Thursday, August 27, 2009

A Model for Assessment: the Road Test

As a native New Yorker, I grew up without any incentive to learn to drive. No one in my high school could drive, and many of my friends were from families that didn't even own a car. Now, after a year of living in Albany and growing increasingly frustrated with its inadequate public transportation, I am finally learning to drive. And boy, do I suck at it.

I have lots of excuses. You see, I am a very focused person. When I'm concentrating, I hardly notice my surroundings. You can call my name and I won't even look up. So driving, which requires diffused attention, just isn't my brain's style. Furthermore, I'm slow to process what my eyes see. I can look in my blind spot and still not notice the car on my left. I'm just not much of a "visual thinker"; I'm too abstract. Probably the thing that really makes driving difficult for me is that I'm plain terrified of it. Everyone tells me to relax, but I think everyone else should get a little less complacent: driving is dangerous!

Of course, none of this makes one whit of a difference to the DMV. No matter how much I've improved, no matter how far I've come in overcoming my fears, no matter what kinds of mental handicaps I have, I am not going to get a driver's license until I demonstrate that I am a competent driver, the same as everyone else.

This is in stark contrast, of course, to the sorts of accommodations I've sometimes been asked to make for students who for various reasons "don't test well" or just "aren't math people". Now, I absolutely believe that there are people for whom the kind of thinking required in mathematics is a particular challenge, just as I believe driving is a particular challenge for me. But the difference is that some students and parents think this means that they should receive different assessments, or a different grading structure, so that their children have a "fair chance".

Implicit in these requests is the notion that competency in math, unlike driving, is not essential for the general populace. If people actually thought that the math test they were being asked to pass tested important, meaningful understanding that everyone should have, it would be obvious that making accommodations would do a disservice to the student as well as the public. Rather, most people seem to hold the opinion that math tests are meaningless hurdles and that effort, rather than competency, is what the grade should reflect.

Clearly, I disagree with this position: I do think there is a real danger in a totally innumerate population, and that mathematical understanding is something everyone should possess to some degree. However, that doesn't mean that my assessments automatically reflect that. If I want to toe the hard line and impress upon my students and their families that an A means excellence, not effort, that knowing what I'm teaching is worthwhile, even if it's hard for you and you're scared of it, then I have to do what I can to align my grading with those ideals.

Here again, the DMV provides a model. I failed my road test today. To a truly embarassing degree. (I cut off a guy in a tow truck and he got mad and started tailgating me.) However, at the end of the test, instead of receiving a big F or a judgmental sermon on how terrible I was, I got a lot of sound advice on where I needed to improve, and some tips that would make it easier. I even got a printed list of every stupid error I just made. Now initially, being faced with a long list of your specific failures hurts. The DMV doesn't sugar coat it when they say "dangerous violation". But this is targeted remediation of the most helpful sort. I know exactly what I should be working on as I continue to practice.

And there's the best part: I will continue to practice because I get to try again. It's not all over. In fact, I can just keep taking it as many times as I need to until it's safe to let me pass. The DMV doesn't care that I learn to drive by a specific date (heck, I'm many years past due by most standards) just that I learn to do it safely once I've decided to make it a goal.

My takeaway lesson from this is that in so far as any assessment is supposed to promote skill mastery, it should provide meaningful feedback, hold students to fixed standard of performance, and allow students to make multiple attempts.

All of this is probably preaching to the choir. What actually surprised me in thinking about this is the contrapositive. In so far as my assessments do not reflect these principles, they are NOT promoting skill mastery. Which is usually just because they are promoting something else. I let students retake quizzes but not tests, for grading purposes but also because deadlines are a part of life. I don't hold all students to the same time-limit for tests because even though I think speed is an aspect of mastery, I think it's expendable. I believe it is fine to make concessions like this, because there actually is more to assessment than the specific skills they test, but it was kind of unsettling to realize that there's a little bit of "meaningless hurdle" in my assessments, after all.

....................................................................................

This post is already too long, and the analogy is imperfect, but I still can't stop thinking about what a great assessment the road test is. In fact, there are two aspects of it that are so wonderful they are mostly beyond my ability to mimic and thus were left out of the previous analysis. The first is that what it actually assesses, more than skills, are habits. I am actually an ace at parallel parking by now because I practiced it over and over. But what got me on the road test was just dealing with intersections I had never seen before. Learning to drive is not just learning to repeat a set of motions. It's learning judgment, informed by an understanding of the rules and experience dealing with a variety of situations. On the road test, you don't get a comprehensive run through of every possible turn and intersection, but you get a reasonable sample that's a little different than you've seen before. Would that my math tests were the same.

The second aspect, which I can never match, is that you don't have to take the road test until you choose to. In fact, you don't have to take it at all. Generally, there is a great incentive for teenagers to learn to drive, but if you had forced me to take it at age 17, I'm not sure I could have survived the process. It is an immensely difficult task for me, and though every else seems to remember it as easy, I really think that's just the rosy tint of nostalgia. Having made the personal decision to drive, out of an awareness of the difference it will make in my life, is the only thing that is motivating me through the process. I often regret that my students do not have the same aid in their struggles with mathematics. They can't imagine how it could help them, and I can't let them just hang around until they figure that out. The best I can do is try to paint a picture for them, of what it would be like to speed down the freeway, with the wind in your hair, and get weirded out by the fact that your speedometer seems to report your speed over an instant rather than an interval.

6 comments:

  1. Hubert Dreyfus writes a lot about the philosphical debates between intentionality and comportment. The simple distinction would be: do you do something with rules in your head that you apply or do you do something because you are 'comporting' yourself towards it (e.g. knowing how far to stand from a painting).

    Driving is one of his favorite examples: you can't plug and chug your way through a driving test, you have to just flow because you are comfortable with the whole experience (including unpredictable elements).

    Maybe the same thing follows for math? You can't 'teach' orignal problem solving, but you can get people comfortable enough with elements of the experience so that with some practice they might start solving new problems on their own.

    The downside, I guess, is that it this approach questions how much can really be taught, and how much is just shepherding the person through the uncomfortable early practice stages? How to make a teaching method that actually *develops* expertise and doesn't just provide practice is beyond me. Maybe engaging opportunities for practice are the method?

    (http://socrates.berkeley.edu/~hdreyfus/html/paper_socrates.html)

    ReplyDelete
  2. John, thanks for the article. It provides a very well-defined and highly relevant rubric for levels of understanding. I think in the school setting why Dreyfus calls "Novice" is considered competence, and what he calls "Advanced Beginner" is considered expert.


    Even if this approach points to the limits of direct instruction, that doesn't mean people can't be lead to a higher form of competence by some means. Obviously people do master driving, at least. Providing "engaging" opportunities does seem to be the most effective tact for an instructor to take, if by "engaging" we mean "challenging, but not overwhelming". The best thing a driving instructor can do is lead a student into situations which she can handle, but are slightly outside her range of experience and comfort. In terms of mathematical problem solving I don't see why the same principle shouldn't apply.

    ReplyDelete
  3. Maybe this is related to alternative education models: the montessori or the unschooling challenge. You don't need organization of lesson plans, just opportunities for practice that the kids can stick with. Presumably if they continue to practice the skill in an uncertain setting, they can grow into expertise.

    Engaging, from my perspective, just means something that you will continue to do. Even a hard problem that you can't solve might keep you failing or taking risks for hours. It's hard to say that that is not a valuable learning experience. The repeated engagement, or the incentive as you put it, seems to be the critical element in building expertise.

    This leads to your insight about retaking tests as practice. Why not regive a test over and over (maybe with incremental additions). Practice with an incentive! Does it really matter if everyone gets an A?

    Related?
    http://www.gladwell.com/2008/2008_10_20_a_latebloomers.html

    ReplyDelete
  4. John,
    I love reading about teaching models that dispense with lesson plans, but even when I try simply to present a single problem for a class, I find I still need a plan for how I will present it, how I will ask the students to attempt it, what kinds of intervention I will supply, and what kind of follow up work the students should be expected to do. Perhaps I am just not skilled enough yet to do all that off the cuff.

    I absolutely agree that a task can still be a learning experience even if it all you get out of it is a series of failures. However, given enough tasks like this, the student comes to the impression that they are just incompetent which diminishes their inclinations to keep practicing. Part of the challenge is not only finding a task which is a learning experience, but motivational as well. Otherwise instead of repeated engagement you just get misery.

    I actually will allow students to repeat quizzes up to three times. These quizzes are 60% of their grade. I just can't let the retake the one test per quarter I give them, because I give them at the end of the quarter and need to get the grades in. That's how it goes sometimes and kids gotta deal.

    Will read the article. I can't think of a single mathematician who was a "late bloomer".

    ReplyDelete
  5. Nice analogy. It's got me wondering what would a more authentic assessment of mathematical learning look like? Written tests are efficient and easy to manage with a big group of kids, but by and large they're not going to be asked to do that as adults.

    ReplyDelete
  6. good stuff indeed. there would be a lot more
    testing situations like "driving" if schools were
    about learning instead of certifications:
    pass/fail, high stakes, repeat ad lib
    (but with mandatory cooling off periods).
    or so it appears to me.

    weierstrass is sometimes cited as a late bloomer.
    meaning his thirties, i think.
    i haven't read the link in question
    and doubt that its terminology
    applies here though.

    i also recently failed a driving test.
    some coincidence!

    ReplyDelete