Examining Statistical Power and Sample Size Adequacy in Published Health Research from Nigerian Universities

Underpowered studies with inadequate sample sizes produce unreliable effect estimates and contribute to a reproducibility crisis in health research, yet the statistical power and sample size adequacy of published health studies from Nigerian universities has never been systematically reviewed. This study examines statistical power and sample size adequacy in a systematic sample of quantitative health research articles published by Nigerian university authors from 2019 to 2023. A systematic literature review identified 120 eligible quantitative studies from Nigerian journals indexed in AJOL and PubMed. Sample size and effect size data were extracted, and post-hoc power analyses were conducted for each study using the reported sample size and estimated effect size under the primary hypothesis test. G-Power software implemented power computations for each study design type. The median reported sample size was 204 (IQR 120 to 350). Post-hoc power analysis showed that 62 percent of studies had power below 0.80 for detecting the effect sizes they reported as significant. Studies with non-significant findings had a mean power of only 0.41, suggesting many null findings reflect inadequate power rather than true null effects. Sample size justifications were reported in only 38 percent of articles. Cross-sectional surveys had better power characteristics than case-control studies. The study documents a systematic sample size adequacy problem in Nigerian university health research and recommends mandatory reviewer assessment of power calculations and pre-registration of sample size justifications for Nigerian health journals. Keywords: statistical power, sample size, research methodology, Nigerian universities, health research