Convert median and IQR to mean and SD
Meta-analyses need means and SDs, but papers often report medians with quartiles or ranges. Estimate them with Luo 2018 (mean) and Wan 2014 (SD), or get the SD from a standard error or a 95% CI, in your browser, with the methods sentence that cites the method.
Your table
| study | arm | n | median | q1 | q3 |
|---|---|---|---|---|---|
| Study 1 | intervention | 40 | 12 | 9 | 16 |
| Study 1 | control | 38 | 14 | 10 | 18 |
Paste one row per arm with what the paper reports (median with quartiles or range, or mean with SE or CI); each row is converted by the method that fits what it has.
How it is chosen
Median with Q1 and Q3 → Luo 2018 for the mean and Wan 2014 for the SD; with min and max as well, the methods that use all five numbers; SE → SD = SE × √n; 95% CI → SD with the t distribution when n < 60, as in the Cochrane Handbook (section 6.5.2).
You get
- Mean ± SD per arm with the formula and the numbers used
- A methods sentence citing Luo, Wan and the Cochrane Handbook
- Send the converted table straight to the forest plot
R packages: runs in your browser
Questions
Is the converter free?
Yes. It runs in your browser and does not use credits.
Which formula is used for the mean?
Luo 2018: a weighted average of the median and the midpoint of the quartiles, with the weight depending on n.