RT drop got higher percentage correct scores on inclusion trials than on exclusion trials and could explain the training sequence, but non-RT drop got low scores on both trials, showing a lack of explicit learning and could not explain the training sequence.