I am currently running a causal mediation analysis for a clinical trial where my mediator model uses a negative binomial regression model and my outcome model uses a logistic regression model.
However, I get the following error: Error in offset(years.in.study) : object 'years.in.study' not found
Does the mediate package allow for an "offset" function to be in the model because the time each participant contributes varies? If so, how can I correct this error? If not, is there another way to handle these models?
Hello,
I am currently running a causal mediation analysis for a clinical trial where my mediator model uses a negative binomial regression model and my outcome model uses a logistic regression model.
model.x <- glm.nb(total.malaria.incident.events ~ Txarm + offset(years.in.study), data = delivery.analysis)model.y = glm(compositeBO ~ Txarm + total.malaria.incident.events + offset(years.in.study), family = "binomial", data = delivery.analysis)m.out = mediate(malaria.glm.nb, model.3.malaria.compositeBO, treat = "Txarm", mediator = "total.malaria.incident.events", boot=T)However, I get the following error:
Error in offset(years.in.study) : object 'years.in.study' not foundDoes the mediate package allow for an "offset" function to be in the model because the time each participant contributes varies? If so, how can I correct this error? If not, is there another way to handle these models?Thanks in advance!