"CodeGrade provides fully autograded assignments with good feedback to our students that satisfies our high requirements for user experience!"

Dr. Luis de la Fuente Valentín is Assistant Professor at the Universidad Internacional de La Rioja (UNIR). The International University of La Rioja (UNIR) is the largest Spanish-speaking private online university, with students from all over the globe and offices in Mexico (UNIR Mexico), Colombia (UNIR Colombia), Ecuador, Peru and the United States.

At the UNIR School of Engineering, Luis is in charge of continuous improvement of teaching and learning and the design of a new free MOOC course in Python Programming, for which he uses CodeGrade. This Python course was offered completely free of charge and available for about 1000 students who were enrolled in Computer Science courses at UNIR, like Big Data or Artificial Intelligence. It lasts 8 weeks and has biweekly assignments. As it is a completely free and extracurricular course, it is given teacherless; the lectures are pre-recorded and the assignments are fully automatically graded using CodeGrade.

CodeGrade instead of a teacher in a MOOC

Next to their regular curriculum of paid courses, students at UNIR could enroll in the additional Python Programming course for free. Luis explains that this was both for experimental reasons and because of financial constraints: “All of our regular courses have a teacher.” Luis mentions that having a teacher is important to him: “In an online environment, this supports the students. But it obviously increases costs and makes us dependent on the online availability of the teacher. For our free massive online course [an extracurricular add-on to regular courses], we could not assign a teacher and wanted to try out how that would work out for our students.”

It was a logical choice to adopt CodeGrade for this course to ‘replace the teacher’, he mentions: “We believe that teaching a programming course without graded assignments is useless, so we needed a way to grade the students’ work. Simple home-made automatic tests could not fill this gap, as we care for a good integration of the automatic tests, and we did not have a teacher, so we decided to adopt CodeGrade.” For a fully online university, a smooth user experience and good integrations are essential: “CodeGrade provides fully autograded assignments with good feedback to our students that satisfies our high requirements for user experience and fits the flow of the course. Because of CodeGrade, we can provide this course with graded assignments.”

Adopting CodeGrade for the Python MOOC

After selecting CodeGrade as their autograder, Luis mentions the two steps he had to take to start using it in the Python MOOC: “First, we needed a perfect technological integration. With about 1000 students enrolled, we needed both the back-end and the user interface to be working smoothly.” As UNIR is a fully virtual university, they have custom and specific tools CodeGrade had to integrate with. “We have a very homogenous and specific infrastructure, and we needed CodeGrade to integrate perfectly with the way we do things at UNIR. The CodeGrade team helped us very well and developed a custom integration with our learning management system Sakai especially for us.” 

After creating the autograded assignments in CodeGrade, which was done quickly and without any problems by a colleague of Luis, the second step was to present these assignments to the students. Since this was a teacherless course, it was especially important that the workflows and instructions for the students were clear and easy to follow. “With CodeGrade, we had to tell the students exactly what to hand in and how their Jupyter Notebooks [a way of writing Python code, read our blog here] had to look like. This was something everyone had to get used to, but it was very good and smooth for a first iteration, as more than 500 students passed the course without any problems whatsoever.”

Results and going forward

Luis enjoyed using CodeGrade and mentioned the frequent updates and support from the CodeGrade Team was great: “The experience and support with CodeGrade was great, transparent and quick!” And from what he can tell, his students also liked CodeGrade: “It looks like the students also enjoyed this approach, the feeling I have about it is good. There were no complaints, the students understood well what to do and how CodeGrade worked.”

He reflects on the first time they had a fully autograded MOOC: “It is important for students to have a personal connection, especially in an online university. But as this was an extra and free course, it is only possible to offer this if the costs are kept reasonable, so using CodeGrade instead of a teacher. The experience was good for us, we are already working on the next iteration of the course!”

“One of our goals is to explore new innovative ways of teaching, and autograded courses are one of those ways. We want to expand the autograded approach to more courses.” For these upcoming courses, he plans to get even more out of CodeGrade: “I want to fit CodeGrade even better in our procedures and protocols, I want to make CodeGrade ‘ours’. The feedback we could automatically give now was already good, but CodeGrade offers even more possibilities that we want to explore for future courses.”

In his academic career, Luis researched and explored automatic grading in computer science courses and taught using home-made autograders. He concludes with: “Also for courses with a teacher autograding is useful. Our teachers have two main functions: teaching and grading. With autograding, they can spend less time on grading and spend more time on teaching. Here CodeGrade can be a great help for teachers: CodeGrade guarantees that student code meets low level requirements, so the teacher can focus on high level grading and even enjoy their grading tasks again! Next to fully automatically graded courses, I am exploring the use of CodeGrade in regular computer science courses at UNIR too.”