I want to conduct a meta-regression, using the variable. The studies in my meta-analysis include studies, that compore two indipendent groups. My question: to calculate average age of a study - is it enough to simply to add up the age of the two indipendent groups and divided it by two? Or are there any other, better ways to get the average age?
5
-
$\begingroup$ That will give you the average age. But what do you want to do with the average age when you've calculated it? $\endgroup$– Jeremy MilesCommented Sep 23, 2019 at 16:15
-
$\begingroup$ I would like to use it as a moderator variable, in a meta-regression. $\endgroup$– MarthaCommented Sep 23, 2019 at 16:22
-
$\begingroup$ and the two samples are usually not the same size, I guess thats a problem $\endgroup$– MarthaCommented Sep 23, 2019 at 16:41
-
$\begingroup$ I found the right solution in the meantime $\endgroup$– MarthaCommented Sep 23, 2019 at 16:58
-
$\begingroup$ Xa= population size of group A Ya= average of group A Xb= population size of group B Yb= average of group B (Xa/(Xa+Xb))*Ya + (Xb/(Xa+Xb))*Yb = Yc (the average of the combine population) $\endgroup$– MarthaCommented Sep 23, 2019 at 16:59
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
3
Your procedure works if the group sizes are equal. If they are not then you need
$$ \frac{n_1 \bar{x_1} + n_2 \bar{x_2}}{n_1 + n_2} $$ where $n$ is a sample size and $\bar{x}$ is a mean.
But note that if you use this as a moderator in your meta-regression it is only an ecological analysis, you are looking at the effect of being enrolled in a study of people of a certain average age not the effect of age itself.
-
$\begingroup$ Thank you very much! If I want to look at age itself, how should I proceed? $\endgroup$– MarthaCommented Sep 23, 2019 at 17:09
-
1$\begingroup$ You need the individual data. Search for the term individual participant data meta-analysis to find out more. $\endgroup$– mdeweyCommented Sep 23, 2019 at 17:37
-