Measures of Relationship Strength in Occupational Therapy Research

Abstract

Measures of relationship strength (effect size) were computed for 330 statistical tests appearing in the occupational therapy research literature The effect size measures were categorized as “small,” “medium,” or “large” according to criteria proposed by Cohen (1977). Analysis of the various measures of relationship strength revealed that the average effect size fell in the medium range Further analysis indicated that effect sizes categorized as large were frequently reported as statistically significant, while effect sizes labeled as small were rarely associated with statistical significance. The implications of the findings are discussed in relation to research conducted in applied fields such as occupational therapy. The argument is made that due to the nature of the variables under investigation and the lack of experimental control in most clinical situations, measures of relationship strength should be reported routinely in occupational therapy research where statistical hypotheses are evaluated. The reporting of measures of relationship strength will aid researchers in determining the power or sensitivity of experimental manipulations and will provide valuable information related to the clinical relevance of research findings.