0

I have means(x) and Std deviations of 2(intervention and control) groups. I wanted to find Odds ratio with CI.

I calculated the cohends d and CI first using,https://campbellcollaboration.org/escalc/html/EffectSizeCalculator-SMD1.php

then converted cohends d to Odds Ratio using http://www.stat-help.com/spreadsheets/Converting%20effect%20sizes%202012-06-19.xls

But here i do not get Confidence interval for OR, i get only OR

How do i calculate Ci for OR

jyoshma
  • 1

2 Answers2

0

As a general principle if you want to convert from one scale to another and have not only a central estimate but also the limits of the confidence interval the way to proceed is to convert all three of them from the scale given to the scale you want to end up with.

So in your specific case just convert the limits of the interval of $d$ as you did the central value.

mdewey
  • 17,806
0

See here, in the section of Statistical Inference.

The distribution of the log of the odds ratio is Normal, with a standard error $$ SE = \sqrt{\frac1{n_{11}}+\frac1{n_{01}}+\frac1{n_{10}}+\frac1{n_{00}}} $$

So the confidence interval at 95% is given by $OR\pm1.96SE$.

chuse
  • 770