We are trying to compare all paths of one path model between two groups. The groups were given different experimental treatment at the beginning but had to answer the same questions afterwards. Thus, we have the same variables for both groups but expect differences in the outcome due to the experimental treatment.
What our path model looks like:
model_pfad_gesamt <- "
identifikation ~ aeintauchen
realismus ~ beintauchen
hinterfragen ~ c*eintauchen
recherche ~ fidentifikation + jrealismus + nhinterfragen + reintauchen
aktivismus ~ gidentifikation + krealismus + ohinterfragen + seintauchen
indirect_h2c := af
total_h2c:= indirect_h2c + r
indirect_h2d := ag
total_h2d:= indirect_h2d + s
indirect_h3c := bj
total_h3c:= indirect_h3c + r
indirect_h3d := bk
total_h3d:= indirect_h3d + s
indirect_h4c := cn
total_h4c:= indirect_h4c + r
indirect_h4d := co
total_h4d:= indirect_h4d + s
aktivismus ~~ recherche
identifikation ~~ realismus + hinterfragen
realismus ~~ hinterfragen
model_pfad_gesamt_results <- sem(model_pfad_gesamt, data = d_Dokus, se = "boot")
summary(model_pfad_gesamt_results, standardized = T)
Those are the different formulas we tried for the model comparison which all unfortunately did not work:
fit <- sem(model_pfad_gesamt_results, data=d_Dokus, estimator="MLR", missing="FIML", group="Gruppen", group.equal="regressions")
chisq.test(model_pfad_gesamt_results ~ d_Dokus$ZG01)
modellvergleich_test <- manova(model_pfad_gesamt_results ~ Dokus$ZG01)
Is the approach correct? Where do we need to adjust the code?
We'd really appreciate if someone could help us with this!
Desperate regards from Germany <3