Highlights
- •The appropriateness of using parametric analysis on cold pressor data was studied.
- •In most published studies, normality of data distribution was not mentioned.
- •However, the vast majority of these studies used parametric analyses.
- •Cold pressor test measures collected at 8 independent studies were not normally distributed.
- •Monte Carlo simulations reveal that a nonparametric approach is more appropriate.
Abstract
Parametric statistical methods are common in human pain research. They require normally
distributed data, but this assumption is rarely tested. The current study analyzes
the appropriateness of parametric testing for outcomes from the cold pressor test
(CPT), a common human experimental pain test. We systematically reviewed published
CPT studies to quantify how often researchers test for normality and how often they
use parametric versus nonparametric tests. We then measured the normality of CPT data
from 7 independent small to medium cohorts and 1 study of >10,000 subjects. We then
examined the ability of 2 common mathematical transformations to normalize our skewed
data sets. Lastly, we performed Monte Carlo simulations on a representative data set
to compare the statistical power of the parametric t-test versus the nonparametric
Wilcoxon Mann-Whitney test. We found that only 39% of published CPT studies (47/122)
mentioned checking data distribution, yet 72% (88/122) used parametric statistics.
Furthermore, among our 8 data sets, CPT outcomes were virtually always nonnormally
distributed, and mathematical transformations were largely ineffective in normalizing
them. The simulations demonstrated that the nonparametric Wilcoxon Mann-Whitney test
had greater statistical power than the parametric t-test for all scenarios tested:
For small effect sizes, the Wilcoxon Mann-Whitney test had up to 300% more power.
Perspective
These results demonstrate that parametric analyses of CPT data are routine but incorrect
and that they likely increase the chances of failing to detect significant between-group
differences. They suggest that nonparametric analyses become standard for CPT studies
and that assumptions of normality be routinely tested for other types of pain outcomes
as well.
Key words
To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to The Journal of PainAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- A common situation conducive to bizarre distribution shapes.Am Stat. 1977; 31: 147-150
- Robustness?.Br J Math Stat Psychol. 1978; 31: 144-152
- Survey of chronic pain in Europe: Prevalence, impact on daily life, and treatment.Eur J Pain. 2006; 4: 287-333
- Increasing physicians' awareness of the impact of statistics on research outcomes: Comparative power of the t-test and Wilcoxon rank-sum test in small samples applied research.J Clin Epidemiol. 1999; 52: 229-235
- Statistical Power Analysis for the Behavioral Sciences.2nd ed. Erlbaum, Hillsdale, NJ1988
- Evaluation and comparison of the distributions of gastric pH and hydrogen ion concentration.Anesthesiology. 1987; 67: 391-394
- Evidence-based clinical trial design for chronic pain pharmacotherapy: A blueprint for ACTION.Pain. 2011; 152: S107-S115
- Unmet needs in the management of neuropathic pain.J Pain Symptom Manage. 2003; 25: S12-S17
- Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research.National Academies Press (US), Washington, DC2011
- Current challenges and future prospects in management of neuropathic pain.Expert Rev Neurother. 2008; 8: 1743-1756
- Persistent post-surgical pain and experimental pain sensitivity in the Tromso study: Comorbid pain matters.Pain. 2014; 155: 341-348
- Two-week test-retest stability of the cold pressor task procedure at two different temperatures as a measure of pain threshold and tolerance.Pain Pract. 2014; 14: E126-E135
- Randomization, Bootstrap and Monte Carlo Methods in Biology.2nd ed. Chapman-Hall/CRC, Boca Raton, FL1997
- The unicorn, the normal curve, and other improbable creatures.Psychol Bull. 1989; 105: 156-166
- Human experimental pain models 2: The cold pressor model.Methods Mol Biol. 2010; 617: 165-168
- Worrying about normality.Crit Care Resusc. 2002; 4: 316-319
- The tridimensional personality theory and pain: Harm avoidance and reward dependence traits correlate with pain perception in healthy volunteers.Eur J Pain. 2004; 8: 31-38
- Hand dominancy—A feature affecting sensitivity to pain.Neurosci Lett. 2009; 467: 237-240
- Nonparametric tests of interaction in experimental design.Rev Educ Res. 1990; 60: 91-126
- Comments on using alternatives to normal theory statistics in social and behavioral science.Can Psychol. 1993; 34: 398-406
- Comments on using robust statistics in social and behavioral science.Br J Math Stat Psychol. 1998; 51: 49-52
- Fermat, Schubert, Einstein, and Behrens-Fisher: The probable difference between two means when .J Mod Appl Stat Meth. 2002; 1: 461-472
- Misconceptions leading to choosing the t test over the Wilcoxon Mann-Whitney test for shift in location parameter.J Mod Appl Stat Meth. 2005; 4: 598-600
- New effect size rules of thumb.J Mod Appl Stat Meth. 2009; 8: 597-599
- A more realistic look at the robustness and type II error properties of the t test to departures from population normality.Psychol Bull. 1992; 111: 353-360
- Increased pain sensitivity among adults reporting irritable bowel syndrome symptoms in a large population-based study.Pain. 2013; 154: 385-392
- Analysing the SF-36 in population-based research. A comparison of methods of statistical approaches using chronic pain as an example.J Eval Clin Pract. 2009; 15: 328-334
- Dopamine transporter genotype dependent effects of apomorphine on cold pain tolerance in healthy volunteers.PLoS One. 2013; 8: e63808
- Associations between polymorphisms in dopamine neurotransmitter pathway genes and pain response in healthy humans.Pain. 2009; 147: 187-193
- Exploratory Data Analysis.1st ed. Addison-Wesley, Massachusetts1977
- Developing patient-reported outcome measures for pain clinical trials: IMMPACT recommendations.Pain. 2006; 125: 208-215
- Therapeutic management of chronic neuropathic pain: An examination of pharmacologic treatment.Ann N Y Acad Sci. 2006; 1088: 164-186
- Parametric versus non-parametric statistics in the analysis of randomized trials with non-normally distributed data.BMC Med Res Methodol. 2005; 5: 35
- Guidelines for the cold pressor task as an experimental pain stimulus for use with children.J Pain. 2005; 6: 218-227
Article info
Publication history
Published online: March 22, 2015
Accepted:
March 3,
2015
Received in revised form:
February 27,
2015
Received:
August 25,
2014
Footnotes
Supported in part by the Public Health Service (NINDS K24-NS059892).
None of the authors have any conflicts of interest.
Supplementary data accompanying this article are available online at www.jpain.org and www.sciencedirect.com.
Identification
Copyright
© 2015 American Pain Society. Published by Elsevier Inc. All rights reserved.
ScienceDirect
Access this article on ScienceDirectLinked Article
- Letter to the Editor: Experimental Comparison of Parametric Versus Nonparametric Analyses of Data From the Cold Pressor TestThe Journal of PainVol. 17Issue 1
- PreviewWe read with great interest the recent article by Treiser et al,7 published in the June 2015 issue of the Journal of Pain, entitled “Experimental comparison of parametric versus nonparametric analyses of data from the Cold Pressor Test.” On the basis of systematic reviews and intensive Monte Carlo simulations, Treiser et al noted frequent but inappropriate usage of the parametric t-test and recommended use of the nonparametric Wilcoxon-Mann-Whitney (WMW) test for cold pressor test data, as well as routinely checking assumptions of normality for other types of pain outcomes.
- Full-Text
- Preview
