A Smarter Way to Grade
Online portfolios allow engineers to be judged on what they produce, not on a number.
By Chris Rogers
Grades are an interesting conundrum. How do we reduce someone’s understanding of a subject to a single number or letter? What does a 3.2/4 in fluid mechanics mean? Often it means that the student tries hard, is fairly good at memorizing, and has done the homework. It might mean that the student really understood statics but got confused when friction entered the fray—or maybe she did not really understand statics, but Bernoulli’s equation made sense. How well does that single number reflect her ability to watch rain pouring off a roof and quickly model it, estimating the flow velocity, or being able to look at water coming out of a faucet and explain why it is moving the way it is?
The problem is grades are very convenient—we can easily average them all together to give a person a GPA and through that somehow rank him relative to his peers. We can quantitatively analyze the spread of learning across the classroom and get a better understanding of how effective we are as teachers—or can we? Take Sally, a highly creative student in that fluid mechanics class. She thinks a lot about the fluids problem and decides to take a risk (after all, what is creativity if there are no risks or failure?) She attempts to take friction into account but therefore gets the answer “wrong,” and the grader gives her an 8/10. Fred had simply thumbed back into the chapter, seen equation 3.2—it looked appropriate—and applied it to the problem, and got a 10/10. Who actually has a better understanding of fluid mechanics? Is there a reason that we continually hear stories about successful entrepreneurs getting mediocre grades in school?
There is a secondary problem. As teachers and mentors, we are heavily invested in helping our students learn the material. We work with them during office hours, stay after class, answer their e-mails, and generally support, encourage, and push them. After all of this, how are we supposed to objectively score them? How can we impartially judge those we mentor? It is easy if we do not coach, but as soon as the coaching starts we are about as effective at grading as parents are at objectively judging their children.
So what can technology do to help us change how we grade? It allows us to move away from the “one number that rules them all,” and move toward something like art appreciation. No one listens to Joshua Bell because he got a 3.8/4 in violin class. We listen to him because we have heard him play a piece of music we love and we appreciate the clarity and strength of his tone. How can we do the same thing in engineering? Just as we judge Bell based on his recordings, electronic portfolios will allow companies to judge engineers based on the engineering they have done. Imagine you are looking for a roboticist and you come across a website of a student who has built a humanoid robot that reads your emotions and helps you with daily tasks. How would this affect your hiring decision as opposed to the grade he got in his introductory robotics course? Even better, on another student’s website, you see the drone she built, the auto-drive car her team developed, and the LEGO robot she and her friends took into a local classroom. You now have a pretty good picture of her robotics skills—much better than the 3.3 GPA can express.
While robotics is intrinsically project based and therefore lends itself to a portfolio, what about courses like introduction to fluid mechanics or controls or dynamics? Even a portfolio of the types of analytic problems the student has solved helps in better understanding his fluids knowledge, giving us more insight into what he learned and how well he did. More exciting, though, would be to teach an entire fluids class around producing potable water. Students would learn statics as applied to a water source, calculate pressure drops in pipes leading to the purifier, grapple with concepts of friction, model momentum losses in their system, and see how well Bernoulli’s equation predicts the behavior of the purification system they build.
I look forward to the day when the GPA option disappears. We, as teachers, would no longer have to judge but instead would supply problems to solve, encouragement, just-in-time help, and written recommendations to help a student understand her strengths and weaknesses. Her work could stand on its own.
Chris Rogers is a professor of mechanical engineering at Tufts University. crogers@tufts.edu