[R-meta] Results from the meta-regression

Tarun Khanna kh@nn@ @end|ng |rom hert|e-@choo|@org
Tue Feb 23 22:04:02 CET 2021

Dear All,

I ran a meta regression using the rma and rma.mv function. The exact command is below:

rma(yi=zi, vi = vi, method="REML", data = dfinc, test = "knha", mods = ~
                         `Commitment strategies`+ Gamification+
                          Randomisation + StatsMethod + StudyDesign +
                          Weather +
                          OptedIn + interventionTreatmentPeriod +
                          Region + StudyYear )

I re-ran the code again after a while and it seems that the results are slightly different. There are slight differences in the estimated coefficients and variances. As far as I can tell I didn't change the underlying data. Is it possible that results change because of the function that I am using here? Or must it be some change in the code or data?


Tarun Khanna
Research Associate

Hertie School

Friedrichstraße 180
10117 Berlin ∙ Germany
khanna using hertie-school.org ∙ www.hertie-school.org<http://www.hertie-school.org/>

	[[alternative HTML version deleted]]

More information about the R-sig-meta-analysis mailing list