Reflection of NAEP Student Work

Working with the NAEP student samples was a very engaging experience and really opened up my eyes to the importance of using good, well-developed rubrics. My group members and I were tasked with reviewing a data analysis problem samples from a set of 4th grade responses. Additionally, we were provided with the rubric that was created to score this problem. Upon first glimpse, the rubric seemed to be detailed enough to cover all of the bases within grading. However, once we actually started going through each of the sample responses and comparing them to the provided rubric, we had a change of thought about that. It became clear that many of the student responses could, technically, fall in two or more of the provided grading levels. As a result, it was really up to the individual or group grading the response to determine which level the response fell under. This led my group members and I to understand why this problem was taken out of the testing pool in the first place. We managed to get through all of the student sample responses and assign each to a specific score. It was shocking to us how many we scored as off task or minimal, the two lowest scoring possibilities. In addition, we were not completely sure that we scored each and every one of the samples correctly, considering the rubric was not exactly clear. Like I stated earlier, this project really opened my eyes to the use of rubrics. While they are a very useful tool for grading, they also must be written correctly and with enough detail. In addition, the different levels of a rubric must be clearly distinguished without any overlap occurring. This overlap is what leads to inaccurate grading of student work, similar to the grading issues and discrepancies we came across in the NAEP student samples. Rubrics are a fantastic tool, but only if written and utilized correctly! When selecting the three samples to create activities for, my group and I decided to pick three samples to cover the range of scores. Therefore, we chose an extended, partial, and incorrect/off task sample. We felt this was the best way to represent everyone in a classroom. We created one, overarching activity for students to do in which they physically created a human bar graph which was then transferred onto the SmartBoard. Each of these focus students worked on various areas, depending on their individual needs. I thought that it was actually quite difficult to differentiate for these three sample learners. One was clearly advanced and required more challenging work, while the other two were lacking and required some additional assistance in problem solving. However, I felt that this was an important obstacle to tackle, because differentiation of learning is something I will have to do every day in my classroom.

After viewing my colleagues' work on their individual NAEP problems, it became evident that my group was not the only one to struggle with scoring some of the student samples. Additionally, the two other rubrics we looked at also had some major flaws and were difficult to comprehend. One of the rubrics was extremely cut and dry, which was different to ours. I felt that our rubric was more well-developed and the different scoring levels were described a little better than this rubric. The other rubric we examined was very wordy and hard to understand/decipher between some components of the different areas. The other groups explained that they, too, experienced difficulties with determining what scoring level some of the student samples should have been graded as. I think that this was a common trend among all of the questions and sample responses. I believe that a big take-away from this project and experience is that creating and utilizing clear and cohesive rubrics is absolutely the key to accurately and fairly scoring student responses on a mathematical problem like the ones provided.