Attachment | Size |
---|---|
con_se.csv [6] | 10.51 KB |
Dear fellow researchers,
I am currently writing my PhD thesis in psychology. Part of my thesis is a three-level meta-analysis of the relationship between personality (i.e. conscientiousness) and self-efficacy in school students (in primary and secondary education).
My thesis also includes a meta-regression with covariates that contain missing values (MAR). That is why I would like to use the metaSEM package for an FIML estimation of missing covariates. As it seems possible to include multiple covariates simultaneously [through x2= c(X1, X2,...)], in the results I can only find one beta coefficients for all covariates combined. Is there also a way, to get the FIML estimation for each beta coefficient, separately?
My R code is as follows:
con_se.mod.metaSEM <- meta3X(y.c, v.c, cluster= sample_nr,
x2= c(year, interval.months, idv, class, se , pers_rater , pers_retro),
data=con_se, model.name="con_se 3 level metaSEM",
intervals.type="z")
The moderators included are continuous variables (year=year of publication, interval.months=time between assessment of conscientiousness and self-efficacy, idv=individualism index of the country where study took place, class=mean grade level of participants) and factors (se=domain of self-efficacy, pers_rater=self- vs. other-rated personality, pers_retro=whether assessment of personality took place before, after or at the same time as self-efficacy).
Thank you for your time!
Best regards,
Denise Braul