I estimated genotypic correlations and then the path coefficients which gave very strange direct effects (more than 3) and residual (-2.2). I used SAS Proc Reg with the VIF option to remove any unique variable. However, removing 2 variables (having VIF greater than 10) didn't work. The results are still the same. Hope someone has the clue.

More Mohammad Tahir's questions See All
Similar questions and discussions