Three Misconceptions about the Separation of Instruction and Assessment
In a recent article in The EvoLLLution, David Schejbal pointed out some advantages of separating assessment from instruction. The argument, essentially, is that since quality assurance happens as learning is assessed, we can improve consistency by having assessments that don’t depend on an individual instructor’s particular points of view. This argument, predictably, generated some impassioned responses that illustrated some common misconceptions about separating instruction from assessment. I’d like to clear some of these up.
1. Great Instructors Are Great Assessors
We’ve all had great instructors who knew how to make a subject come alive. But that doesn’t make them great at knowing how to assess what students have learned in a way that provides useful information to others. They may do very well at seeing each individual’s growth over time, but when it comes to reliably and clearly indicating whether students have met objectives, they may be just as prone as anyone else to bias. For example, higher grades might go to students who tried really hard and made great advances, as opposed to those who slacked off but who came in knowing more — or their assessments might favor writing ability over thorough analysis.
2. Learning Is Best Done in a Highly Individualized Way, So Assessment Should Be Individualized Too
Each student’s path to a thorough understanding of a subject will be different. Learning to think critically might transcend any particular subject matter. Isn’t it a good thing when the instructor understands each student’s biases and assumptions and challenges them effectively? Yes, it is. But the research methods instructor needs to know whether the students who took statistics really understood what they needed to understand to succeed at research methods. And the employer cares less about how students learned critical thinking skills than about whether they can apply those skills to new contexts. A highly individualized assessment tends to provide very little usable information.
3. Instructors Understand the Context in which Students are Working, So They’re Better Positioned to Create Appropriately Contextualized Assessments
Learning may need a context, but students ultimately need to be able to generalize beyond the specific context, and all too often instructors assess to the specific context. Students end up with grades that may indicate that they know an area of study, but in reality, their knowledge of the subject at large is narrow. In my undergraduate program, I had to pass exams in my major that were set by outsiders, people who were given the learning outcomes and reading list and grilled me on important topics in the field. I certainly didn’t rely solely on what my instructor had said, but rather studied diverse trends in my field to meet the outcomes — and I felt empowered and satisfied when I passed, confident that I had mastered that particular subject.
What all of these misconceptions have in common is the assumption that learning should be done in the same way as demonstrating what one has learned. It’s that assumption that we need to question. Just as teaching should be done by those who really know how to teach, assessing should be done by those who really know how to assess. This doesn’t mean all assessments should be multiple-choice, objectively-scored exams — many competency-based programs use trained raters who use complex rubrics to evaluate authentic student projects — but it does mean that having an unbiased third party not involved in the teaching can often provide better information about what students can do than an instructor who may have no training in assessment and a very small and potentially skewed sample of students to work with.
Author Perspective: Administrator