Effect Size Guidelines, Sample Size Calculations, and Statistical Power in Gerontology

Area of Science:

• Gerontology
• Psychology
• Statistics

Background:

• Standardized guidelines for interpreting effect sizes (e.g., Cohen's r, Cohen's d) lack empirical basis and may not accurately reflect specific research fields.
• Gerontology research, encompassing both individual and group differences, requires field-specific effect size benchmarks for accurate interpretation.

Purpose of the Study:

• To investigate the distribution of effect sizes in gerontology research.
• To establish field-specific benchmarks for small, medium, and large effect sizes in gerontology.
• To provide recommendations for interpreting effect sizes and sample size calculations in gerontology.

Main Methods:

• Extraction of effect sizes (Pearson's r, Cohen's d, Hedges' g) from meta-analyses in top gerontology journals.
• Calculation of 25th, 50th, and 75th percentile ranks to define small, medium, and large effects.
• Conducting a priori power analyses based on observed effect size estimates.

Main Results:

• Established gerontology-specific effect sizes: Pearson's r = .12 (small), .20 (medium), .32 (large).
• Established gerontology-specific effect sizes: Hedges' g = 0.16 (small), 0.38 (medium), 0.76 (large).
• These values suggest Cohen's general guidelines may overestimate effect sizes within gerontology.

Conclusions:

• Recommended gerontology-specific effect size benchmarks: Pearson's r = .10, .20, .30; Cohen's d/Hedges' g = 0.15, 0.40, 0.75.
• Researchers in gerontology should adopt these field-specific benchmarks for more accurate interpretation of results.
• Larger sample sizes are recommended in gerontology research to achieve adequate statistical power with observed effect sizes.