Data Collection for the Learn-OCaml Programming Platform: Modelling How Students Develop Typed Functional Programs

Online programming platforms provide unique opportunities to collect and analyze a wealth of information on how students develop programs. In this work, we give an overview of the data collection infrastructure for the Learn-OCaml programming environment which allows students to write, typecheck and run OCaml code directly in their browser. We collect data for three different events: compile reads student's code and type-checks it; eval compiles and evaluates the code; grade runs the auto-grader on the student's well-typed program and provides feedback on the input-output correctness and code style. The data which we aim to gather across semesters serves as a basis for a wide variety of future studies on understanding how students develop programs in the context of typed functional programming.