In this post-covid world, I have noticed that a lot of people are learning new things on MOOCs - Coursera, EdX, Udacity, Unacademy etc. I tried my hand at an online specialisation as well, but after completing it, I’m left wondering if certificates from MOOCs actually matter. In fact, how authentic are the people brandishing their MOOC certificates?
My skepticism stems from my own experience, which only served to dilute the importance I once gave to the different online learning platforms.While I successfully completed a Management specialisation (5 exhaustive courses and a Capstone Project based on a case study) on Coursera, I came across an interesting problem that is inherent in the most famous online learning platform - and maybe other MOOCs as well.
Coursera relies heavily on multiple-choice quizzes and peer-graded assignments to evaluate the learners attending a course. I noticed the following things:
- All multiple-choice quizzes allow the learner different attempts to pass. Coursera allows 3 attempts every 8 hours. That means anyone with plenty of time can try out different combinations of guesses until they get a perfect score. Which also means that someone may not go through the entire content, they can simply hazard best guesses to clear the quizzes quickly. It only takes a maximum of 4 attempts to score a 100% on a multiple choice quiz with 4 options.
- Peer-grading crowdsources the painstaking work of scoring to the learners. Courses require the learner to grade at least 2 peers before their grade is calculated. The scoring metric (or rubric) is displayed to everyone who submits their assignment. Coursera also shows a Honor Code to all students who are about to attempt an assignment. It’s read as much as the text above “I agree” button while installing new software. Here are the problems with the peer review mechanism:
a. Scorers can give grades arbitrarily. Since the scorers are not evaluated on their grading capability, they can fail others or give full marks to their friends.
Quiscustodietipsoscustodes?
b. Learners can edit their submission until their assignment is not graded. I found a lot of my peers had copy-pasted the rubric into their submission. They would submit a blank submission early to gain access to the rubric, and then paste the scoring sheet directly into their answer before evaluations. I scored such assignments a solid ‘0’. But would everyone enrolled in a course do that? Since the assignment exactly matches a rubric, wouldn’t someone give it a full-score?
c. Checking plagiarism is not easy. I spent an entire weekend collecting data for a case study to write a 500 word essay for a Digital Transformation case study. On starting reviews the next day, I found my assignment was copied exactly with just my name replaced in the document. Even the filename was left unchanged. I was lucky to find the copycat immediately. But what if there were others? I reviewed only 2 of my peers, what if there was another copycat who got credits from my efforts by simply pasting their name on my work? I would never know.
- Reporting misuse - Coursera does have a reporting mechanism where inappropriate or plagiarised submissions can be flagged. As someone in IT, I know that the flag report would end up in some sort of ticketing system being tracked by a handful of people. Coursera currently serves over 2700 courses - even 1 report per course per week is a huge number. I know the system is understaffed/heavily backlogged because most people that I had reported were able to complete the course without any action.
Which brings us to the main problem at hand. Although MOOCs enable learning, there are not enough checks and balances to confirm whether someone who completed a certificate did learn something. This problem is exacerbated by considering there are learners distributed across geographies and timezones. In a post-covid online world, how can you bring exam proctoring and evaluations to massively open online education platforms?
Please mention your thoughts in the comments below.
