More Important Things to Do With Student Writing Than Just Grade It

A recent study by University of Akron comparing the scoring of student essay writing by computer software programs versus that done by "trained human readers" has caused quite a bit of controversy among college level English teachers.

The authors of the study rightfully point out that these findings should not be taken as an excuse to rush to replace English teachers, but rather as an indication that these software programs could be a useful resource for student writers as a way to get additional feedback on their writing. 

Coincidentally, the English faculty at the community college where I teach, just met to conduct our annual review of student performance across our writing program. One thing we do at these sessions is a blind cross-score by teachers of essays written by our students on the essay portion of the Proficiency Profile test from ETS. The essays are first scored using the eRater software program (this is one of the same programs evaluated in the Akron study); then we compare the scores of the teachers to those given by the e-Rater program. Our results, however, are the exact opposite of the larger study. For five years running, we have found scores given by teachers to be significantly different (almost always lower) than those given by the software.

Teachers of English, however, are not just "trained human readers" or warm-blooded scoring machines, and we have much more important things to do with our students' writing than assign it a grade or score. Writing is first of all an exchange of ideas. Our students are sharing their thoughts with us, and before anything else, teachers should be respectful, thoughtful readers of those ideas. Like many of my colleagues, I respond to my students' writing with questions and comments--on what they are saying (e.g., "What makes you say this? Could you tell me more about this point? Do you have more evidence or examples to support this particular point? What an effective and delightful analogy!") 

Our next responsibility is, as fellow writers, to help these budding younger writers develop in the craft of writing. We help them learn how to shape, clarify, and express their ideas for a variety of audiences and purposes. Part of that, of course, is teaching the conventions of formal usage. One way to accomplish both those goals is to expose them to a wide variety of great writers.

When I sit down next to each of my students or with a small group of them (physically or virtually) to share their most recent work; to reflect on how they have grown as writers since their last piece or since the start of the course; to give sincere, critical feedback on what would help make it better---I am doing what no software program can copy. It is through these very human interactions that I also show them that they have worth far beyond a number on a scale. I also learn some very real and fascinating things from my students' writings. They also have experiences, points of view, information, and writing techniques to share with me and the world. It is a wonderful place to be, and one that no software program can take from me.....yet.

 

 

Comments

[David B. Cohen]

Thanks for writing this, Renee. I haven't run any complex linguistic algorithms on your post, but I do see that you have used a certain word and sentence-level complexity, and there are words that suggest transitions and development. This post will be rated an 8.

But more seriously, my take on this whole thing is that if a computer can match the ratings produced by humans, we haven't asked enough of the humans.

[Mary E. Ward, NBCT- 2022]

Renee, I agree with you wholeheartedly. Yes, it is time consuming, but grading students' work with our eyes, head, heart, and hands produce more than grades along a numerical gradient. I like to call it meta-cognition dialogue. Such interchange addresses the question in students minds of of "How can I, in what ways am I, which technique am I able to employ to make this richer, flow easier, probe deeper?" Computer grading can't address such inquiries that are in the minds of students and come forth through students/teachers interacting--focusing upon what went well rather than what didn't.

[TeachMoore]

Thanks for your passionate responses. Teachers & parents need to resist efforts to dehumanize education.

[Paul M.Rutherford, PhD]

Prior to teaching, I was a naval officer in the aviation community. Attempting to replace a teacher in the spectrum of assessment is analagous to replacing a naval aviator with predator-type(pilotless) aircraft or UAVs. If you REALLY want to accomplish the mission, you put a PILOT in the cockpit. Same holds true for assessment software; do you want truly, reliably, validly desire to assess student performance?, then put a TEACHER 'in the cockpit'!!

[Nancy Flanagan]

Wonderful piece--and exactly right.

[Suzanne Klein]

Thanks for your thoughtful contribution to this conversation. I have a feeling that computerized assessments may still be coming (at least to K-12 education) because writing has been restored to its position as one of the 3 Rs by the new Common Core Standards. Assessing students' writing skills is thus going to be more widespread, and it costs more money to pay teachers to do the scoring. While computerized scoring may be practical, we need to think about what will be lost in this equation. Perhaps teachers' voices will make a difference.

Suzanne Klein
Founder, WriteSteps
www.writestepswriting.com

[Marie@ Concise Writing Consultancy]

Excellent post Ms. Moore. And I agree as well. There’s so much more to writing than just the technical aspects, so many deeper levels of learning within one piece of writing that can only be delved into through human interaction. I don’t think computer programs could ever achieve such a feat, and I don’t think they should. It would devalue human interactions, and that’s not right at all.

[David Olsen]

Hi http://davidolsenmarketing.com/ here. Great blog you have here. I will keep on visiting your site and will wait for your new post.