Hello everyone,
I am writing a systematic review about if X experimental group has superior effect than Y comparison group on outcome.
Please note that the lower the scores of the outcome, the better the results are.
to answer if there is superiority: I wanted to calcuate the effect size of each study using Hedgs g formula. I used the Comprehensive Meta-analysis (CMA) software to calculate these for me with visualization. However, I was confused when the g values were negative, so I did not know whether this shows that X group has superior effect than Y or how do I know?
other question is, some studies reported median of the outcome, instead of mean, I searched a lot how to calcuate the effect size in this case, but I did not find any. I look forward very much to your suggestions.
Best,
Ghazal Naser (senior master student)
Hedgs G is pretty similar to Cohen's d; in fact, it is kind of like a correction for Cohen's d (see page 4: https://pubs.asha.org/doi/pdf/10.1044/cicsd_33_S_42#:~:text=This%20means%20that%20if%20repeated,a%20high%20of%201.42%20SD.)
Therefore, I'm pretty sure that a negative value interpretation is the same as with Cohen's d: i.e., negative usually means the control group had a higher mean than the treatment group. I think a lot of studies tend to use the absolute value of Cohen's d and then you look at the averages to see which is bigger, which might be what is confusing you.
If a study only reports a median, I don't know if you can calculate an effect size that requires an average from it. In some cases, it is quite impossilbe if the variance should be pooled (for many dependent samples cases). You might try emailing the authors and seeing if they're willing to either share the data or provide the average for you?
- Badges
- Science method
- Similar topics
- Biological Science
- Histology
- Staining