A review of high impact journals found that misinterpretation of non-statistically significant results from randomized trials was common
Bibliographic record
Abstract
OBJECTIVES: To determine the prevalence of poor interpretation practices, such as conflating evidence of absence with absence of evidence and over-emphasis of statistical non-significance in abstract conclusions, in a sample of randomized controlled trials (RCTs) with non-statistically significant primary outcomes published after the 2016 American Statistical Association statement on the interpretation of P-values. DESIGN AND SETTING: Review of 50 two-arm individually randomized superiority trials with non-statistically significant results in four high impact journals published between 2017 and 2020, to determine the proportion that conclude evidence of no impact (thus, likely conflating evidence of absence with absence of evidence) or place emphasis on statistical non-significance (technically correct but arguably uninformative) in the abstract conclusion. RESULTS: Of the 50 RCTs with non-statistically significant results for primary outcomes, 28 (56%) of abstract were classified as concluding there was no difference between the two treatments; 19 (38%) placed an over-emphasis on statistical significance; only one acknowledged any uncertainty and the remaining 2 (4%) concluded that one treatment was more effective. Only four studies provided any justification for a finding of no difference, for example that the confidence interval gave no support to values of importance. CONCLUSIONS: RCTs with non-statistically significant primary outcomes almost always present their conclusion in the abstract as evidence of no impact or ambiguously as "not statistically significant" without giving due attention to values supported by the confidence interval.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.260 | 0.761 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.005 | 0.005 |
| Bibliometrics | 0.047 | 0.039 |
| Science and technology studies | 0.002 | 0.005 |
| Scholarly communication | 0.008 | 0.010 |
| Open science | 0.003 | 0.006 |
| Research integrity | 0.004 | 0.003 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".