Hello, and welcome to the first ever Weekly Wilkins! This is a weekly blog post dedicated opinion articles that I find. Today's article is Grading the Graders: Human vs. Robot a New York Times opinion article about how "robots" are not as good at grading essays. With that article, though, I will also bring in another article: Facing a Robo-Grader? Just Keep Obfuscating Mellifluously which is about a new machine designed for grading papers more efficiently.
E-Rater (the grading machine) grades by looking for grammatical errors, big words and sentence fragments, and gives points accordingly. One thing this machine does not look for, however, is if the information is correct. Which means that, as it is put in the article Facing a Robo-Grader? Just Keep Obfuscating Mellifluously "E-Rater doesn't care if you say the war of 1812 started in 1945". And the machine gives better grades to longer essays, even if a short essay has better material in it. Therefore, it is very easy to trick it into giving a better grade.
In the Grading the Graders article, the author believes that the machine is not a good tool because the writers had great writing that machines could not pick up. I agree with him. I think that the E-Rater should only grade papers on grammatical errors, if even that. For now, at least. The robots might get better in a few years, but right now, grading should be left to the professionals: teachers.
No comments:
Post a Comment