Interaction effect mixed effects meta-analysis (Metafor)

The following two lines of code generates a mixed model with two moderators and asks a summary of that model:

lor.rma.mixed.npinteraction <- rma(EFFECTSIZE1, ESTVAR1, method = "DL", mods = cbind(FERTP, FERTN), data = dfSel)
summary(lor.rma.mixed.npinteraction)

I would like to add an interaction effect between the moderators FERTP and FERTN to this. Intuitively I would add ',FERTP*FERTN' behind FERTN like this:

lor.rma.mixed.npinteraction <- rma(EFFECTSIZE1, ESTVAR1, method = "DL", mods = cbind(FERTP, FERTN, FERTP*FERTN), data = dfSel)

But this does not add an interaction effect to the summary of my model. I looked it up and they did something similar, basically just this:

lor.rma.mixed.npinteraction <- rma(EFFECTSIZE1, ESTVAR1, method = "DL", mods = ~ cbind(FERTP*FERTN), data = dfSel)

They added a tilde before cbind and only added FERTP*FERTN to the model and got the interaction effect and the individual effects of both moderators. This also didn't work for me. Does anyone have a suggestion on what to do?

FERTP and FERTN are both binary categorical variables either indicating 'Pyes, Pno' or 'Nyes, Nno' relatively. They indicate whether phosphor or nitrogen fertiliser was added yes or no. The effect size is a measure of how root fungal-plant relationship effects plant growth.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.