Introductory programming (CS1) remains a major barrier to persistence in computer science, with high DFW (D, F, Withdrawal) rates and wide disparities in prior programming experience. At Tennessee Technological University, historical CS1 DFW rates ranged from 25-31%. Existing solutions reduce inequities but often add credit requirements or create administrative burdens. To address these challenges, we developed a self-selected, experience-based grouping (EBG) model in CS1. After receiving instructor guidance in the first week, students choose between Purple (novice/slower pace) or Gold (experienced/faster pace) groups, with structured opportunities to switch groups during the semester. Unlike track-based models, the EBG approach operates within a single course section, offering flexibility without altering credit structures. Grounded in Self-Determination Theory, this design embeds autonomy, competence, and peer connectedness to promote motivation and belonging. Across three semesters, EBG reduced DFW rates from historical averages of 25–31% to 18–19%. Course GPAs also improved, alongside stronger survey-reported confidence, belonging, and motivation. In Spring 2024, these results led my institution to adopt the EBG intervention across all CS1 sections. My dissertation includes this work as well as investigating the longitudinal impacts of using the EBG model on student preparedness, motivation, confidence, and persistence in CS2/CS3, and persistence in the computer science major.