Does trial registration make an impact on publication bias? Knowing the answer could earn you a cash prize!
Macartan Humphreys (Columbia, Political Science) and collaborators Albert Fang and Grant Gordon are doing research on how publication (and publication bias) changed after the introduction of registration in clinical trials. They also want you to guess what the changes were. The bidder with the closest guess will win a $200 cash prize. Click here to read more and enter a guess.
Enthusiastic supporters of research transparency are often keen on advocating for the registration of trial experiments. But in the social sciences the practice remains fairly rare and its impact on publication bias is relatively unknown. Fortunately, social scientists can learn from their peers in the medical sciences who have been required to register their medical trials since 2005. The research of Humphreys et al. will look to see if there was a change in the share of p values just below 0.01 and 0.05 before and after 2005 in published medical trials. Their results will provide valuable insight as to whether or not registration should be a high priority on the transparency agenda.