Research · 06
Feeling the Future and the replication storm that followed_
Nine experiments, one high-profile journal, and a decade of methodological reform — the single most consequential psi paper of the modern era, mostly for reasons unrelated to psi.
- 2011 – 2016
- Daryl Bem; Ritchie, Wiseman & French; Bem, Tressoldi, Rabeyron & Duggan
- Journal of Personality and Social Psychology / F1000Research
- Precognition
- Replication
- Pre-registration
The claim
Bem reported nine experiments in which standard psychological effects — priming, habituation, recall practice — appeared to work backwards in time, with the causal stimulus presented after the response was recorded. Eight of the nine reached significance, with a mean effect size around d = 0.22.
The response
Independent replications by Ritchie, Wiseman and French returned null, and initially struggled to find a journal willing to publish a non-result. That difficulty, more than the original claim, is what pushed the field toward registered reports.
Bem's later meta-analysis of 90 experiments reports a small persistent effect; critics attribute it to flexible analysis and publication bias. The disagreement is unresolved and probably cannot be resolved with the existing corpus.
Why it matters to us
This episode is the reason our defaults are what they are: pre-registered analysis plans, adversarial replication, and publication of nulls on the same footing as positives.
We treat Bem's designs as a source of well-specified tasks rather than as evidence. Any of them we run, we run pre-registered, powered, and with the analysis fixed before the data exists.
Reference: Bem, D. J. (2011). Feeling the future. Journal of Personality and Social Psychology, 100(3), 407–425; Bem, D. et al. (2016). Feeling the future: a meta-analysis of 90 experiments. F1000Research, 4:1188.
Read the study ↗← All researchMore research