STATISTICAL AND REPORTING ERRORS IN APPLIED BIOMECHANICS RESEARCH

  • Duane Knudson
Keywords: type I error, sampling, reliability, effect size

Abstract

Applied biomechanics research reports in the 2004 volumes of the ISBS Proceedings (n=94) and the Journal of Applied Biomechanics (n = 11) were analyzed for statistical errors. There was no significant difference in the distribution of ratings of quality based on statistical errors. The percentages of various statistical errors in reporting data were also quite similar. Both sources of applied biomechanics research had unacceptably large percentages of papers with errors in reporting statistical testing and results that limit readers' ability to interpret the findings. Improvements need to be made in the reporting and peer review of applied biomechanics papers in ISBS Proceedings in order to achieve the mission of ISBS.
Published
2008-03-17