Cited by 331Meta-analysis and systematic reviewsPsychology Research and BibliometricsDecision-Making and Behavioral Economics

Using prediction markets to estimate the reproducibility of scientific research

Open access
Dreber, Anna, Pfeiffer, Thomas, Almenberg, Johan, Isaksson, Siri, Wilson, Brad, Chen, Yiling, Nosek, Brian A., Johannesson, Magnus·2015·Proceedings of the National Academy of Sciences

Abstract

Concerns about a lack of reproducibility of statistically significant results have recently been raised in many fields, and it has been argued that this lack comes at substantial economic costs. We here report the results from prediction markets set up to quantify the reproducibility of 44 studies published in prominent psychology journals and replicated in the Reproducibility Project: Psychology. The prediction markets predict the outcomes of the replications well and outperform a survey of market participants' individual forecasts. This shows that prediction markets are a promising tool for assessing the reproducibility of published scientific results. The prediction markets also allow us to estimate probabilities for the hypotheses being true at different testing stages, which provides valuable information regarding the temporal dynamics of scientific discovery. We find that the hypotheses being tested in psychology typically have low prior probabilities of being true (median, 9%) and that a "statistically significant" finding needs to be confirmed in a well-powered replication to have a high probability of being true. We argue that prediction markets could be used to obtain speedy information about reproducibility at low cost and could potentially even be used to determine which studies to replicate to optimally allocate limited resources into replications.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1073/pnas.1516179112

Cites

79 of 79 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • In the comparison of reproducibility indicators, the fraction of studies that replicated with P<0.05 in the original direction was significantly higher in this experimental economics project (61.1%, 11/18) than in the RPP psychological sciences project (36.1%, 35/97) (chi-square χ2=3.96, P=0.047).

    “To test if the fraction of studies that 'Replicated with P&lt;0.05 in original direction' differed between the studies (61.1% (11/18) versus 36.1% (35/97)) we used a chi-square test that was significant (  2 =3.96; P=0.047).”

    From Evaluating replicability of laboratory experiments in economics (2016) · cites this paper as [26]

    The paper defines successful replication as P<0.05 in the original direction (introductory section, page unavailable), and its Discussion says that more than half of the RPP psychology findings failed to replicate. It does not report the stated 35/97 psychology or 11/18 experimental-economics counts, nor the χ²=3.96 comparison (searched those figures and the economics comparison). (LLM, conf=0.76)

    Experimental — accuracy not yet measured
  • 'Prediction market beliefs' differed significantly between this experimental economics project (75.1%, n=18) and the RPP psychological sciences project (55.1%, n=44) (Mann-Whitney U test, z=3.21, P=0.001).

    “The same test was used to compare 'Prediction market beliefs' between the studies, and this difference (75.1% (n=18) versus 55.1% (n=44)) was significant (z=3.21, P=0.001).”

    From Evaluating replicability of laboratory experiments in economics (2016) · cites this paper as [26]

    The Results report a mean final prediction-market price of 55% for 44 psychology studies, consistent with the cited 55.1% after rounding. Our copy does not report the experimental-economics figure of 75.1% for 18 studies or the between-project Mann–Whitney test (z=3.21, P=0.001). (LLM, conf=0.88)

    Experimental — accuracy not yet measured
  • The prediction market methodology used in this study is similar to the one used in Dreber et al. (26).

    “The prediction market methodology used in this study is similar to the one used in Dreber et al. ( 26 ).”

    From Evaluating replicability of laboratory experiments in economics (2016) · cites this paper as Dreber et al. [26]

  • The RPP prediction market study found a significant Pearson correlation between prediction market beliefs and replication outcomes of 0.42.

    “That study found a significant Pearson correlation between the prediction market beliefs and the replication outcomes of 0.42,”
  • The trading platforms and the participant pool differed between the two prediction market studies (a sample of psychologists participated in the Dreber et al. (26) study and a sample of economists participated in this study).

    “The trading platforms and the participant pool differed between the two prediction market studies (a sample of psychologists participated in the Dreber et al. ( 26 ) study and a sample of economists, see above, participated in this study).”

    From Evaluating replicability of laboratory experiments in economics (2016) · cites this paper as Dreber et al. [26]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 3

  1. Maniadis Z, Tufano F, List JA (2014) One swallow doesn't make a summer: How economists (mis-) use experimental methods and their results. Am Econ Rev 104(1): 277-290.
  2. Hanson R (2006) Decision markets for policy advice. Promoting the General Welfare: American Democracy and the Political Economy of Government Performance, eds Patashnik EM, Gerber AS (Brookings Institution Press, Washington, DC), pp 151-173.
  3. Hanson R (2007) Logarithmic market scoring rules for modular combinatorial in- formation aggregation. J Pred Markets 1(1):3-15.

Cited by

Sorted by citations

Authors
Dreber, A., Pfeiffer, T., Almenberg, J., Isaksson, S., Wilson, B., Chen, Y., Nosek, B. A., & Johannesson, M.
Affiliations
  • Stockholm School of Economics (SE)
  • Institute for Advanced Study (DE)
  • Swedish National Bank (SE)
  • Consensus Point, Nashville, TN 37203;
  • Harvard University (US)
  • Center for Open Science (US)
Venue
Proceedings of the National Academy of Sciences
Published
2015-11-09
Volume
112
Issue
50
Pages
15343-15347
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
OpenAlex
W2145409614
Cited by
331
References
78
Citations by year
2026: 15 · 2025: 21 · 2024: 22 · 2023: 31 · 2022: 53 · 2021: 41 · 2020: 55 · 2019: 34 · 2018: 26 · 2017: 17 · 2016: 14 · 2015: 2
FWCI
19.77
Type
journal article
Language
en
Open access
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion)
  • PubMed Central (publishedVersion)
  • PubMed (publishedVersion)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
Topics
  • Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Psychology Research and Bibliometrics
  • Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
Keywords
prediction markets, reproducibility, replication, Reproducibility Project: Psychology, statistical significance, hypothesis testing, replication study
MeSH terms
Commerce, Forecasting, Probability, Surveys and Questionnaires, Research, Science, Reproducibility of Results

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review