Analysis Of The Electronic Essay Rater (AES) System

1533 Words 7 Pages
Among the advanced new technologies offered by the progress in artificial intelligence research, many essay scoring software programs have been developed with the ability to score students’ essays and tender feedback immediately. The automated essay grading (AES) systems, which have both benefits and drawbacks regarding their implementation, have raised questions about the appropriate usage of these software options.
The Electronic Essay Rater (e-rater), developed by the Educational Testing Service
(ETS), uses syntactic variety, discourse structure and content analysis, and is based on natural language processing technology (Monaghan & Bridgeman). It applies statistical analysis to linguistic features, and also considers vocabulary and topical
…show more content…
This allows students to receive guidance along the way rather than submitting multiple drafts to their teacher (Markoff). Without this software, students often won’t receive any criteria they can base their progress on until they receive their grade.
Most importantly, online courses that often feature class sizes over 500 students shows the key benefit that automatic essay grading software provides. The ease of reaching many students at once with an online lecture faces a bottleneck if one teacher is expected to grade the submissions from such a large class. This can be alleviated by adding teacher’s assistants or paid graders, but this introduces significant cost to the organization and has the potential for inconsistency between the graders. In contrast, an online course with automated essay grading can allow students to submit their essay and receive instant feedback with the same grading criteria provided to everyone enrolled (Miller).
There are clear drawbacks for putting too much faith in automated grading as a replacement for a human grader. The current technology excels at identifying variety of word choice, the word count of sentences and paragraphs, or glaring errors, but many blind
…show more content…
Although the decision will ultimately depend on the organization’s budget, the management can prefer to use low-cost standard packages to leave the basic grading to the software, while the instructor grades the higher-level thinking in the essay. The tailored, costlier software packages may offer advanced features such as configurable grading preferences and machine learning.
This more advanced software could be helpful where graduate-level or sophisticated material is offered in class. Lastly, the management should consider the organizational environment in terms of computer literacy. Computer usage and access might be an issue for organizations and student bodies in low-income areas. If the majority of the instructors aren’t proficient with computers or the students don’t have easy access to computers, then the management might need to supplement the investment in AES software with complimentary investment, such as employee training or providing the necessary equipment.
As AI continues to progress at an accelerating rate, many researchers have been

Related Documents