The New Statistics¶
Cumming, G. (2014). The New Statistics: Why and How. Psychological Science, 25(1), 7-29.
Cited by¶
6 citations across 6 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Confidence Intervals
- The CI framework reveals interconnections with other primes: duality with #434 (hypothesis testing — a 95% CI excludes a value iff a two-sided test rejects at α=0.05); complementarity with #435 (statistical significance / p-value — CI communicates magnitude and precision while p-value communicates binary significance); synergy with #447 (effect size — CI-for-effect-size provides calibrated uncertainty around magnitude estimates); planning value with #437 (statistical power — expected CI width under alternative hypotheses informs sample-size calculations); dependence on #433 (sampling representativeness — valid CIs require design-based variance estimation for complex sampling); alternative to #444 (Bayesian updating — Bayesian credible intervals as conceptually distinct but numerically similar alternative)
This sourceThe defining manifesto of the 'New Statistics' (estimation, effect sizes, confidence intervals, and meta-analysis in place of dichotomous NHST/p-values).
Supported in partVerified against the publisher's abstract
Cumming (2014)'s abstract shows a programmatic advocacy paper for estimation and confidence intervals in place of NHST, topically at home in a CI cluster, but the abstract does not attest that the paper is 'canonical'.
“The new statistics refers to recommended practices, including estimation based on effect sizes, confidence intervals, and meta-analysis.”
- The CI framework reveals interconnections with other primes: duality with #434 (hypothesis testing — a 95% CI excludes a value iff a two-sided test rejects at α=0.05); complementarity with #435 (statistical significance / p-value — CI communicates magnitude and precision while p-value communicates binary significance); synergy with #447 (effect size — CI-for-effect-size provides calibrated uncertainty around magnitude estimates); planning value with #437 (statistical power — expected CI width under alternative hypotheses informs sample-size calculations); dependence on #433 (sampling representativeness — valid CIs require design-based variance estimation for complex sampling); alternative to #444 (Bayesian updating — Bayesian credible intervals as conceptually distinct but numerically similar alternative)
- Effect Size
- This shift reflects the recognition that business decisions depend on effect magnitude, not on significance thresholds
Supported in partVerified against the publisher's abstract
The abstract backs a discipline-wide shift from NHST to estimation via effect sizes, but says nothing about business decisions depending on effect magnitude rather than significance thresholds.
“in response to renewed recognition of the severe flaws of null-hypothesis significance testing (NHST), we need to shift from reliance on NHST to estimation and other preferred techniques. The new statistics refers to recommended practices, including estimation based on effect sizes, confidence intervals, and meta-analysis.”
- This shift reflects the recognition that business decisions depend on effect magnitude, not on significance thresholds
- Hypothesis Testing (Null vs. Alternative)
- A statistically significant result can be trivially small in magnitude; a non-significant result can reflect either small effect or insufficient power
This sourceCumming new statistics effect-size confidence intervals point estimate plus uncertainty reporting discipline.
- A statistically significant result can be trivially small in magnitude; a non-significant result can reflect either small effect or insufficient power
- Statistical Power
- The CI includes zero and is centered near zero; the 5-BPA clinical-threshold for meaningful yield penalty is outside the CI upper bound (+1.7 < 5.0)
This sourceCumming new statistics effect-size confidence intervals point estimate plus uncertainty reporting discipline.
- The CI includes zero and is centered near zero; the 5-BPA clinical-threshold for meaningful yield penalty is outside the CI upper bound (+1.7 < 5.0)
- Statistical Significance (p-Value)
- Contemporary practice often employs the dichotomous threshold as if it were epistemically fundamental, losing continuous information
This sourceCumming new statistics effect-size confidence intervals point estimate plus uncertainty reporting discipline.
Supported in partVerified against the publisher's abstract
Cumming's abstract endorses abandoning null-hypothesis significance testing in favour of estimation with effect sizes and confidence intervals, the general stance the claim's critique of dichotomous thresholding rests on.
“in response to renewed recognition of the severe flaws of null-hypothesis significance testing (NHST), we need to shift from reliance on NHST to estimation and other preferred techniques.”
- Contemporary practice often employs the dichotomous threshold as if it were epistemically fundamental, losing continuous information
- Type I & Type II Errors
This sourceCumming new statistics effect-size confidence intervals point estimate plus uncertainty reporting discipline.
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 3 of 6 citations: 3 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:94721c3ac3f0 · see in the full table