Skip to content
Studies / Precognition / Why psychologists must change the way th…

Future Sight? Psychologists Rethink the Data

Eric‐Jan Wagenmakers, Ruud Wetzels, Denny Borsboom, Han L. J. van der MaasJournal of Personality and Social Psychology, 2011 Peer-Reviewed
On this page
✦ Imagine …

Can statistics lie about psychic abilities?

Imagine a psychology experiment that seemed to show people could sense the future — and then imagine the scientific earthquake that followed. In 2011, researcher Daryl Bem published studies suggesting that participants could unconsciously predict random events before they happened, sending shockwaves through psychology. But then a team of statisticians took a closer look at the numbers and declared: 'Wait a minute — this isn't about psychic abilities at all.' What they found would challenge how an entire field conducts research.

Reanalysis of famous precognition study finds no evidence for psychic abilities.

In 2011, Cornell psychologist Daryl Bem published explosive research claiming to prove precognition - the ability to sense future events. His studies with over 1,000 participants suggested people could unconsciously predict what would happen next. The findings sent shockwaves through psychology, prompting fierce debate about both psychic abilities and research methods.

💡

This study showed that even seemingly impressive statistical results can disappear when analyzed with more rigorous methods — revealing a crisis in how psychology research is conducted.

🔍

Key Findings

  • The reanalysis painted a very different picture than Bem's original conclusions.
  • Using Bayesian methods, the evidence for precognition ranged from weak to completely absent.
  • The researchers also found problems with how Bem analyzed his data, including exploratory analysis presented as confirmatory testing.

What Is This About?

These researchers didn't conduct new experiments - instead, they took Bem's original data and reanalyzed it using different statistical methods. They used Bayesian analysis, which weighs evidence differently than traditional p-value testing. They also examined whether Bem's analysis methods were appropriate, looking for signs that he might have tweaked his approach after seeing the data (called 'data fishing').

Methodology

The researchers reanalyzed data from Bem's 9 precognition experiments using Bayesian statistical methods instead of traditional p-value analysis.

Outcomes

The reanalysis found weak to nonexistent evidence for precognition, contradicting Bem's original conclusions that suggested psychic abilities exist.

How Good Is the Evidence?

#

Over 1,000 participants were involved in Bem's original studies - a large sample size that should have provided reliable results. However, the Bayesian reanalysis found the evidence strength was much weaker than traditional statistics suggested.

Anecdotal5/100
AnecdotalPreliminarySolidStrongOverwhelming

Supporters of psi research argue that Bem's studies were well-designed and that dismissing them based on statistical preferences is unfair. They contend that extraordinary claims shouldn't require extraordinary statistical methods. Skeptics counter that extraordinary claims do require more rigorous evidence, and that Bem's analysis methods were flawed from the start. They argue that proper Bayesian analysis reveals the true weakness of the evidence.

↔ Interpretation Spectrum

Mainstream: Statistical reanalysis definitively disproves precognition claims and highlights serious methodological problems in parapsychology. Moderate: The reanalysis raises valid concerns about statistical methods, but the debate reflects broader issues in psychological research rather than definitively settling the psi question. Frontier: Bayesian methods may be overly conservative for detecting subtle psi effects, and the reanalysis represents statistical bias against anomalous phenomena.

Common Misconception

Many people think statistical significance automatically means a finding is real and important. However, this study shows that the same data can appear significant or insignificant depending on which statistical method you use and how you conduct your analysis.

Convincing Checklist
2 of 5 criteria met
Met2/5
Large sample (N>100)
Peer-reviewed journal
Replicated
Significant effect
DOI available

To settle the precognition debate would require large-scale, pre-registered studies with independent replication across multiple labs using agreed-upon statistical methods. This reanalysis study meets the criteria for transparent data use and rigorous statistical examination, but represents only one side of an ongoing methodological debate rather than new experimental evidence.

We conclude that Bem's p values do not indicate evidence in favor of precognition; instead, they indicate that experimental psychologists need to change the way they conduct their experiments and analyze their data.

Stance: Skeptical

What Does It Mean?

This paper turned a study about potential psychic abilities into a mirror that reflected deep flaws in scientific methodology itself. The real 'extraordinary phenomenon' wasn't precognition — it was how easily our statistical tools can fool us into seeing patterns that aren't really there.

It's like having two accountants review the same financial records and reaching opposite conclusions about whether a company is profitable - the numbers are the same, but the methods of analysis lead to different interpretations.

If this critique is correct, it suggests that many published psychological findings might be statistical mirages rather than real discoveries. This would mean that our understanding of human behavior, built on decades of research, might need fundamental revision. The implications extend beyond psychology to any field that relies on statistical significance to validate findings.

🎓
Science Literacy Tip

The same data can tell different stories depending on which statistical lens you use to examine it - this is why pre-registering analysis plans before seeing the data is crucial for reliable science.

Understanding Terms

📖
Bayesian Analysis
A statistical method that updates the probability of a hypothesis as new evidence becomes available, often more conservative than traditional p-value testing
📖
P-hacking
The practice of manipulating data analysis to achieve statistically significant results, also called data fishing or cherry-picking
📖
Confirmatory vs Exploratory
Confirmatory studies test pre-planned hypotheses, while exploratory studies look for patterns in data - mixing the two can lead to false discoveries

What This Study Claims

Findings

Bayesian reanalysis shows evidence for psi is weak to nonexistent

strong

Methodology

Controversial claims require strictly confirmatory studies and conservative statistical tests

moderate

One-sided p values may overstate the statistical evidence against the null hypothesis

moderate

Bem's data analysis was partly exploratory rather than strictly confirmatory

moderate

Implications

Experimental psychologists need to change how they conduct experiments and analyze data for controversial claims

moderate

This summary is for general information about current research. It does not constitute medical advice. The scientific interpretation of these results is debated among researchers. If personally affected, please consult qualified professionals.