From Model Organisms to Humans, the Opportunity for More Rigor in Methodologic and Statistical Analysis, Design, and Interpretation of Aging and Senescence Research

Abstract

This review identifies frequent design and analysis errors in aging and senescence research and discusses best practices in study design, statistical methods, analyses, and interpretation. Recommendations are offered for how to avoid these problems. The following issues are addressed: 1) errors in randomization, 2) errors related to testing within-group instead of between-group differences, 3) failing to account for clustering, 4) failing to consider interference effects, 5) standardizing metrics of effect size, 6) maximum lifespan testing, 7) testing for effects beyond the mean, 8) tests for power and sample size, 9) compression of morbidity versus survival curve-squaring, and 10) other hot topics, including modeling high-dimensional data and complex relationships and assessing model assumptions and biases. We hope that bringing increased awareness of these topics to the scientific community will emphasize the importance of employing sound statistical practices in all aspects of aging and senescence research.

Re-Analysis and Additional Information Needed to Inform Conclusions. Comment on Halenova et al. Deuterium-Depleted Water as Adjuvant Therapeutic Agent for Treatment of Diet-Induced Obesity in Rats. Molecules 2020, 25, 23

“…In addition to group-housing altering standard errors and degrees of freedom for any between-group tests, the cage-level treatment makes any food or water data represent averages across rats resulting from two independent measurements per treatment, not ten. Both scenarios (randomizing by cage and treating animals at the cage-level) need to be accounted for in power calculations to estimate sample size [6,7,9]; a sample size calculation was not reported by the authors. As in any power calculation, power is dependent on variability in the number of independent observations.…”