in which we fail to replicate our own high-impact study [ https://osf.io/8vwd7/ ] 1/7

In 2007 @vhsouthgate, Atsushi Senju and I reported that 2-year-olds anticipate someone’s impending action on the basis of a false belief attributed to her ( https://is.gd/dDNbsm ). Since then, this study has been cited close to 1000 times according to Google Scholar. 2/7

While this finding has been replicated in various populations, including non-human primates, there were also several reports of non-replication. However, none of these replication attempts were close enough to the original study. 3/7

We (with @dora_kampis, Petra Kármán, @vhsouthgate, and @MikolajHernik) have attempted a direct replication of the study using a relatively large sample across two labs. We found no evidence of action anticipation or belief attribution. 4/7

The details of the study, including links to all the data we collected and the analyses we performed, are found in the preprint, and will also be reported at #BCCCD21 next week ( http://bcccd.org ) 5/7

While this result is certainly not the last word in determining whether young children can attribute false beliefs to others, it knocks off a crucial piece of evidence that has so far supported arguments for the existence of this competence early in life. 6/7

The #Manybabies2 project ( http://manybabies.stanford.edu/MB2/ ) will test further whether action predictions derived from attributed epistemic states are manifested in children’s anticipatory looks. 7/7