Automated Essay Scoring (AES for short) can be described both as evidence that technology in the world today is rapidly advancing and also that it is still not fully advanced yet. AES can grade an astounding 16,000 papers in only 20 seconds (Winerip 1); however the average human grader can approximately grade 30 essays in one hour (Winerip 1). Now right away one might think that since it is so much more efficient then why we don’t just use it for assessment tests such as the SAT’s. Unfortunately however, there is more to it than just that. A student can easily prepare to get a good grade rather than to actually be an exceptional writer. Another discomforting reason on why we shouldn’t use AES is that it cannot decipher whether or not a fact is true. For example if one were to write, “Dinosaurs roamed the Earth in 2003, scavenging old buildings, thrashing the monkey-made skyscrapers, and consuming all the tacos,” it would think that it is an A+ sentence. Also we have the fact that it may not be biased towards a student, as many skeptical students presume about their teachers, although it can be biased about an essay. An essay that has longer words, sentences, and paragraphs that are not written as well as an essay with smaller words, sentences, and paragraphs will still get a better grade. In addition to that it is a computer, a soulless piece of metal that does not experience emotion. Nor does it grade papers that exemplify emotion as well as informational text with good structure. Furthermore I personally believe that we can repair AES and actually make it better through an evolution of programming to the next generation of artificial intelligence. That may not be for a long time, but it may be nearer than one might assume.



Works Cited

Winerip, Michael. “Facing a Robo-Grader? Just Keep Obfuscating Mellifluosly” The New York Times, 22 April 2012