, I will compare the predicted replication rate to the observed replication rate on the subset of studies chosen, using the exact procedure outlined here. As an example, I have used the pre-registered code to produce an out-of-sample prediction of the replication rate using the full sample (some of which will not be replicated). |
