The Perfect Assessment
by Terry Heick
Nothing is perfect, but we can dream. So let’s dream about assessment.
First, what is an assessment? A measurement? A snapshot? A kind of bar for students to clear? In most important, I theorized that “the benefit of assessments for learning isn’t merely a more clear picture of understanding; Used properly, it can also inform the rest of the learning process, from curriculum mapping (what do we learn when?) to instruction (how will it be learned?) to assessment design (how should future learning ideally be measured?)”
If the goal of our collective craft is understanding, than the tools we use should promote understanding, both directly and indirectly. Assessment is one of those tools–one widely misunderstood by teachers, and causing anxiety in the students we’re trying to serve. So then, can we do better with our assessments? I want to think of assessments in a completely different light, especially in light of modern technology. Take a clean sheet design approach to how we think of the word “test” (the patriarch of the assessment family).
So what would the perfect assessment be like? If we can design anything–not just digitize multiple choice questions, but start from scratch? What would the perfect assessment do? How would it be formed? What data would it yield? What effect would it have on the student?
How would it be used to improve learning? How could it server students–like stairs–instead of being obstacles to clear, like hurdles?
How can it promote understanding without haunting students?
There’s no single answer here because there are too many moving parts. So we can’t hope for perfection, but we can hope for perfect alignment between goals and function: What we want, and how we hope to achieve it. It’s not impossible, then, to begin to identify a set of indicators of a near perfect assessment–what it would and would not do, for starters.
Below, I guess at some of these indicators. I didn’t have a particular assessment form (MC, essay, performance task, project, etc.) or mode (norm-referenced, criterion-based, etc.) in mind. Clean sheet design and all. (You can read more about different types of assessment, or see 10 assessments you can perform in 90 seconds or less as well.)
I was more interested in the function of assessment as a tool for learning, and what we might be missing.
32 Characteristics Of The Perfect Test
- …will be in the form and mode that will help the students reach their goals, not the institution reach its goals
- …will provide data to revise planned instruction
- …will show both short and long-term progress
- …will adjust in real-time and scale (in complexity, knowledge demands, etc.)
- …will compel students to respond with their best effort and particular genius
- …will produce easy-to-extract, usable data that both teachers and students can understand and use
- …will use transfer as an indicator or degree of understanding
- …will be based on a specific learning taxonomy
- …would allow the students to enter a state of “flow”–a complete–and perhaps playful–state of mental and emotional immersion, where they give themselves entirely to the task
- …will have multiple entry or starting points
- …will uncover both equally what a student does and does not understand
- …would be a learning experience in and of itself
- …would be fun (i.e., as a basketball game is an immersive and entertaining “test” for athletes)
- …will align exactly with the stated goal–an academic standard, future aspiration, personalized learning desire, community need, etc.
- …will use a scoring system that reflects degrees of understanding, progress, or mastery of individual line items/standards (rather than a gross score for a mash of “things)
- …will provide a clear starting point forward for both the student and the teacher
- …will allow the students to use their inherent strengths to compensate for their weaknesses (as adults do in their daily lives)
- …will give students hope
- …will yield compelling artifacts to bolster student portfolios and/or improve that student’s human circumstance (e.g., a product that improves their lives/community)
- …won’t mistake confusing with complex
- …won’t drown the teachers with follow-up work or other processes that keep them from doing anything else other than using that data to revise planned instruction for each student
- …won’t be designed to yield unusable or irrelevant data
- …won’t be widely misunderstood by parents, families, and communities at large
- …won’t required students to sit in a desk in a room
- …won’t be a matter of “pass or fail,” but rather start here and move forward
- …won’t be designed in a such a way that one error here can allow several errors there (e.g., a math problem where if the first problem is solved incorrectly, the rest of the problem can’t possible be right)
- …won’t have built-in barriers that obscure understanding of exactly what’s being assessed (e.g., a complex text that demands strong reading skills when it’s knowledge of the water cycle that’s being assessed)
- …won’t resort to distractors or tricks as a test of a student’s “grasp” of the content
- …won’t have inherent cultural biases (e.g., in regards to race, gender, economic class, faith, or other “human” factor–see Diane Ravitch’s The Language Police for some background here)
- …may use simulations or scenarios that invite students to use contextual cognitive processing–thinking in relation to circumstances they can play with or are naturally drawn to or are authentic–to agitate and coalesce academic knowledge
- …may provide scenarios for students to think their way through, providing an authentic context, need to know, and opportunity to transfer understanding
- …may not have a beginning or an end, but rather function as an ongoing, iterative effort
*previously published at TeachThought.com