Publication Trends and Statistical Validity of Articles Published in The Asian Journal of Kinesiology

OBJECTIVES This study investigated the publication status of article types, sample sizes, statistical programs, and statistical techniques in The Asian Journal of Kinesiology (AJK). In particular, the study analyzed statistical errors, focusing on the application oft-tests and analysis of variance (ANOVA). METHODS A total of 444 articles were reviewed. Publication type, statistical program, and statistical techniques were classified using a statistical checklist. Among the 347 articles employing inferential statistics, errors in the application oft-tests and ANOVA were examined in detail. RESULTS Of the 444 articles, 347 (78%) utilized inferential statistics. SPSS was the most frequently used program (89.0%), and approximately 96% of analyses adopted parametric techniques. Thet-test (94 articles) and ANOVA (207 articles) were the most frequently applied statistical methods, together appearing in 301 studies (77.4%). The error rate for t-tests was 18.1%, primarily due to inconsistencies between research design and the selected method. For ANOVA, the error rate was 39.1%, with most errors related to post-hoc procedures and some discrepancies between design and application noted. CONCLUSIONS In kinesiology research published in AJK, careful alignment between research design and statistical techniques is essential. The t-test should be applied only when appropriate to the design, with consideration of two-way ANOVA when feasible. For ANOVA, researchers must ensure an understanding of its assumptions, apply methods aligned with experimental design, and select appropriate post-hoc or contrast procedures to ensure statistical validity