I was on a discussion forum connected to the “Getting and Cleaning Data” MOOC on Coursera. During the peer evaluation of the course’s project, one question puzzled me and I wanted to get some clues. The question is in no way related to the content of the project or the course, so I feel free to report it here. It was: “As far as you can determine, does it appear that the work submitted for this project is the work of the student who submitted it?”.
I was puzzled. I simply do not understand why somebody would submit work from another person Why, before being evaluated would you believe that somebody else is better than you? And what would you learn. I was not alone and this discussion uncovered very simple facts. By reading the forum posts. it became evident that there is a considerable number of people that actually does it! A simple google search on the script file name required by the project revealed hundreds of results. And somebody remarked it in the forum, and you can report such supposed cases of plagiarism to Coursera. Now, about that, there are several points to discuss.
- A MOOC student is not a policeman. The fact that a repository is old may mean that the student is re-sitting the course. The fact that you found two or more evidently identical repositories does not automatically reveal who is the pirate. So you could be ending up accusing the wrong person. Do not get me wrong I am totally for the punishment of the wrongdoer, but I would not condone myself if I would accuse the wrong person.
- This phenomenon decreases the value of the certifications provided by the platform. It is probably difficult for someone to achieve a full specialization based on plagiarism but it may be possible. So, what would be the credits in front of a possible employer?
- It is now clear to my eyes why some platforms like udacity have abolished certificates.
- The platform should do something to reduce the phenomena, because this is possibly hampering its reputation and its capability to sell certifications. The current verification policy is insufficient. So, coursera should invent a way to – at least – make plagiarism more difficult. I do not have solutions, but for instance, assigning unique repository names and file names by student with some automation and encoding of user data would possibly reduce the problem.
