Understanding how students engage with programming is critical for supporting those who struggle and fostering long-term improvement. This research investigates how programming process data can be used to model student engagement and development over time. Drawing on Self-Determination Theory, the work explores how behavioral patterns, may reflect underlying motivational states. While prior research in learning analytics has focused primarily on error-based metrics to predict performance, this study aims to connect engagement to constructs of motivation. Using methods from learning analytics, such as clustering and predictive modeling, this research seeks to identify pathways of motivation in introductory programming. This work contributes to bridging the gap between learning analytics and theory to understand student motivation.