This transformation is necessary to consider studies that provide non-parametric descriptive values for a meta-analytic study.
Dear Arturo,
When I am conducting SR-MA, sometimes I find papers only describe information in terms os median and IQR. You do not want to exclude this papers from your MA. In this case, and only considering that the data follows a normal distribution (or symmetrical), you can apply this formula:
Mean=median, and SD = IQR/1.35.
Again, please aware about the distribution of the data. You can read more about this issue here:
Hope this helps!
http://handbook.cochrane.org/chapter_7/7_7_3_5_mediansand_interquartile_ranges.htm
Thank you very much for the help. Unfortunately, when authors describe descriptive data by median and IQR it is usually because their scores are not normally distributed.
You can estimate it (with various assumptions), but you can't calculate it since information is lost. The norm with meta-analyses is usually just to write to the authors if they are still around.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Values