Should a Computer Grade Your Essays?

1499 Words Nov 22nd, 2014 6 Pages
Case Study 11: Should a Computer Grade Your Essays?
The case study discusses the April 2013 launch of Harvard/MIT’s joint venture MOOC (massively open online course) essay scoring program, utilizing AI (artificial intelligence) technology to grade educational essays and short answers, with immediate feedback and ability to revise, resubmit, and improve grades.
In 2012, a group of colleagues, Les Perleman, Mark Shermis, and Ben Hamner, introduced over 16,000, K7-12 standardized school tests to the AES (automated essay scoring software) to compare results of hand-scored essays; the results appeared to produce a more accurate grade, however, no official statistical studies were performed to accurately maintain the claim.
Essay Raters
…show more content…
Currently, EdX tools are free and tout trainability by educators. As demonstrated in the video IBM Watson: The Science Behind an Answer (IBM, 2011), the system uses algorithms gathered from the data stored in the system and “learns” answers by utilizing similarity of questions technique. The instructors grade 100 essays and/or questions and as they are input, which inevitably become the tools that becoming the grading metrics.
Massively open online courses (MOOC) platform rests on “self-service” and utilizing EdX to allow students the opportunity to obtain certificates of completion, versus degrees. The university will pay EdX a fee based on generated revenue for initial courses and “production help” revenue base for course development, with a potential for future licensing. The passing rate is anticipated to increase an estimate 60 to 91 percent, causing scrutiny and angst amongst traditional block and mortar educators; this translates to a speculated decline in educational quality and as well as the demise of trained educators.
The first case question asks to define the intelligence level of the automated essay grading. The EdX system requires educators or graders to first, grade 100 essays then allow the system, using learning techniques; second, it trains itself to grade future
Open Document