Teacher Shannon Snowball Discusses the Impact of Automated Essay Grading on her Students' Writing
Thursday, November 13, 2008 at 12:00PM Shannon Snowball is a third grade teacher from Peralta Trail Elementary School in Apache Junction Unified School District, Arizona, with fifteen years of teaching experience. She teaches using NWREL's 6+1 Traits ® and has been using LearningExpress's eFolio(TM) program for the past year. I asked Ms. Snowball to tell us about her experiences with her students and the impact of automated essay scoring on their learning.
I introduce the 6+1 Traits of Writing® over the first half of the school year. As I teach each trait, I make that trait a part of the rubric I use to assess the students' writing. I usually start with Voice and Point of View. Conventions or grammar rules are taught throughout the entire year. When I target a trait, I have them practice it first. Then, when I think they understand the requirements I am looking for (based on a 6-point rubric), I assign an essay to start the writing process. I score the essay using the rubric, only scoring the traits they have practiced so far. Eventually, I have them score their own papers using the rubric and compare their assessments to mine. This practice helps me to see if they understand the expectations. To help the students understand each trait, I use multiple strategies such as creative lessons, examples from literature, modeling, the writing process, student samples, technology, and hands-on learning.
My principal asked me to use an automated scoring system last year. At first I thought, "How accurately can a computer score a student's writing?" After using it, however, I was impressed. I used a standard rubric to score the student essays manually. Then, I took the same essays and entered them into the automated scoring system. The results were very similar. When the students scored their own writing, I found that their self-evaluations were also very close to the program's results. The automated system made it easy to track the students' progress and compare their growth. It also gives an unbiased score, which helps with consistency.
Automated scoring is another tool that provides feedback. It helps reinforce accountability. The scores break down each trait, which makes it easier to see what traits need more attention.
I definitely see a difference now that I am working with an automated essay scoring system. I am able to show them the improvement in their scores over time, which helps them to feel validated. Seeing their improvements also motivates them to work harder, and they don't mind having to do additional drafts. I keep all of their rough drafts, essays, and scores so they can clearly see their progress. Scoring their papers and then having the program give a similar score reinforces their opinions about their work. When their scores go from something like 2.1 to 4.4, they get excited. The lower writers may not score fours and fives, but they feel good when they see their scores go from ones to threes.
I also think the feedback motivates them to want to write more. I make overhead copies of some of the essays that have different scores (I remove the students' names, of course), and we review what they scored and why, as well as what they did well and what they could have done better. This helps them understand the way the papers are scored.
I had one student last year who started with a score of three in all traits. She was determined to "beat" the program by scoring perfect sixes. The computer does not score sixes very often; the work has to be very high quality. This student even stayed after school to work on her essay. She scored a four, two fives and two sixes. She was so proud of herself. Also, a group of ten students wrote essays for a contest, and they challenged themselves so much that one of them won first place and another was fourth in the Pinal County Character Counts essay contest.
I think using the automated scoring makes me more aware of the students' consistency, growth, strengths, and weaknesses, which helps me tailor my lessons to meet their needs. This, in turn, produces more success in their writing.
The only suggestion I can make for changing the program is that last year, the program broke down the traits for each assignment, but did not show a comparison of those traits over time. It showed the trend for the student's overall score, but not for each individual trait. I understand that this is a feature in the new version of eFolioTM, however.
Last year I saw dramatic improvement, even though it was the first year I used this program and I did not enter the first essays until the end of October. The students showed significant growth from the October essay to the February benchmark. For example, my block-one students went from an average score of 2.70 to 3.82 in Ideas and Content from October to February. In Voice, they went from an average of 2.95 to 4.05. Their average raw score went from 2.89 to 3.88.
The scores continued to increase every time I used the program. By AIMS testing time (Arizona's state test), most of my students' averages increased another point. 93% of the third grade 'Met' or 'Exceeded' the state standards on the AIMS writing test.
This year, I entered the students' writing at the end of August, so I will have even earlier scores to compare and evaluate. According to the program, my students need to work on the Ideas and Content trait. I have been focusing on this area since the month of October.
Shannon, thank you for allowing this interview regarding your experience with 6+1 Traits of Writing. It has been a very informative discussion, and I look forward to seeing your continued results over the course of this school year.



