Digital Assessment for Digital Learning: Grades that Mean Something

by

We are long overdue for a fresh look at the way we do assessment. I’m sure we all realize that measuring student progress is not the same as measuring a window for drapes. The very act of measurement can have a huge motivational or de-motivational effect on the very students we are measuring. It’s the Heisenberg effect writ large. Keeners behave like lawyers, wheedling and pestering their teachers for every possible extra half-point, while those at the opposite end of the bell curve go through the motions, trying to get the bare minimum to “pass” a course. Students think grades are the end goal, and want them to be “better,” when the watchword should be “accurate.”

Even teachers may get stuck on the model of “grade as a reward.” When a student is “so close,” the temptation is often overwhelming to make the student jump through a hoop or two so we can rationalize giving the student the exact number of bonus points they need to get a passing grade. Do we really feel we’ve provided some kind of transformative learning experience as a result? No, we aren’t thinking about learning at all. We’re in pinball mode. Five points for this assignment, fifteen for that. Some students are shooting for the high score, while others are just trying to get through the round. In either case, they’re missing the point, and so are we.

This doesn’t hurt learning just for weaker students. Our top students go on to plan out their university course schedule to maximize their GPA rather than stretching themselves as learners (they may already practise this skill in high school). No high-achiever is going to risk her med school application on an interpretive art course where she can’t be certain of earning an A. The future lawyer won’t tread into unproven ground on a major course essay. The high achievers may actually be amongst the least likely to embrace the ideals of a liberal arts education, and it’s the fault of our pinball point system for punishing students for challenging themselves, and divorcing grades from actual learning.

In a recent article, Robert Sternberg facetiously argued that college classes should be replaced by test prep courses (“Who is Really Adrift?”, Inside Higher Ed). The immediate sense of wrongness this image provokes is enlightening. We actually do understand that cramming for a test and struggling to really understand something are two different things. It’s just that assigning points and determining some kind of average is all most of us know.

We have another problem. Ask yourself, what is the significance of earning 70% in a mathematics class? Does a student have 70% of the skills taught in that class (and if so, which 70%)? Can he utilize all the skills, but with only a 70% success rate? Is he better than 70% of his classmates? Perhaps it’s one of several billion combinations of these things. And right here we’ve already put our finger on one of our grading system’s fundamental flaws. Determining final grade is always an irreversible operation. We can arbitrarily pick any number of tasks to give some kid a final grade, but we can’t take that same grade and reproduce his work. So not only are we encouraging students (and teachers) to focus on the wrong thing, we don’t even know what that thing actually means.

If you really break it down, learning is digital. Every skill or concept is like a single bit of data with two possible values: yes or no.

We don’t have to do it this way. If you really break it down, learning is digital. Every skill or concept is like a single bit of data with two possible values: yes or no. Our true assessment data, with no loss of information, should be a list of learning outcomes and a yes or no answer on whether a student has mastered each one. Really, this is very obvious, and recognition of this issue is not original to me.

Working within the confines of the current system, a few schools have tried to change the current mode of thinking. Some US charter schools require students to achieve an A in every course in order to pass. Though it doesn’t get to the root of the issue, they have effectively disallowed the “just enough points” attitude at the low end of achievement. A more flexible and, I think more promising approach is Standards-Based Grading (SBG), which aims to more closely tie grades to learning outcomes.

The core idea of SBG is simply that assessment is intended to give information about a student’s mastery of outcomes, and not about their attitude, charm, or even time management and work ethic (though some might argue that those could also be outcomes). In practice, this means that the only thing determining final grade is what a student actually knows and can do by the end of the course. A student who shows no understanding of reducing fractions on a test early in the term but later displays a clear and sustained mastery of that skill will be deemed to have met that outcome. As long as the term is still in session, a student potentially has the opportunity to master any outcomes that they may have had trouble with. But the key is mastery, not make-work.

Ultimately, most schools using SBG still require a final numerical grade. One approach is described by teacher and blogger, Kelly O’Shea. Her school’s science department has categorized curricular outcomes into basic A objectives and deeper, more complex B objectives. Passing the course requires mastering every single A objective, while getting a grade of 90-plus additionally requires all B objectives. The upside of this approach is that it could be implemented at the classroom level by a teacher working alone, while still promoting a very different attitude of what students are there for.

The downside is that we’re still tied to this idea of the one number, which tries to sum up an entire course. We bundle together our learning outcomes into packages we call English Lit or Honours Chemistry because these skills are related to and dependent on each other, and it makes it easier to plan our teaching schedules. But along with this, rather than making students accountable for each outcome, we make them responsible for the course as a whole, knowing full well that many students receiving course credit have managed to avoid achieving at least some of the outcomes.

There’s a different way we could do this. It seems to me that the natural next step from Standards-Based Grading would be something like “Standards-Based Credits.” We could cut our outcomes loose from the straitjacket of a standardized course. According to O’Shea’s blog (“31 Reasons Why Kids Like SBG”), students like being able to move at their own pace, still being held accountable for mastering each skill and concept, but not within a narrow learning schedule. But the obvious extension of SBG is to allow students to meet outcomes individually whenever they are able to, even after a course is finished.

This isn’t an invitation to take it easy and get caught up later. It’s a recognition that some students and some concepts may need to be revisited a few times before mastery occurs. And it hopefully gives credit where credit is due, particularly for students who challenge themselves by going outside their comfort zone academically. With Standards-Based Credits, students can be recognized for the outcomes they have mastered, rather than denigrated for the ones they have not.

Under this system, courses would become less of an absolute list of required topics and more of a forum for learning related skills. A student wouldn’t end up with a percentage representation of his understanding in the course, but a list of specific learning outcomes where consistent and sustained mastery was shown. I don’t think our teaching would have to change dramatically under this system, though I do hope our students might shed some of their extrinsically-motivated behaviours and focus more on genuinely understanding something, and perhaps take some risks.

Our current high school transcripts are analog, treating courses as a single unit, with the possibility of virtually any percentage grade for each of those units. But under Standards-Based Credits reports would be digital (more specifically, binary), with just a yes or no on outcomes achieved. A student who sat in on an advanced forum that he’s not confident he can handle would not risk an F by trying. At worst there would be no mention of his academic exploration, at best, one or more outcomes he mastered in the course would be included in his report. Under this system, a student’s cumulative report records would look very much like a curriculum vitae of skills.

We may still award high school diplomas for those students who have met a certain list of required outcomes, those which we feel that every good citizen must master. Schools could certainly continue to recognize special student achievement in whatever way they felt appropriate. Somebody who has met a large number of outcomes in history, particularly some of the higher-level outcomes, might be a “history scholar”; a person who has met a large number of total outcomes in a variety of different subject areas might be recognized for their willingness to stretch their learning.

Behaviours that we wish to promote and recognize but don’t currently do so in any formal way could now be more easily accommodated. This could include responsibility, punctuality, time management, leadership, even courtesy. Little things that were too small to be considered a credit can certainly be considered as learning outcomes.

Post-secondary institutions could now be more specific about what they want prospective students to really know rather than vaguely requiring a minimum percentage average in certain courses. Students might choose to sit in on a more advanced mathematics forum just to learn one particular set of skills required for a job or training program, where in the past perhaps they would have thought that educational path to be out of their reach.

In the end, I think this system would hold students more accountable, they would learn more, and we would have more complete and accurate information as to the nature of this learning. As with any untried system, there are likely to be bumps, unintended consequences, and surprises along the way. But we can’t know unless we test it. What I am proposing is that we at least consider that the way we’ve done things for the last hundred years isn’t necessarily the only way.


ABOUT THE AUTHOR

J.J.S. Boyce
J.J.S. Boyce is a Canadian science teacher and writer, and tries to use both sides of his brain regularly. His articles on science and education have appeared at the University of British Columbia’s Terry Web Site, the Science Creative Quarterly, and Suite101. He also writes regularly about culture and literature for the Green Man Review, Sleeping Hedgehog, and Blogcritics.


This article is from Canadian Teacher Magazine’s Sept/Oct 2011 issue.

You may also like