Never mind, I figured it out! But if anyone out there is reading this and had the same question, the answer is: I had done an "assess variable importance" analysis, and the predictors were automatically re-ordered to match their relative influence on the model, i.e., from most to least important.
Anderson B. Mayfield