Before publishing this book, I taught the information presented here for 9 months with students and gathered feedback (6 of those 9 months were entirely online). When I first got the idea for this book, I was opposed to the idea of artificial intelligence evaluating writing. I believed that automated essay graders (AEG) compromised the social nature of writing. As I dug into the process and began to unlock the mystery of how the writing was actually evaluated, I began to see a way to use this technology to help students. COVID19 sent everyone into some form of remote or online learning format. Teachers, students, and parents became overwhelmed by learning and working at home. Parents suddenly became teachers and writing took on a new emphasis in a digital environment. It can be argued that digital writing environments present a more complex communication environment than print. Teaching and learning took on a different level of time commitment. As school systems grapple with how they will “open back up” and teachers and students grapple with the decision to return to the classrooms, artificial intelligence and its role in education may be taking on a new meaning. Consequently, my thoughts on the topic of automated essay graders (AEG) have come full circle.
I presented the information in this book at an Oregon GED Summit 2019 conference, and saw hope in the eyes of teachers who had previously been frustrated, because they didn’t understand the “reading” mechanism behind automated essay graders (AEG). I have seen hope in the eyes of students because AEG “reads” the same way every time, and the consistency is like a video game. Students feel like they can “level up” by learning what matters to the AEG, the rules of the game so to speak. For some students, knowing how the AEG will “read” takes the fear out of writing. Students can submit a paper to AEG, receive a score, re-write the paper, and see if they can improve the score. Human graders can be inconsistent and not always evaluate writing in the same manner. Human graders can vary their interpretation of what is “good” writing. The automated essay grader has been programmed to compare an essay written by a student to sample essays which are the foundation of the programming. It’s true that the sample essays are proprietary, and there is likely a bias built into the sample essays, but understanding the nature of how an automated essay grader will “read,” and what is valuable to the reader, is just another way of “knowing your audience”.
During the time I started working with the concepts of the book, COVID-19 hit the country and schools were forced to go online or remote as a delivery model. Teachers became overwhelmed with trying to deliver a classroom to students who were now learning at home. An automated essay grader suddenly had a different role. I found a free AEG source (paperrater.com) that students could use to evaluate their own writing and give themselves feedback in much the way Ellis B. Page had once envisioned. Students began to understand why word count and word choice mattered in a way they hadn’t previously understood. Students could understand the importance of how a sentence is started and how the use of transitional words would impact the grade their paper would get. The automated essay grader also added a level of consistency to my own evaluation process and gave me specific directions to help my students improve their writing in a definitive manner.
Overall, the creation of this book on automated essay graders has been an evolutionary journey for me. I now see automated essay graders in much the same way I see using my phone for help or directions. Automated essay graders are not useful for all types of writing. Creativity will never be measured through programming, but not all writing is meant to be creative. Writing this book has helped me think about strategies to help my students be more successful in their written communications. It has also helped me think about what words really “matter” and the importance of helping students understand the relationship between cause and effect. It seems especially important to develop those skills in students during this historical moment of COVID-19. As schools, as well as the workplace, become more automated, and remote or distance learning/working becomes the “new normal” understanding and leveraging artificial intelligence will become a critical skill.