Yule-Simpson’s paradox (YSP) often occurs when statistical data are aggregated from multiple groups, research sites, or time points. If YSP-type effects go unrecognized, health educators risk yielding erroneous and spurious findings or misinterpreting study data. These paradoxical effects may be avoided if health researchers carefully consider strategies to avoid YSP at the onset of and throughout research studies. This article presents strategies to reduce or avoid YSP-type effects across multiple stages of the research process. A practical checklist is also provided to minimize the likelihood YSP will go unnoticed in research studies.
BakerS. G.KramerB. S. (2002). The transitive fallacy for randomized trials: If A bests B and B bests C in separate trials, is A better than C?BMC Medical Research Methodology, 2(13). Retrieved from http://www.biomedcentral.com/1471-2288/2/13
2.
ClarkA. G.BoerwinkleE.HixsonJ.SingC. F. (2005). Determinants of the success of whole-genome association testing. Genome Research, 15, 1463-1467.
3.
EdwardsA. (2007). A cautionary tale. Significance, 2, 47-48.
4.
GoltzH.SmithM. L. (2010). Yule-Simpson’s paradox in research. Practical Assessment, Research & Evaluation, 15(14). Retrieved from http://pareonline.net/pdf/v15n15.pdf
5.
KraemerH. C.WilsonK. A.HaywardC. (2006). Lifetime prevalence in pseudocomorbidity in psychiatric research. Archives of General Psychiatry, 63, 604-608.
6.
LièvreM.CucheratM.LeizoroviczA. (2002). Pooling, meta-analysis, and the evaluation of drug safety. Current Controlled Trials in Cardiovascular Medicine, 3, 6-9.
7.
ReifsnyderP. C.ChurchillG.LeiterE. H. (2000). Maternal environment and genotype interact to establish diabesity in mice. Genome Research, 10, 1568-1578.
8.
ThompsonB. (2006). Foundations of behavioral statistics: An insight-based approach. New York, NY: Guilford Press.
9.
WilcoxA. J. (2006). Invited commentary: The perils of birth weight–a lesson from directed acyclic graphs. American Journal of Epidemiology, 164, 1121-1123.
10.
BaronR.KennyD. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 113-1182.
11.
KiviluM. w. (2003). Understanding the structure of data when planning for analysis: application of hierarchical linear models. South African Journal of Education, 23, 249-253.
12.
RobertsJ. K. (2004). An introductory primer on multilevel and hierarchical linear modeling. Learning Disabilities: A Contemporary Journal, 2(1), 30-38.
13.
SimpsonE. H. (1951). The interpretation of interaction in contingency tables. Journal of the Royal Statistical Society, 13, 238-241.
14.
SmithM. L. (2009). Instrumentation and psychometric property reporting in current health education literature. American Journal of Health Studies, 24, 232-239.
15.
SmithM. L.HonoréH. H. (2008). Effect size reporting in health education literature. American Journal of Health Studies, 23, 130-135.
16.
YuleG. U. (1903). Notes on the theory of association of attributes in statistics. Biometrika, 2, 121-134.
