Although I have a limited knowledge of statistics, I need to understand statistical procedures used in an article on aphasia intervention completely and apply them to my own data. In a table in the article, a column with the mean, SD, t, df and p-value of a exemplar treatment of anomia has been provided along with the treatment's effect size (90% CI). There is no further explanation for the formula used to calculate it. I surfed the internet for it, but using Cohen's d does not yield the similar results. I am very confused now
!
Are you looking for an absolute effect size (e.g. a mean difference) or a relative effect size (e.g. Cohen's d, which is the mean difference divided by the SD)? What did the authors mean with "effect size"?
Effect size is a simple way of quantifying the difference between two groups that has many advantages over the use of tests of statistical significance alone!!
Effect size emphasizes the size of the difference rather than confounding this with sample size. However, primary reports rarely mention effect sizes and few textbooks, research methods courses or computer packages address the concept.
Try this link:
https://www.leeds.ac.uk/educol/documents/00002182.htm
Many journals are requiring some type of effect size measure to be reported along with t-tests, ANOVA, etc.
Here's another reference for you:
Schuele, C.M., and Justice, L.M., The importance of effect sizes in the interpretation of research. Primer on Research: Part 3. The ASHA Leader, 11, 14-27, August, 2006.
Thank you all for your great support and guidance. Unfortunately, I am still confused! I think I should mention the article I have trouble with. It is Howard and Nickels's article titled "Optimising the design of intervention studies: critiques and ways forward". If there is anyone who is familiar with this article, I would be so grateful if he/she can explain the effect size formula mentioned in the table 6.A
The effect sizes in that table are about the statistics calculated from the pre- and post-tests (according to WEST-COL, WEST-Trend and WEST-ROC in columns D,E,F). The first two (D,E) are the sum of the scores divided by 2. The third (F) is the sum itself.
I understood it differently than Jochen did.
WEST-COL = Pre1(-1) + Pre2(-1) + Post(2)
WEST-Trend = Pre1(-1) + Pre2(0) + Post(1)
WEST-ROC = Pre1(1) + Pre2(-2) + Post(1)
HTH.
Bruce Weaver, yes, that is how these statistics are defined. But the table gives an "effect size" below, which is for the first two the sum/2 and for the last one the sum of these statistics.
Thank you Jochen and Bruce for taking the trouble to help me. I really appreciate it.
If the test statistic was a correlated t and you wish to have a standardized effect size estimate (Cohen's d) you will need to take into account the correlation between the pretest and posttest scores. James Algina has provided SAS code for doing this. See http://core.ecu.edu/psyc/WuenschK/docs30/CI_d_RelatedSamples.docx
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning