Dear All,
May I please ask for your help with the following problem: I am running meta-regression analyses using robumeta (robust variance estimation meta-analysis package).
In order to control for large effect sizes, I would like to use the leave-one-out method. This means: I always drop one study from the analyses and then compare the average effect sizes of each analysis against each other.
My code generally runs fine.
However, I would really appreciate some help from much more advanced STATA users with the following:
1. Save the coefficients from each meta-analysis as a variable or in a matrix. This will be helpful to plot all coefficients from each meta-analysis and see if there are any outliers that drive the effect,
and
2. it would be super helpful to know which study was taken out, hence, which drove the effects. Is it possible to display the author or study id somehow?
This is my code:
foreach num of numlist 1/60 {
robumeta cohensd if studyid!=`num', variance(cohensd_var) study(studyid) weighttype(random) rho(0.8)
dis e(QE)
}
All the best,
Sophia
PS: This is the overall code but probably not needed to answer my question!
drop if not_ma==1
preserve
drop if post_int_categorical==2 | post_int_categorical==3
drop if harshmalspectrum_stricter==0
tab post_int_categorical
foreach num of numlist 1/60 {
dis "These are the main effect analyses for `var'"
robumeta cohensd if studyid!=`num', variance(cohensd_var) study(studyid) weighttype(random) rho(0.8)
dis e(QE)
}
restore
Related Posts with Saving coefficients from meta regression as variable (ROBUMETA)
Recover population size by groupHi, I am trying to construct a "population" variable from household survey data. Data is collected …
Prefix ‘by’ when running post-estimation statistics for a PCAHi I have a question regarding the use of the prefix ‘by’ when running post-estimation statistics fo…
Question analysing data - counting different values of variables if another variable has a set valueHi there I have a Dataset of lots of observations (People) with lots of variables. For example I hav…
🐛 Stata inserts ESC as a character into inputI've been looking for a way to reproduce this since about the times of Stata 10, and finally today f…
stsplit at specific date which is different for every observationHi there, I'm trying to asses if is there an increased risk of virological failure in one year follo…
Subscribe to:
Post Comments (Atom)
0 Response to Saving coefficients from meta regression as variable (ROBUMETA)
Post a Comment