Statistical Versus Practical Significance

In hypothesis testing, there are two types of outcomes. One is the determination of statistical significance, which is present when one rejects the null hypothesis. When one rejects the null hypothesis, however, one only knows that there is likely a nonzero relationship between the variables in the population. Practical significance is based on the potential usefulness of the research finding. Researchers use measures of effect size or strength of association to aid in determining practical significance. Effect sizes are measures of observed difference in units of standard deviation, such as Cohen's d, Glass's Δ, and Hedges's g. Strength-of-association measures estimate the proportion of variation in the dependent variable attributed to the independent variable. Examples are eta-squared, omega-squared, the intraclass correlation, and Cramer's V.