Greetings,
I'm using a m-STAR Durbin (spmstardxt) model with panel data and getting strange results regarding the estimates of direct and indirect effects. More specifically, for every estimated coefficient, the direct and indirect effects are always opposite signs. Has anyone else encountered this?
Thanks (in advance)
I'm using a m-STAR Durbin (spmstardxt) model with panel data and getting strange results regarding the estimates of direct and indirect effects. More specifically, for every estimated coefficient, the direct and indirect effects are always opposite signs. Has anyone else encountered this?
Thanks (in advance)
