I applied SEM and according to modification indices I fixed an error variance of an indicator to 0. But it is not the reference indicator. Then I got standardized direct effect and squared multiple correlation as 1. Is this normal? How can I interpret this? Or am I not going to interpret this coefficient?