What exercise and sport scientists don't understand

Area of Science:

• Exercise Science
• Sports Medicine
• Sport Psychology

Background:

• Research design and statistical power are critical for valid scientific findings.
• Previous analyses have indicated potential issues with statistical power in various scientific fields.

Purpose of the Study:

• To evaluate the statistical power, effect size reporting, and Type I error control in studies published in the Australian Journal of Science and Medicine in Sport (AJSMS) between 1996 and 1997.
• To identify potential limitations in research design and interpretation within exercise and sport science.

Main Methods:

• Analysis of 29 studies published in AJSMS in 1996-1997.
• Computation of statistical power for small, medium, and large effects using Cohen's conventions.
• Examination of effect size reporting and interpretation, and control of experiment-wise (EW) Type I error rates.

Main Results:

• Median power to detect small, medium, and large effects was .14, .65, and .97, respectively.
• Effect size reporting was infrequent, and their interpretation was rarer.
• The mean EW Type I error rate across studies was .49.

Conclusions:

• Research in exercise and sport science, as represented in AJSMS during this period, was likely underpowered, especially for small effects.
• Inadequate attention to effect sizes and experiment-wise error rates may lead to substantial Type I and Type II errors.
• Researchers are urged to conduct power analyses, report and interpret effect sizes, and control for EW error to improve research quality.