MétaCan
Menu
Back to cohort
Record W3109355298 · doi:10.7717/peerj.10387

Comparing multiple comparisons: practical guidance for choosing the best multiple comparisons test

2020· article· en· W3109355298 on OpenAlexaff
Stephen R. Midway, Matthew Robertson, Shane Flinn, Michael D. Kaller

Bibliographic record

VenuePeerJ · 2020
Typearticle
Languageen
FieldDecision Sciences
TopicMeta-analysis and systematic reviews
Canadian institutionsMemorial University of Newfoundland
FundersNational Institute of Food and Agriculture
KeywordsBonferroni correctionMultiple comparisons problemPairwise comparisonWilcoxon signed-rank testComputer scienceSelection (genetic algorithm)Test (biology)Variety (cybernetics)Statistical hypothesis testingStatisticsMachine learningArtificial intelligenceMann–Whitney U testMathematics

Abstract

fetched live from OpenAlex

Multiple comparisons tests (MCTs) include the statistical tests used to compare groups (treatments) often following a significant effect reported in one of many types of linear models. Due to a variety of data and statistical considerations, several dozen MCTs have been developed over the decades, with tests ranging from very similar to each other to very different from each other. Many scientific disciplines use MCTs, including >40,000 reports of their use in ecological journals in the last 60 years. Despite the ubiquity and utility of MCTs, several issues remain in terms of their correct use and reporting. In this study, we evaluated 17 different MCTs. We first reviewed the published literature for recommendations on their correct use. Second, we created a simulation that evaluated the performance of nine common MCTs. The tests examined in the simulation were those that often overlapped in usage, meaning the selection of the test based on fit to the data is not unique and that the simulations could inform the selection of one or more tests when a researcher has choices. Based on the literature review and recommendations: planned comparisons are overwhelmingly recommended over unplanned comparisons, for planned non-parametric comparisons the Mann-Whitney-Wilcoxon U test is recommended, Scheffé’s S test is recommended for any linear combination of (unplanned) means, Tukey’s HSD and the Bonferroni or the Dunn-Sidak tests are recommended for pairwise comparisons of groups, and that many other tests exist for particular types of data. All code and data used to generate this paper are available at: https://github.com/stevemidway/MultipleComparisons .

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 imitation

Not 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.

metaresearch head score (Codex)0.213
metaresearch head score (Gemma)0.590
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesMetaresearch
Consensus categoriesMetaresearch
DomainCandidate signal: Methods · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.787
Threshold uncertainty score0.970

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.2130.590
Meta-epidemiology (narrow)0.0050.005
Meta-epidemiology (broad)0.0060.008
Bibliometrics0.0140.013
Science and technology studies0.0040.009
Scholarly communication0.0100.014
Open science0.0120.008
Research integrity0.0110.022
Insufficient payload (model declined to judge)0.0670.033

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.

Opus teacher head0.870
GPT teacher head0.555
Teacher spread0.315 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.

Study designTheoretical or conceptual
DomainMethods
GenreMethods

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".

Quick stats

Citations429
Published2020
Admission routes1
Has abstractyes

Explore more

Same venuePeerJSame topicMeta-analysis and systematic reviewsFrench-language works237,207