Reproducibility Project

A team of psychologists hopes “to illuminate the extent to which studies fail when they are reproduced by a different set of researchers,” reports Sally L. Satel in The New York Times (2/24/13). Their impetus is the discrepancy between original research on the “goal-priming effect” and subsequent efforts to reproduce it. The “goal-priming effect” is when “subjects automatically and unintentionally alter their thoughts or behavior when prompted by various kinds of information.”

It’s a type of research made famous by Malcolm Gladwell in his best-selling book, Blink, such as the 1996 study in which NYU students walked more slowly after being “primed” by “words typically associated with older people, like ‘Florida,’ ‘bingo,’ and ‘gray’.” The suggestion is that “if stimuli we are not aware of can influence us, then perhaps we are not as accountable for our actions as others might want to hold us.” The problem is that an attempt to replicate this research “found no difference of walking between goal-primed and unprimed subjects.”

Some of the discrepancy may be because of “little tweaks in research design.” Or, perhaps the original “finding was modest to begin with”– there was “only a one-second difference in walking duration between primed and unprimed students,” for example. Then there’s the tendency of scientific journals to publish only positive results and a bias against reproduced experiments. The Reproducibility Project hopes to address this and perhaps blunt the “glut of neat results that are long on mass appeal but short on scientific confirmation.”