I am trying to conduct a meta-analysis using the difference in mean change between intervention and control groups. I have one study with data on mean (SD) for baseline, after and change in both intervention and control groups. Therefore, I want to calculate the mean change and its corresponding standard deviation to be used in meta-analysis; for this, I will need to use the paper which reported the mean and standard deviation values for baseline, after and change to calculate the correlation r and use it to calculate the mean (SD) for change for other papers. Is it valid to calculate correlation r for just one study and use it for all other studies included in meta-analysis? If no, how many papers are reasonable?
No it isn't valid, for two reasons. First a single study will be vulnerable to sampling error. Second the 'true' correlation will vary between studies because of factors such as range restriction. If I were doing it this way I would use all the data and try and account for range restriction. If you are using a spreadsheet to collate the study data it should increase the amount of work too much.
I wouldn't go as far as to say it wasn't valid Any evidence of the likely correlation is better than no evidence. Perhaps you can try a sensitivity analysis, using the correlation from the paper, but seeing what happens when you go (say) +/- 0.1 away from it.
Dear, Thom and Adrian
Thank you so much for your answer. I also think that it is more logical to use the correlation coeficient and check the sensitivity. Because, using the mean change (in the intervention and control groups) for estimating the mean difference can better show the possible effect of interventions and this method is superior compared to calculating the effect size based on after intervention period values.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis