Manul: An organized autograding approach to enhance learning and assessment of Data Science students

The following work proposes a general workflow for scoring Data Science exercises and tests in a curated development environment using the Manul library, evaluating the student's progress through the subject of Data Science in an educational context. The Manul library was implemented with the following general workflow as a reference for use: first, the instructor designs the exercise with a number of problems of their choosing; then, they implement the evaluation instructions and the tests necessary to score the exercise; following that, the exercise must be deployed in a container so students can solve it according to instructor's guidance; once finished, the Manul library will automatically score the student's exercise. The results show a positive reaction among the students who participated, providing us with necessary feedback to improve our workflows.