Automated Essay Graders: Impact on Education and Writing Skills

This essay provides a comprehensive overview of automated essay graders (AES) in educational settings. It discusses the use of specialized computer programs to assess written essays, emphasizing their role in classifying facts and terms within the text. The essay highlights the cost-effectiveness and technological aspects of AES, driven by pressures on education systems to maintain standards and reduce costs through advanced information technology. It examines the advantages of AES, such as saving time for professors and providing instant feedback to students, while also acknowledging limitations, particularly in grading based on ideas and thought processes. The essay explores the use of AES as a tool during the writing process, emphasizing the importance of human evaluation for final grades. It also references studies on the predictive capabilities of computer models in scoring writing and discusses the benefits of AES for administrative purposes and providing valuable feedback. The essay concludes by emphasizing the importance of using AES in conjunction with human instructors to ensure a thorough evaluation of student work, addressing both the strengths and weaknesses of automated assessment methods.

ESSAY ON AUTOMATED ESSAY GRADERS
1

Automatic Essay Grader can be considered as the use of specialized computer programs
to consign grades to written essay in an educational setting. It is also regarded as a method of
educational assessment which analyses the efficacy of language proficiency used in the written
essay. The main objective of essay grader is to classify the facts and terms which is being used in
the written essay (Zhang, Mo, Jing Chen, and Chunyi Ruan, 2015). It is also used as rating tool
that gives marking to online writing tasks. Rating are given according to 1-6 in which different
aspects are accessed mentioned in the essay. Hence, it is considered as a problem of statistical
classification. The grade system is cost effective and is also technologically concerned; hence
most of the teachers and professionals uses the approach. The cost of education has increased
and that has led pressure to hold the education system. The system is accountable for results
through imposing several standards. This is the result of advanced information technology that
measures educational achievement at reduced cost (Wilson, Joshua, and Amanda Czik, 2016).
Writing is recognized as a critical skill in business, education and other layers of social
engagement which not only describes subject matter; but also increases the ways of writing
components. At the time of thinking professor grading papers, teachers can imagine a dark room
illuminated by a single florescent lamp that’s been pushed back to the farthest corner of a desk
by an expanding clutter of papers. This is somewhat a mixed up of papers that is used to be
reviewed for assessment purpose (Shermis, Mark D., Sue Lottridge, and Elijah Mayfield, 2015).
Although this may just be an image developed by movies, wherein it is believed that grading a
massive stack of essays would be rather stressful and overwhelming. Thus, because of this aspect
people are constantly striving to find easier ways of doing things. Hence, in many cases
automated essay graders have adopted because it is a helpful tool for teachers to find out the
efficiency level of essay writing. Further, it is also unbelievable that professors, at a college
level, should use these grade system when giving a final grade on a student’s paper, but rather,
the tool can be useful during the writing process for the students and as an added tool when
grading papers (Powers, Donald E., David S. Escoffery, and Matthew P. Duchnowski, 2015).
Automated essay graders are unable to grade based on ideas, and restrains the professor
from being able to understand what their students thought processes are like and what they need
help with. Although AES should not be depended on when giving a final grade, AES should be
used as a tool during the grading process. Automated essay scoring can save a professor time
when grading a large amount of papers, and can allow a professor to focus on grading based on
2

the students “voice” rather than their grammar and sentence structure. Finding reliable and
efficient ways to assess writing highly increases the standard of test writing components and it
also generates new tools for computer formats (Ng, Sing Yii, et al, 2015). The software uses
algorithms for the purpose of measuring more than 500 text- level variables which also yield
scores and feedback regarding the writing quality. Using such tool can assist the teacher to
diagnose the information on each writer and give them more time to address problems and assist
students to comprehend things that can be managed without machine such as content, reasoning
and writing.
The method of assessment should always be judged on the basis of validity, fairness and
reliability. Therefore, any instrument can be termed as valid if it actually measures the trait that
is significant to measure. It has been observed that teachers are turning to essay grading software
to critique student writing; however at the same time it also aids them to critics to serious flaws
in the technology (McNamara, Danielle S., et al, 2015). Although automated essay grading is a
useful tool for both the professor and the student during the process of writing. However, at the
time of perfecting an essay, the professor should give the final grade on a paper after they
personally evaluate the student’s work. The AES system has shown to be in the early stages of its
grading abilities focusing primarily on grammar, sentence structure and style, and word count.
Because of its lack of algorithms composed to create this tool, automatic essay graders lack the
ability to grade based on the students thought process and logic behind their ideas that they
portray in their work.
It has also been analysed that AES program instantly produces a high scoring grade.
While writing an essay, it is essential for the writer to hold the importance of meaningful
sentence so that the writing can be evaluated accordingly (Hoang, Giang Thi Linh, and Antony
John Kunnan, 2016). This sentence needs to use proper grammar and structure and it should also
hold appropriate meaning. Hence, with the help of grading system, teacher can look into all these
matters so as to evaluate the capability of students. Therefore, the instructor of the class should
still be required to evaluate the students work when using an automated essay grader. It is also
explored that those who use standard feedback methods without automated scoring spend more
time in discussing about spelling, punctuation, grammar and capitalization. Several researchers
have established that computer models are highly predictive and that gives idea about capability
3

of humans in scoring good marks in writing (Cummins, Ronan, Meng Zhang, and Ted Briscoe,
2016).
In addition to this, it has also been analysed that the benefits of automation are great
when considering from administrative view point. If computer models provide acceptable
evaluation and feedback, they help in reducing the amount of needed training for human scorers.
Another reason for which the AES programs are used is that in this tool, instructor can
understand their student’s thought process. If the professor of an English course does not take the
time to read any of their students work and only know the grade that the AES program spits out,
how will they know what their student’s need help understanding. Although some students will
talk aloud in class and present their ideas, allowing the professor to assume where they stand in
the class and their progression throughout the course (Buzick, Heather, et al, 2016). The AES
program is not able to inform the professor what their student’s ideas are, the professor should
find that out through evaluating the student’s work. The automated assessment helps the teachers
in providing valuable feedback to the students and through this, they can adopt new measures for
improvement. Further, students can also reinforce and demonstrate the principles and rules of
writing. This is highly essential for developing the writing skills of writer and at the same time,
efforts of teachers can be reduced due to automated assessment methods. Valuable feedback can
also help the students to become more competent (Barrett and Catherine, 2015).
A benefit of using this tool is the instant results that the students can receive from the
teachers. Rather than waiting to receive feedback roughly a week later when the class has moved
on and the student’s ideas are not fresh in their head. This allows the students to write more
frequent practice essays while still being able to track their progress. It can be difficult for an
instructor to assign essays as frequent as on to two a week because of the strenuous grading.
With the AES tool, students can practice writing properly structure essays without having to
focus on only writing a few essays (Zhang, Mo, Jing Chen, and Chunyi Ruan, 2015). The system
is also accurate in predicting human scores through applying fairly simple scoring method; hence
this could be deceived by students who get higher grades in longer essays. Automated essay
scoring is highly useful for teachers because through this, they can thoroughly analyze the
content of the essay and at the same time, they can also organize their own schedule. This is also
regarded as the best method of applying thorough understanding in essay writing.
4

Henceforth, the method is greatly recommended to English teachers because through this,
systematic use of quotations and punctuations can be identified. Moreover, it can also amend the
practicality in analysing the large scale assessments of writing ability. However, on the other
hand, employing human rater could be an expensive method in terms of time and resources.
Thus, it is vital to include more than one rater in large scale writing assessment. This can also
reduce biasness in individual scorers (Wilson, Joshua, and Amanda Czik, 2016). Automated
ratings would surpass the accuracy of the usual two judges. The same tool has many weaknesses
as well such as it highly stresses the lack of human interaction as well as the sense of the writer
while writing the essay. Another criticism is based on objections; hence that is the major reason
computer count variables are considered less important.
5

REFERENCES
Barrett, Catherine M. "Automated essay evaluation and the computational paradigm: Machine
scoring enters the classroom." (2015).
Buzick, Heather, et al. "Comparing Human and Automated Essay Scoring for Prospective
Graduate Students with Learning Disabilities and/or ADHD."Applied Measurement in
Education 29.3 (2016): 161-172.
Cummins, Ronan, Meng Zhang, and Ted Briscoe. "Constrained Multi-Task Learning for
Automated Essay Scoring." Association for Computational Linguistics, 2016.
Hoang, Giang Thi Linh, and Antony John Kunnan. "Automated Essay Evaluation for English
Language Learners: A Case Study of MY Access."Language Assessment
Quarterly (2016): 1-18.
McNamara, Danielle S., et al. "A hierarchical classification approach to automated essay
scoring." Assessing Writing 23 (2015): 35-59.
Ng, Sing Yii, et al. "Automated Essay Scoring Feedback (AESF): An Innovative Writing
Solution to the Malaysian University English Test (MUET)." (2015).
Powers, Donald E., David S. Escoffery, and Matthew P. Duchnowski. "Validating Automated
Essay Scoring: A (Modest) Refinement of the “Gold Standard”." Applied Measurement
in Education 28.2 (2015): 130-142.
Shermis, Mark D., Sue Lottridge, and Elijah Mayfield. "The Impact of Anonymization for
Automated Essay Scoring." Journal of Educational Measurement 52.4 (2015): 419-436.
Wilson, Joshua, and Amanda Czik. "Automated essay evaluation software in English Language
Arts classrooms: Effects on teacher feedback, student motivation, and writing
quality." Computers & Education 100 (2016): 94-109.
Zhang, Mo, Jing Chen, and Chunyi Ruan. "Evaluating the Detection of Aberrant Responses in
Automated Essay Scoring." Quantitative Psychology Research. Springer International
Publishing, 2015. 191-208.
6