In Spring 2013 I taught a 5-week MAT142 class as the second half of a linked MAT121-MAT142. In order to alleviate some of the students’ math anxiety associated with exams and to demonstrate practical applications, I switched from exam-based assessments to project-based assessments with promising results.
This fall, I chose to teach both a 5-week and a 16-week MAT142 course, and ran them both as project-based. Results were mixed. I ran an analysis of variance on overall scores of 173 of my MAT142 students and found no significant difference in scores due to the method of assessment or the duration of the course. There may be two reasons for this. First, I had no data from a 5-week course that was exam-based to use for comparison. Second, all the classes used My Math Lab for homework and quizzes. Since My Math Lab is expensive, many students in the shorter course did not purchase permanent access. They failed to complete all the assignments, lowering their overall scores. If I were to continue this analysis, I would obtain final score data from 5 week MAT142 courses that don't use My Math Lab.
Marianne - You hit the very heart of why CATS exists! You are trying out something new, assessing it, and making future instructional decisions based on your data. Even though you did not get the results you wished for, you still learned from your efforts, and student learning will, I'm sure, follow closely behind. Job well done!
Yes, this is very familiar . . .
I think the project based assessment is an exceptional idea. I really hope you continue this research to determine if there is an effect. You did an exceptional job identifying confounding variables (e.g., 5-week students not purchasing the full math lab, etc.). I would suggest doing this assessment again with the two classes of the same length to reduce the number of confounding variables or controlling for these variables as you indicated. Great job!
I like the use of ANOVA. You are bringing a lot of rigor to your analysis!