One team on our campus gave all multiple choice tests last year. Each test was graded solely by a scanner. Students received only overall percentages scores. Teachers spent their grading time doing item analysis of the tests and planning specific interventions for strategic groups based on this analysis. Students did not receive formative feedback on their actual work.
This team had not only the highest proficiency level on our campus, but also the highest grades, and an 11% growth in their proficiency rates.
The other two teams graded all assessments on a rubric and provided some formative feedback for students between assessments. Some strategic interventions were planned, but much was left to the students to get tutoring and re-take assessments a la standards based grading. One team had a 5% proficiency loss. The other also had an 11% growth in proficiency, but the level was significantly lower than the other successful team.
As a department leader, I’m not sure what to do with all of this. Each of these teams was at a different grade level. While I see the successes of one method, I’m not sure that the others can find the same success by simply copying what was done. My gut says that grade-level definitely plays a significant factor in this success. I say this not to discount this team’s awesome job, but to hopefully not take steps backward from our growth for the other grades. Teachers talked about how they felt the standards-based grading train really helped students feel successful and built confidence in a way that they hadn’t seen previously. However, we also talked about how perhaps we micro-managed skills too much. Our skills lists were very detailed.
We have not come to any conclusion with all of this. At this point, we’ve agreed to test the waters by perhaps keeping things at the status quo for quizzes but to move tests to the multiple choice format. We’ll see how that goes. For now, I’m looking for some help. I wonder what my standards-based grading friends would say about all of this. What are strengths or weaknesses to multiple choice tests? To free response tests? What do you use? Why?