Discover and read the best of Twitter Threads about #opensciecne
Most recents (1)
Commentary by Dijksterhuis, author of original study on behaviour priming that failed to replicate in most recent @RegReports, falls into usual post hoc explanations and lack of falsifiability. 1/n psychologicalscience.org/redesign/wp-co… #replicability #opensciecne
All degrees of freedom that are required to make the effect significant post hoc were foreseeable a priori. Eg exclude those who reporting being suspicious of the study's intent. If this data was collected + the original theory state it's an unconscious effect, 2/n
then why wasn't this employed as an exclusion criterion? Author reports not expecting such a high rate of suspicion, but it if was a priori relevant the rate wouldn't matter or could have had a design contingency. 3/n
