New effect size and sample size guidelines in dentistry

Cohen has emphasized that the recommended thresholds for effect sizes should only be used in the absence of detailed information about effect size distributions within specific fields.

The study aimed to establish updated effect size thresholds (Cohen's d, Hedges' g and Pearson's r) tailored for research in dentistry.

Following methodologies from prior research on effect sizes, the data was extracted from meta-analyses published in the top 10 ranked dentistry journals. The 25th, 50th and 75th percentiles were calculated for Pearson's r values, as well as for Cohen's d or Hedges' g. A total of 4,250 studies were analyzed, with statistical analyses conducted using the R programming language.

The 25th, 50th and 75th percentiles for Pearson's r in individual differences research were 0.16, 0.40 and 0.67, respectively. For Hedges' g, the percentiles corresponding to small, medium and large effect sizes were 0.10, 0.35 and 0.86, respectively.

In light of these findings, researchers in the field of dentistry are encouraged to adopt the following thresholds: for Pearson's r, 0.20 for small effects, 0.40 for medium effects and 0.70 for large effects; and for Cohen's d or Hedges' g, 0.10 for small effects, 0.40 for medium effects and 0.90 for large effects. These updated thresholds can improve the rigor and quality of dental research, ultimately benefiting patients through enhanced diagnostics and treatment strategies.