Arguably, our estimate underweights large-scale replication projects that report many direct replications in a single paper (eg. Registered Replication Report articles). Our estimate is not the same as the number of studies that are replicated. (9/9)