We all (at least vaguely) remember those statistics and research methods classes in which we learned about ANOVA, p values, and all of that other complicated statistical stuff. At our SLP Hill Country Institute this week, Dr. Ron Gillam, from Utah State University, gave me a reminder of the one statistic to which we SHOULD pay attention: Effect Size. Now, all of the American Speech Language Hearing Association journals require that intervention studies report effect size. So, what exactly is it, and what do you need to know?
Effect sizes “estimate the magnitude of effect or association between two or more variables” (p. 1, Ferguson, 2009). Basically, it tells how big of a change was seen in the participants in the study. While the other statistics tell how likely that result would have happened by chance, effect size tells whether or not the change was meaningful.
While there are a number of different indexes of effect size, one of the commonly used is Cohen’s d. Cohen’s d is the difference between two means divided by the standard deviation. Here’s a simple example: What if you planned to do an intervention, and you were pre- and post-testing with the CELF. On the pre-test, the student got a standard score of 80, and after intervention, on the post-test, the student got a standard score of 90. The difference between the pretest and posttest score is 10 points. To determine the effect size, you would divide the difference by the standard deviation, which is 15. So, 10/15 would be .67. What does that mean?
Generally, a large effect size is .8 or higher, which is almost 1 standard deviation of change. That kind of effect can be observed just by looking at the child. A medium effect is .5 to .79. That effect could be seen through the administration of some kind of test, but is not readily observable. A small effect is .2 to .49. A small effect is a barely noticeable change, even with a test.
So, the next time you are reading an article on an intervention in your ASHA journals, I hope you take a moment to look for the effect size to help you determine if the evidence indicates that it is an intervention you might want to use in therapy.
Ferguson, C J., (2009). An Effect Size Primer: A Guide for Clinicians and Researchers. Professional Psychology: Research and Practice, 40 (5), 532-538.