This is a rough one... On the one hand, I want to ensure that our students are learning and we must create measurement tools. On the other hand, I don't have time to write 13 different assessments for my classes. When I was teaching high school band, I gave a written, scantron pre/post test for my ensemble because it was required by the administration that every class must produce pre/post test data for measurement. No exceptions. But here was my problem: I had among the most hetergeneous classes in the school. I had the senior valedictorian taking the same kind of assessment as a freshmen level 2 RTI student. So you can imagine the stats that were generated as a result of me giving the test in this form.

This is an extreme example of the opposite of the idea of differentiated assessment. Not only was the data completely worthless, it was a chore to collect. A better model of assessment is described on the following wiki page:

http://differentiatedstrategies.wikispaces.com/Differentiated+Strategies+for+Assessment

In order to differentiate assessment, we must go beyond the multiple guess variety of questions. In addition to being accomodating, assessment must also guide practice and help students set goals for themselves. Although open ended assessment questions tend to be messier than multiple guess, their flexibility makes them more accomodating for a wide variety of student abilities.