A lack of statistical significance supports the conclusion that the null hypothesis is true.

the verdict

REFUTED

the evidence says no

Recorded sources

0 sources for · 5 against

Counts group repeated records of the same source within each side. They do not measure evidence strength or source independence.

Is it true that a lack of statistical significance supports the conclusion that the null hypothesis is true?

No. A lack of statistical significance (a non-significant p-value) only indicates that the data are compatible with the null hypothesis under a specific model, not that the null hypothesis is proven true. Bayesian methods and proper uncertainty quantification are required to actually measure evidence in favor of the null.

Verdict: REFUTED · 0 sources for, 5 against · Last checked 4 Oct 2026

The analysis

The claim asserts that a lack of statistical significance supports the conclusion that the null hypothesis is true. In statistical theory (particularly frequentist hypothesis testing), failing to reject the null hypothesis is not the same as proving or supporting it; it simply means there is insufficient evidence against it. Furthermore, modern methodological papers (e.g., papers 0, 2, 8) emphasize that non-significant p-values can be inconclusive and that confirming a null hypothesis requires proper Bayesian analyses (like Bayes factors) rather than merely failing to reach p < 0.05. Therefore, the claim is refuted by statistical consensus.

Evidence against · 5

Beyond the p Value Dichotomy: Alternatives for Statistical Inference-A Critical Review.

2026

Critiques the dichotomous interpretation of p-values, emphasizing that non-significance does not equate to proof of the null hypothesis.

Get citation

MHL, F., LC, C., & NC, J. (2026). Beyond the p Value Dichotomy: Alternatives for Statistical Inference-A Critical Review.. doi.org. https://doi.org/10.1111/jep.70373

See more details

More against · 4

Exploring the proper use of p-values and confidence intervals in leading epidemiology journals.

2026

Emphasizes that non-significant results and confidence intervals require careful contextualization rather than assuming the null is true.

Get citation

MK, H., AJ, R., G, M., & I, B. (2026). Exploring the proper use of <i>p</i>-values and confidence intervals in leading epidemiology journals.. doi.org. https://doi.org/10.1016/j.gloepi.2026.100247

A Bayesian reevaluation of randomized controlled trials in assisted reproductive technology: quantifying evidence strength for null and alternative hypotheses.

2026

Shows via Bayesian reanalysis that non-significant p-values can often be inconclusive regarding the null hypothesis.

Get citation

J, G., T, T., K, K., Y, Y., Y, C., Y, F., R, C., F, L., Y, Z., & Y, W. (2026). A Bayesian reevaluation of randomized controlled trials in assisted reproductive technology: quantifying evidence strength for null and alternative hypotheses.. doi.org. https://doi.org/10.3389/fendo.2026.1860725

The Fragility of Significance: Why P-values and the Fragility Index Are Not Enough.

2026

Argues that null hypothesis significance testing is fundamentally flawed and that Bayesian methods are necessary to properly quantify evidence for hypotheses.

Get citation

M, K. (2026). The Fragility of Significance: Why P-values and the Fragility Index Are Not Enough.. doi.org. https://doi.org/10.1007/s13193-026-02531-9

Reporting of second-generation p-values, Bayes factors and fragility indices in trials enhances assessment of clinical importance beyond p-values: a meta-epidemiologic study.

2026

Demonstrates that non-significant results often fail to reliably distinguish between clinically important effects and true null effects.

Get citation

TM, E., L, M., N, R., N, Y., DJ, S., E, K., JC, L., & L, T. (2026). Reporting of second-generation <i>p</i>-values, Bayes factors and fragility indices in trials enhances assessment of clinical importance beyond <i>p</i>-values: a meta-epidemiologic study.. doi.org. https://doi.org/10.1016/j.conctc.2026.101667

Receipt History

Anyone with this link can read the claim and its public receipt, including any personal information in that text. Open permanent receipt.

Check your own claim

Challenge the receipt

I have a source that contradicts this verdict I have a source that supports it No source: request a receipt review Requests are recorded for review. This does not start an automatic check or guarantee a response time.

See the evidence timeline

Citation formatting by citeproc-js (Frank Bennett) and the Citation Style Language project. Source and licenses.

This receipt is an automated verdict against our published method · not an opinion about any author or publication.
Terms · Privacy · How verdicts work