The search for a grade

The first semester at my school ends tomorrow. We do not have semester exams. Some AP teachers at my school give an exam in the evening. I chose not to. I think the main reason my colleagues opt to give an exam is to give the students the experience of sitting through an extended test in order to prepare them for the exam in May.

I am confident that all of  my students have sat a 3-hour exam once before. All but one were students at my school last spring, and they sat final exams at that time.  Therefore, I don't feel the need to give them a long test.

I do want them to be frequently reviewing and revisiting previous topics. So, I am giving them three Free Response Questions that cover the first four chapters. But, those three questions will not combine to form some large percentage of their semester average.

After today, my students will have been graded on 13 free response questions. Two of them were of the Investigative Task variety (that is, Free Response Question #6 from past exams). The others were either taken directly from past exams or were in that style. I am dropping their lowest two scores.

My assessments in AP Statistics this semester have mostly been a departure from how I have evolved otherwise. I have another (dormant) blog regarding my thoughts on assessments and evaluation. My tenth grade integrated mathematics course has zero period-long tests. I have given homework quizzes. But, in that class, grades are largely calculated from projects and presentations. I believe these forms of assessment give me the best indication of what a student can do and knows.

My AP Statistics have completed two projects this semester and they account for a large portion of their grade. I think the projects have given the students a worthwhile opportunity to engage meaningfully with the course topics and skills in a way that will prepare them for the exam. But, I think they also help the students see the utility and relevance of the topics beyond the AP exam. And, I think they provide me with a much better insight into what the students really understand and their abilities to apply the methods they have been studying.

I dislike multiple choice items a lot. (I wrote about this dislike in this blog here and here.) I do not think they are fair nor do I think they tell me much about what a student understands. Perhaps, in a 40-item multiple choice exam, they give a better picture. The larger number of items reduces the chance that a score is based largely on lucky guesses and increases its probability of being due to a student's understanding.

The free response questions are better, but still contrived as they must fit into an arbitrary allotted amount of time and they are written in a way that takes much of the decision-making away from the student.

I do wonder if I could teach a project-based AP Statistics class and still prepare my students for the big, bad AP exam. Seems possible. I guess I have my goal for next year's course.

Anybody have experience teaching AP Statistics as a project-based course? I would love to read about your experiences, recommendations, ideas, and thoughts.