ELLEN
BERG
Diary #27
Helping Students Zoom In on Scoring Guides
What was the mother's life like? How do you know that?
A: I think she was maybe poor. B: His mother's life has been bad because it says life for her ain't been no crystal stair. It had tacks in it, splinters, and boards torn up. C: She thinks now that she has a very good son to hear, and she wants him to be the best.We have been working on a poetry unit for the past three weeks. Last week we read Langston Hughes' poem "Mother to Son" in class. I was curious to see how well my students could comprehend the figurative language in poetry, so I assigned five constructed response questions about the poem. Above, I've listed the second question and three students' answers. They range from excellent to, as Simon Cowell is known to say, ghastly. After reading through my students' answers, I learned that while they were pretty successful at understanding the deeper themes of the poem, they did not offer any explanations or examples to support their generalizations. At a workshop I attended a few months back, the presenter, Willy Wood, noted that if a student can make a correct generalization the hard part is already mastered. The student's brain used the clues from the text to make the generalization, though the student may not be aware of it. Wood reasoned that students do not include the proof with their answers because they are not in the habit of doing so. They have either been inundated with selected response questions like multiple choice or true/false, or their teachers have accepted the generalization alone. It is not that our students cannot provide the proof; it is simply that they do not know they have to. The lack of proof causes a problem on many levels. First, as students move through high school and college, their instructors are looking for them to be able to back their assertions. Second, our high stakes test, the MAP (Missouri Assessment Program) consists of all rubric-scored, constructed response questions. In order to get a perfect score on a question, students must supply the generalization and two support items. If students supply a correct generalization, their score drops to half.
Helping students back up their assertions
After analyzing the results of my students' papers, I decided to spend some time looking at constructed response questions within the context of poetry. I typed up a list of anonymous responses to the "Mother to Son" questions without scores. I presented the scoring guide to the students, and we talked about how a 2, 1, and 0 looked. As a class, we discussed the question and answers I included at the beginning of this entry. I had students write down what score they thought each response should receive and why they would score the item that way. We voted as a class, students explained their rationale for scoring, and then I revealed the actual score.
After working one question together, I released students to score the rest of the questions and answers. As they worked, students asked questions to clarify their thinking, debated with one another about the scores, and were deep in thought regarding each answer. Even the underachieving class I wrote about last week was fully engaged and on task. We went through the results, and I was excited to see most students really understood how to score the answers. We had an interesting discussion at one point when a student wanted to score a response lower because of grammar. I asked her where she saw grammar on the scoring guide, and something finally clicked for her. "When a teacher gives you a scoring guide for your work," I explained, "she is held to whatever is in the scoring guide. I can't decide later that I want to grade on something I didn't tell you about ahead of time just because it's a problem." I saw the usefulness of scoring guides rise in my students' estimation. That exercise was a success, but I wondered if their new knowledge on how constructed response questions are scored would carry over to their own responses. I decided to test it out with a new poem and questions.Applying what we learned
We read "The Highwayman" by Alfred Noyes, talked about some of the vocabulary and the setting, reviewed the scoring guide posted on our SmartBoard, then students got to work on the assigned questions.
As students worked, I observed several interesting events. First, I watched them return to the text again and again instead of relying on their memory. An unwillingness to return to the text has been a huge problem in testing and other situations, although I have stated repeatedly that good readers return to the text. Here, students were doing it without being told. Second, students were asking both me and their peers if their answers were 3's on the scoring guide. I tried to return students to the scoring guide to analyze their answers with probing questions, and students seemed to do the same. At the very least they would tell the questioner what was missing and why they needed to include it. The third result was that students were not satisfied with scoring less than perfect. It seemed that knowing how to get a perfect score motivated them to reach that score. The process of modeling and practicing scoring with a scoring guide was so simple I am kind of embarrassed to be writing about it here. However, I'd wager there are many of us who use scoring guides and rubrics with our kids but fail to spend much time teaching kids how to use it. Time is a major factor, but having the sample work to score is a roadblock as well. I find that when I try to write a sample project that is less than a perfect score, I cannot do it very well and results in an inauthentic sample. Still, we must find ways to guide our students through the process of actually using and making rubrics and scoring guides meaningful. I think that once my students were able to get inside the evaluator's head and inside the problem, they were able to keep the evaluator's mindset as they worked. Just another example of moving the zoom lens a little closer in...
