Monday, 8 April 2013

Who is grading MOOCs' students?


EdX announced last week the introduction of an automated grading software with a focus on providing speedy feedback to students’ essays: “There is a huge value in learning with instant feedback, students are telling us they learn much better with instant feedback.” says Anant Agarwal edX’s president.


The idea of an artificial intelligence grading essays raised quite a few eyebrows but if not AI, who’s grading the MOOCs’ students?


  • Multiple Choice Questions

MCQs and other types of close ended questions form a large part of quizzes, either in the form of ‘in-lecture pop-ups’ to maintain engagement or in the form of assessments contributing towards the final grade.
Those are obviously automatically graded although instant feedback is sometimes not provided in details, encouraging students to figure out their errors by themselves.

  • Programming assignments

Although there can numerous options for tackling a specific programming assignment, the instructions can be coded, allowing grading programs running on the platform’s servers to not only accept or reject a student’s submission but to provide some level of comments.

  • Essays, etc.
For fields of study in which obtaining a valuable certificate of completion without submitting an essay is not an option, Coursera has pioneered an interesting peer-assessment system.

After submitting its own work, each student is provided with a very detailed evaluation matrix and is required to grade at least 3 to 4 submissions from  his fellow students. On most courses, the student also has an opportunity to self-evaluate before its average grade and the detailed comments of its peers are made available.
Peer-grading is not limited to essays. For programming courses, it is a good way for students see how differently their peers might have approached a problem.
In a Canvas MOOC on project management run by the French engineering school Centrale Lille, peer-assessment was also used to grade Gantt charts and mind maps but was ultimately personally validated by one of the course organisers.


It will be interesting to see edX’s AI grader in action. Its expected speed will probably be of some value. However, peer-grading is a fantastic way of learning by casting a critical eye on one’s own work and sampling the work others.


Write a comment below or share this post!
Post a Comment