The Deep Dive
In January 2011, the Journal of Personality and Social Psychology—one of the most respected publications in all of psychology—published a paper by Cornell University professor Daryl Bem titled 'Feeling the Future: Experimental Evidence for Anomalous Retroactive Influences on Cognition and Affect.' The paper described nine experiments involving over 1,000 participants that appeared to demonstrate humans could sense future events before they occurred. In Bem's most striking experiment, participants were shown two curtains on a computer screen and asked to guess which one concealed an image. The computer randomly placed the image behind one curtain only after the participant made their choice. When the hidden image was erotic in nature, participants guessed correctly 53.1 percent of the time—a small but statistically significant deviation from the expected 50 percent. Bem interpreted this as evidence that the emotional arousal of the future image reached backward in time to influence the present decision. Other experiments in the series used variations on well-established psychological paradigms. In one, participants performed better on a memory test for words they would later be asked to practice—as though the future practice session retroactively improved their past performance. The paper used conventional statistical methods (frequentist analysis with p-values below .05) and followed the standard formatting of any mainstream psychology paper. This was precisely what made it so unsettling. Bem had not used fringe methods or published in an obscure journal. He had played by every rule of mainstream psychology and still arrived at an impossible conclusion. The paper forced an uncomfortable question: if standard methods could produce evidence for precognition, what did that say about the methods themselves? Bem was no outsider or crank. He was a tenured professor with decades of respected work in social psychology, best known for his self-perception theory—a mainstream contribution to attitude research. His stature made it impossible for the journal to reject the paper on reputational grounds alone, and the peer reviewers had confirmed the methodology met the field's accepted standards. The paper became a lightning rod for a conversation that had been brewing in psychology for years: whether the standard p-value threshold of .05 was simply too permissive, whether researcher degrees of freedom (choices about analysis that are made after data collection) were inflating false-positive rates across the entire discipline, and whether the publication system's preference for novel, surprising results was distorting the scientific record.
How It Is Used in Marketing
Online psychics who market themselves as offering precognitive readings—predicting your romantic future, career trajectory, or health outcomes—frequently reference the Bem study as peer-reviewed proof that humans possess the ability to perceive events that have not yet happened. The pitch typically goes something like: 'A Cornell professor proved in a scientific journal that we can feel the future. I have spent decades refining that natural ability.' Some sites go further, claiming Bem's research validates specific practices like tarot reading or astrology by framing card spreads as tools that 'tap into the same precognitive faculty Bem documented.' What they omit is the catastrophic aftermath of the paper's publication. You may also see Bem's name invoked on sites that sell precognition training courses, usually promising to help clients develop their 'latent retrocausal abilities' through meditation or visualization exercises priced at several hundred dollars per module. The scientific veneer of a Cornell-published, JPSP-accepted paper is the perfect sales tool for these products.
The Skeptic's Verdict
The fallout from Bem's paper was seismic, but not in the way parapsychology advocates hoped. Three separate teams of researchers attempted to replicate Bem's experiments using his exact protocols and pre-registered their analysis plans in advance. All three failed to find any evidence of precognition. Psychologists Eric-Jan Wagenmakers and colleagues published a Bayesian reanalysis of Bem's original data, arguing that the evidence actually favored the null hypothesis (no precognition) once you accounted for the implausibility of the claim. Stuart Ritchie, Richard Wiseman, and Chris French conducted a particularly rigorous replication of Bem's most compelling experiment and found hit rates indistinguishable from chance, publishing their null results in PLOS ONE after the Journal of Personality and Social Psychology declined to publish the replication—a decision that itself sparked outrage about publication bias in psychology. The Bem affair became a pivotal moment in what is now called psychology's replication crisis. It exposed how flexible statistical practices—running multiple analyses and reporting only the significant ones, adjusting sample sizes mid-experiment, and cherry-picking dependent variables—could produce 'evidence' for virtually anything, including time-traveling erotic images. The lasting legacy of Bem's paper is not evidence for precognition but rather a fundamental reform in how psychology conducts and reports research, including the widespread adoption of pre-registration and open data practices.