“You are smarter than a calculator,” elementary school teachers often tell their students. “The human mind is a wonderful tool.” But as academic technology improves over the years, the first statement might not remain true. Recently, the New York Times announced that Harvard and the Massachusetts Institute of Technology are developing software that would allow a computer to evaluate students’ written short-answer and essay responses.
EdX, the company working on this technology, hopes to make grading more efficient for both professors and university students. Unlike other answer-detection systems such as the Scantron, which has been used since 1972, this software goes to the extent of analyzing arguments and forming interpretations. Although impressive-sounding at first, the importance of meaning over mechanics and the emotion embedded in reading is too much to sacrifice for grading efficiency. This software can also make it easier for students to cheat.
The first plausible error that EdX entails is the failure of a computer to comprehend the meaning of an essay that contains spelling and grammatical errors. Even in the most recent version of Microsoft Office, a large amount of words are missing from the “dictionary.” This is because Microsoft Office does not detect proper nouns, slang, or words recently added to real English dictionaries. For example, “lol” is considered incorrect by SpellCheck even though it was added to the Oxford English Dictionary two years ago. Although words like this are not common in academia, they can be used to address a current event or prove a point in writing. Most professors would agree that spelling and grammar errors are minute when compared to the overall meaning of a written work. If the technology would issue an automatic failure due to misunderstanding, it is unfair to the writer. Only real people can decipher those technical errors made in the haste of a 40-minute writing period and popular culture references that are so commonly used in writing.
Another downfall to EdX is the inability of the professor to contribute genuine reaction and response that have always been used in essay-based subjects. EdX so far only promises to create responses that state if an essay is “on topic or not.” Must we forget the wonderful feeling our grade-school teachers gave us with comments like “Well done!” or “Great understanding!”? Sure, a computer can shuffle through and spit out random responses, but surely no student wants to read fake enthusiasm. Students also need constructive criticism. In order to develop one’s writing skills, we must know where we can improve or expand our writing.
EdX has also failed to acknowledge that technology can easily be tampered with. Students have accessed Scantron answer keys in the past, used illegal notecards, showing willingness to cheat on tests. Who is to say that EdX software cannot be hacked by students more proficient at using technology than taking a test? Even if one can determine the cues that the software detects, they can already be on their way to a passing paper. There is no way to “hack” a professor’s mind, aside from the information one is willing to share.
Supporters of this technology are quick to say that professors deserve a break from their busy lives. However, there are other ways to resolve this issue. Teaching assistants already grade many essays throughout college courses. This is not to suggest instead that they should do more work, but perhaps department leaders can decrease the overly high student to TA ratio that currently exists. It goes without questioning if professors and assistants of essay-based courses are even asking for this type of help. It is typically up to the professor to decide how to distribute coursework, so they have the authority to decrease the amount of essays assigned in the first place.
When it comes to composition, technology is not a suitable tool to evaluate students’ success. An algorithm cannot be applied to an English paper. Although there may be a theoretical “formula” to writing a paper, each individual student is entitled to his or her own voice and perspective, which only a human reader can truly comprehend. Examples used by writers cannot be understood by a computer with no context or capability of reasoning. And students deserve to know exactly what their professor or teaching assistant thinks about their writing. This technology only enables students to write essays mechanically. Essay-grading software would only make it easier for students to write like robots, teachers to grade like robots, and university to be turned into one large factory.
