In response to the 2/23/2011 post, “Common Assessment Data Analysis – Part 1,” Doc Seabolt writes:
As to the first paragraph, I have to admit I am not familiar with “Test Blueprint,” however, I do agree with the writer, in theory. That is, it is certainly possible to construct an assessment instrument that tests at a variety of levels. In fact, I would argue that this is exactly what most standardized tests ATTEMPT to do. This is a very daunting task. For example, the TAKS test attempts to assess across all grade levels and ability all in a single test. Most educators yell foul, yet I have heard no other than Bill Daggett, on more than one occasion, pronounce the TAKS test is a pretty good test.
Now, going back to the premise of a teacher creating a test to reliably test across a variety of abilities, I am with Cain on this one. I think it is theoretically possible, but I also think adult bias will creep into the test, yielding self-fulfilling results. Again, test writing is a profession in it's own right. As I have said before, teachers fancy themselves as great assessment writers; in general, we are not.
However, having said that, the poster certainly speaks of a type of differentiation I am fond of: grade differentiation. That is, I like the idea of contracting for a grade as a differentiation practice. A grade of C may only require a demonstration of knowledge/comprehension mastery with very low relevance. This is the “I know the correct answer when I see it” level, which is the ability to pass a standardized test. This may use a rubric: for a C, you must do three of the following assignments. For a B, you must do the work for a C grade, plus at least 2 of these 4 assignments. The B assignments are at the application/analysis level with a relevance level at least at the application level in the content area. For an A, you must do C, B, and 1 of three possible advanced assignments. The A level assignments are at the synthesis/evaluation level, high relevance. In my mind this is using the grading policy (some kids care about this) in order to promote differentiation.
The writer's basic premise is consistent with the ASCD approach to differentiation. I don't totally reject the idea of differentiating tests, but then again, I don't embrace the idea of any test, as I see them as very limited in their ability to assess.
I am not sure what the writer means when they say “Differentiation is differentiating objectives, not interest right?” Given the ASCD model, differentiation means providing opportunities to stretch children according to their ability level. In the ASCD model, students may initially select their own level of stretch, with the teacher serving as the safety mechanism to insure students don't “slack” themselves. Certainly in the ASCD model the objectives for each level of learner is differentiated, and as I pointed out, a modified grading system is ideal to promote this idea. I would say then differentiation is providing students with various levels of rigorous instruction as defined by Bloom and relevance as defined by Daggett, Ericson, LYS, et al. Can this be done in a test? Maybe.
As Cain points out, differentiation can be interpreted as “the differentiation of strategy, practice and intervention to insure success.” I have pointed out valid methods to differentiate, which are consistent with both the writer's and Cain's interpretations. I do not reject testing as a way to differentiate, but I would certainly not embrace it due to the complexity of creating valid tests. I would prefer the assignment/project model in the classroom.
T.W.A. - Doc Seabolt
SC Response
I have to think about the concept of using a variety of assignments selections to determine a grade. It seems a little like what Glasser was writing about in his “Quality Schools,” books. Part of me likes the idea. Part of me is loath to the idea of letting students determine when they are going to stop. I can see adults using the idea of “student choice” as justification for a lack of student motivation. As you know, I’m not a fan of letting people choose not to be the best (the absence of leadership). But, in the right setting, with the right mix of staff and students, I’d pilot it for a semester.
Think. Work. Achieve.
Your turn...
Follow Sean Cain at www.Twitter.com/LYSNation
Pre-order your campus copies of "The Fundamental 5: The Formula for Quality Instruction" www.TheFundamentalFive.com
Attend the LYS presentations at TASSP and TASB in June