The Kalamazoo Promise (KPromise) is amongst the most well-known and generous tuition-free policies. This study advances the understanding of Promise student performance and persistence. We used a weighted-least square means and variance adjusted (WLSMV) SEM approach and k-nearest neighbors (k-NN) to deal with missing data. The main model suggested first-year college GPA ([beta] = -0.48) possessed the strongest effect on a first-year stop out followed by socioeconomic advantage ([beta] = -0.26), high school performance ([beta] = -0.25), immediate enrollment ([beta] = -0.22), and KPromise funding ([beta] =0.06). Model differences from 06-10 and 11-15 cohorts, illustrate that in the later cohorts socioeconomic advantage's effect on a first-year stop out lessened.