Comparison of exact tests for association in unordered contingency tables using standard, mid- p , and randomized test versions
The DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index.
Abstract
Pearson’s chi-square (Pe), likelihood ratio (LR), and Fisher (Fi)–Freeman–Halton test statistics are commonly used to test the association of an unordered r×c contingency table. Asymptotically, these test statistics follow a chi-square distribution. For small sample cases, the asymptotic chi-square approximations are unreliable. Therefore, the exact p-value is frequently computed conditional on the row- and column-sums. One drawback of the exact p-value is that it is conservative. Different adjustments have been suggested, such as Lancaster’s mid-p version and randomized tests. In this paper, we have considered 3×2, 2×3, and 3×3 tables and compared the exact power and significance level of these test’s standard, mid-p, and randomized versions. The mid-p and randomized test versions have approximately the same power and higher power than that of the standard test versions. The mid-p type-I error probability seldom exceeds the nominal level. For a given set of parameters, the power of Pe, LR, and Fi differs approximately the same way for standard, mid-p, and randomized test versions. Although there is no general ranking of these tests, in some situations, especially when averaged over the parameter space, Pe and Fi have the same power and slightly higher power than LR. When the sample sizes (i.e., the row sums) are equal, the differences are small, otherwise the observed differences can be 10% or more. In some cases, perhaps characterized by poorly balanced designs, LR has the…
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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 citing paper
“The choice of test statistic for the exact or mid-p tests has little influence on the significance level (results are not shown here).”
From Choice of test for association in small sample unordered r × c tables (2007) · cites this paper as [6]
In the citing paper
“For r × c tables, Lydersen et al. [6] found that the mid-p type I error typically is approximately equal to or slightly less than the nominal level.”
From Choice of test for association in small sample unordered r × c tables (2007) · cites this paper as [6]
In the citing paper
“Lydersen et al. [6] have compared the test statistics for exact conditional, exact conditional mid-p, and randomized tests in small to moderate samples and have concluded that in most practical designs, Pearson's chi-squared statistic and Fisher's statistic perform slightly better than does the LR statistic.”
From Choice of test for association in small sample unordered r × c tables (2007) · cites this paper as [6]
Citations
Cites
Agresti, A. (2002) · Wiley Series in Probability and Statistics · doi:10.1002/0471249688 · Cited by 6,746
Ripley, B. D. (1987) · Wiley Series in Probability and Statistics · doi:10.1002/9780470316726 · Cited by 2,241
Note on an Exact Treatment of Contingency, Goodness of Fit and Other Problems of Significance
Freeman, G. H., & Halton, J. H. (1951) · Biometrika · doi:10.2307/2332323 · Cited by 954 · cited 2×
“Journal of Statistical Computation and Simulation Vol. 75, No. 6, June 2005, 447–458 Comparison of exact tests for association in unordered contingency tables using standard, mid-p, and randomized test versions STIAN LYDERSEN*†, VIVEK PRADHAN‡, PRALAY SENCHAUDHURI‡ and PETTER LAAKE§ †NTNU, Unit for Applied Clinical Re…”
MethodSignificance Tests in Discrete Distributions
Lancaster, H. O. (1961) · Journal of the American Statistical Association · doi:10.1080/01621459.1961.10482105 · Cited by 272 · cited 1×
“(9) In a mid-p test version [8], reject H0 if mid-p-value ≤ α, where mid-p-value = P(T (x) > T (n)) + 1 2 P(T (x) = T (n)) = P(T (x) ≥ T (n)) − 1 2 P(T (x) = T (n)).”
Exact inference for categorical data: recent advances and continuing controversies
Agresti, A. (2001) · Statistics in Medicine · doi:10.1002/sim.738 · Cited by 162 · cited 1×
“However, it is often much lower than α, and this fact causes much of the ongoing controversy about exact test [6].”
BackgroundSignificance Tests in Discrete Distributions
Lancaster, H. O. (1961) · Journal of the American Statistical Association · doi:10.2307/2282247 · Cited by 59
Exact Inference for Categorical Data
Corcoran, C. D., Senchaudhuri, P., Mehta, C. R., & Patel, N. R. (2014) · Wiley eBooks · doi:10.1002/9781118445112.stat04864 · Cited by 49
Self-citationExact Tests for 2 × 2 Contingency Tables
Davis, L. J. (1986) · The American Statistician · doi:10.1080/00031305.1986.10475377 · Cited by 33 · cited 1×
“For 2 × 2 tables with equal row (or column) sums, Pe, LR, and Fi tests always give the same p-values and conclusions [4].”
BackgroundLydersen, S., & Laake, P. (2003) · Statistics in Medicine · doi:10.1002/sim.1671 · Cited by 24 · cited 5×
“Some further remarks and references on the test versions are given in Lydersen and Laake [5].”
BackgroundMethodSelf-citationExact Tests for 2 × 2 Contingency Tables
Davis, L. J. (1986) · The American Statistician · doi:10.2307/2684874 · Cited by 17
Research works printed but not linked: 2
- 2003, A Statistical Software for Exact Nonparametric Inference . StatXact 6 User Manual , Cambridge, MA : Cytel Software Corporation .
- 10.1093/biomet/38.1-2.141 · doi:10.1093/biomet/38.1-2.141 · Paper page
Cited by
Recommended tests for association in 2×2 tables
Lydersen, S., Fagerland, M. W., & Laake, P. (2009) · Statistics in Medicine · doi:10.1002/sim.3531 · Cited by 308
“The same conclusions may be made for rxc tables with one fixed margin [33].”
BackgroundSelf-citationGood practice in testing for an association in contingency tables
Ruxton, G. D., & Neuhäuser, M. (2010) · Behavioral Ecology and Sociobiology · doi:10.1007/s00265-010-1014-0 · Cited by 95
“In particular, the mid-P version of this Fisher–Freeman–Halton test appears to have the best properties (Agresti 2001; Lydersen et al. 2005, 2007 ).”
Strengthening Families Program (10-14): Effects on the Family Environment
Riesch, S. K., Brown, R., Anderson, L., Wang, K., Canty‐Mitchell, J., & Johnson, D. L. (2011) · Western Journal of Nursing Research · doi:10.1177/0193945911399108 · Cited by 70
Choice of test for association in small sample unordered r × c tables
Lydersen, S., Pradhan, V., Senchaudhuri, P., & Laake, P. (2007) · Statistics in Medicine · doi:10.1002/sim.2839 · Cited by 48
“[6] have compared the test statistics for exact conditional, exact conditional mid-p, and randomized tests in small to moderate samples and have concluded that in most practical designs, Pearson’s chi-squared statistic and Fisher’s statistic perform slightly better than does the LR statistic.”
BackgroundMethodResultInfluentialSelf-citationLi, X., Xu, R., & Jiang, X. (2026) · Journal of Psycholinguistic Research · doi:10.1007/s10936-026-10259-4
“…was used to test for association in r × c contingency tables, given its capacity to accommodate the small-sample size dataset and given that the expected value in some cells of the contin - gency table was below 5 (Kim, 2017 ; Larntz, 1978 ; Lydersen et al., 2005 , 2009 ; Lydersen & Laake, 2003 ).”
Metadata
- Authors
- Lydersen, S., Pradhan, V., Senchaudhuri, P., & Laake, P.
- Affiliations
- Norwegian University of Science and Technology (NO)
- Cytel (United States) (US)
- University of Oslo (NO)
- Corresponding author
- Lydersen, Stian
- Venue
- Journal of Statistical Computation and Simulation
- Publisher
- Informa UK Limited
- Published
- 2005-06
- Volume
- 75
- Issue
- 6
- Pages
- 447-458
- ISSN
- 0094-9655 (ISSN-L)
- 1563-5163
- OpenAlex
- W2042803824
- Semantic Scholar
- 97ce01459cb88cead03696343faa0f5698e8e3c2
- Cited by
- 5
- References
- 9
- Citations by year
- 2026: 1
- FWCI
- 0.35
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Journal of Statistical Computation and Simulation (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Probability and Statistical Research
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Applications
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
- Keywords
- randomized tests, Pearson chi-square test, likelihood ratio test, statistical power, type I error, small sample inference
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Comparison of exact tests for association in unordered contingency tables using standard, mid- p , and randomized test versions. https://app.cochat.ai/papers/10.1080/00949650410001729454
- The paper
- Lydersen, S., Pradhan, V., Senchaudhuri, P., & Laake, P. (2005). Comparison of exact tests for association in unordered contingency tables using standard, mid- p , and randomized test versions. Journal of Statistical Computation and Simulation. https://doi.org/10.1080/00949650410001729454
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.