Robo-graders

Robo-graders, or software that grades essays have been there for some time. But how good/useful is an algorithm to assess the quality of writing?

Les Perelman, a critic, says these algorithms don’t understand the meaning (obviously) and just use certain metrics like “length and the presence of pretentious language”:

“The fallacy underlying this approach is confusing association with causation. A person makes the observation that many smart college professors wear tweed jackets and then believes that if she wears a tweed jacket, she will be a smart college professor.”

Further, Perelman says that once kids know (or guess) how the algorithm works, they will write to suit the algorithm without worrying about content! A couple of students at Harvard demonstrated just that by writing an application that generates gibberish and yet gets rated highly by the algorithm.

Then again, as Chaitanya Ramineni, an ETS (Educational Testing Service) researcher, says, “human scoring suffers from flaws” too. Remember the dread with which most parents view the marking of board exams?

Another problem is rooted in the fact that any grading algorithm would have to be created by analyzing the qualities of good essays (as graded by humans). But this then means that such an algorithm would inevitably “punish creativity”. Let this article explain how:

“The computer judges “topical analysis” by favoring “similarity of the essay's vocabulary to other previously scored essays in the top score category.”

Ergo, anything different would, by definition, be considered worse!

But others do see benefits to algorithmic grading. Andrew Klobucar says that:

“His students are willing to revise their essays, even multiple times, when their work is being reviewed by a computer and not by a human teacher.”

Klobucar feels that happens because a human reviewer feels “corrective, even punitive” whereas an algorithm feels individualized yet not “personal”. Turns out there is even a term for this effect of technology on human behavior, the “disinhibition effect” (Outside education, it’s also seen in how honestly patients answer questions to software v/s doctors. Guess the eponymous Dr.House was right in asserting that patients always lie!). Amusingly, other researchers feel part of the reason why students are willing to revise and resubmit to a computer is that “the process of improving their writing appeared to take on a game-like quality, boosting their motivation to get better”!

Maybe the robo-grader has a place in the education system after all, just not to give grades.