The results from a factorial experiment with two factors involved (factor A with 4 levels and factor B with 2 levels), which was completed with 3 replicates [a total of (4x2x3) runs] are being analyzed). If the model must be kept hierarchical, how many degrees of freedom (DF) will be in the error term if "factors A and B" were found to be NOT significant but the "interaction AB" was Significant