From regan.bernhard at gmail.com Thu Jan 9 16:43:46 2020 From: regan.bernhard at gmail.com (Regan Bernhard) Date: Thu, 9 Jan 2020 10:43:46 -0500 Subject: [Mediation-information] problem with multimed Message-ID: Hello! I am trying to use multimed but keep getting the following error: Error: variable 'cong_incong' was fitted with type "factor" but type "numeric" was supplied Here is the command I am using: lab_force.multi_med<-multimed(outcome="force", med.main="norm", med.alt="value", treat="cong_incong", data=labdata). force, norm, and value are all continuous numeric variables. cong_incong has been made into a factor with two levels using the factor function and every other function in my script seems to be treating it as a factor, however, when I check its type using typeof(cong_incong) it tells me it is an integer. For what it's worth, I am able to run standard mediations for norm and value separately with no issues. I saw that this question came up here once before. The suggestion was to make sure the treatment was a numeric variable with values 0 and 1. I tried that and got this error: In mean.default(data.b[, treat]) : argument is not numeric or logical : returning NA Again I used typeof to and the data is of type double. I'm stumped and would love some suggestions! Thanks! Regan -------------- next part -------------- An HTML attachment was scrubbed... URL: