In this study, we explored the impact of the decision agency (Student vs. Tutor) and granularity (Problem vs. Step) across students with different levels of incoming competence (High vs. Low). In a 2x2 classroom experiment, 279 students were randomly assigned to four conditions. They were split into High and Low groups based on their pre-test scores in the posthoc analysis. Our overall results showed that there were significant different impacts of the decision agency and granularity between High and Low students on learning performance. More specifically, for High students step level decisions can be more effective than problem level decisions; for Low students, they learned significantly better when they were making problem-level decisions than making step-level decisions, but no significant difference was not found when the decisions were made by the tutor. To our surprise, both High and Low groups showed strong decision-making preference for problem solving over worked example.