Retro-priming, priming, and double testing: psi and replication in a test–retest design

This study tackled a deceptively simple question: if someone scores high on a test of 'retro-priming' -- the spooky idea that a picture you haven't seen yet can somehow influence how fast you classify a word right now -- can they do it again? Researcher Thomas Rabeyron ran 162 people through Daryl Bem's backward-priming task, where you sort words before a matching or mismatching image flashes up afterward. The overall result was a dud -- no significant effect. But digging into subgroups turned up hints: students and especially men showed something (men hit a noteworthy correlation of 0.41). So the 28 top scorers were invited back for round two. Here's where it gets really interesting -- and a bit deflating. Not only did those high scorers fail to repeat their success, their scores actually flipped negative. There was a striking negative correlation between sessions, meaning the people who did best the first time did worst the second time. Meanwhile, ordinary forward priming (where you see the picture before the word, the normal way cause-and-effect works) kept humming along just fine with a strong effect. No personality trait or psychological profile could pick out who would be a consistent 'psi star.' This sign-reversal pattern is fascinating because it feeds directly into ongoing debates about why psi effects seem to evaporate on retest -- researchers call this the 'decline effect' -- and whether that vanishing act is itself meaningful or just statistics doing what statistics do.