Gas station without pumps

2014 May 21

Establishing the habit of writing

Filed under: Circuits course — gasstationwithoutpumps @ 09:19
Tags: , , ,

In Preparing for AP Physics 1: establishing the habit of writing Greg Jacobs writes

I’m in the infant stages of planning my AP Physics 1 course. The big trick is going to be establishing my students’ ability and willingness to write their reasoning, to get them to focus on communication rather than on getting a correct numerical answer. Once it’s clear that they are not taking a math course—once they see that the solution to a problem looks much more like what they’ve done in biology or economics than in calculus—I think the students will be able to move along quickly and enthusiastically through the material.

Students must get comfortable with calculation. However—as was correctly pointed out to me at the AP consultant meeting in April—if we start the course with lots of pure calculation, students will think that getting the answer is the holy grail of physics problems. If instead we begin the course demanding description, explanation, and all sorts of prose, students may become accepting of the idea that a numerical answer is merely the result of careful reasoning.

If this change in AP Physics actually works (something I’m always skeptical about in any curriculum reform, particularly at the high school level), it may help engineering students in college. Engineers do far more writing than most professions, with far less training at doing it.

I don’t think that a prompt that just says “In a clear, coherent, paragraph-length explanation, describe how you would figure out …” is going to do the trick, though. If they could already write clear, coherent paragraphs about how they would figure something out, then they would not need the curriculum change—they might not even need a physics class at the level of Physics 1.

I’m struggling with this problem in my applied circuits course, in which I require weekly design reports for the circuits they design and build. The students are staying in lab until they finish the designs and demo them, so they are clearly capable of doing the work (though not always as quickly as they should). But only a few students can explain their computations for the design parameters (like gain, corner frequency, and component values) clearly—others put down any nonsense that has a few of the right buzzwords in it.

The top students have gotten better at their explanations as a result of feedback, but the bottom students are still often producing word salad. Although there is some indication of a general writing problem (lack of topic sentences, poor grammar, and misused vocabulary), the problem is most pronounced when they are trying to explain how they selected component values. The more steps that there are in the underlying math, the more jumbled their explanations, even if the problem is just a chain of multiplications.

From time to time, I’ve suspected that the students don’t produce coherent sentences about how they computed something may not have actually done the computation, but “borrowed” the result.  This is not an explanation I believe in strongly, though, as the students have been (mostly) coming up with different solutions to the design tasks, so there isn’t simple copying going on. I’ve also seen the design process the students use, as they have been doing their pre-lab work in lab (instead of at home), so I hear them discussing the problems.  They do ask each other not just what answer they got, but how to get the answers, so they are trying to learn the method.

In looking at the pre-lab homeworks that were turned in on Monday I realized what part of the problem is—the students keep absolutely awful design notes. What the students turned in on Monday (even the top students) was mostly incomprehensible scribbling of numbers, with no indication where the numbers come from or what they were attempting to compute.  Half an hour after writing down the notes, I’m pretty sure that they could not reconstruct their reasoning—hence the often magical methods in their design reports, where they copy numbers out of their notes (some of which are correct), but can’t put together a coherent chain of reasoning that leads to those numbers. On the long multi-step computations needed to figure out what gain an amplifier needs, they can usually do each step (though often needing coaching on one or two of the steps, either by me or by one of the better students in the course), but they don’t record the meaning of each step or even what the sequence of steps is, and the “answer-getting” mentality causes them to flush the process from their minds as soon as they have a number.

I’ve seen a lot of lab exercises for other courses that try to scaffold the process by providing worksheets that give the step-by-step process and have the students fill it out as they go along. I don’t think that this is helpful though, as it encourages students to solve one step at a time and then forget about it—the scaffold prevents the students from exercising the very skill that I most need them to learn. Showing them worked examples, as I have done in class, doesn’t seem to help much either—they can follow along as I break the problem down with them, and think they understand, but then not be able to do the same thing themselves.  Again, the scaffolding prevents them from exercising the skill I most need them to learn—identifying problems and them into subproblems.

For next year, I’m probably going to have to come up with some exercises which get students to organize their thoughts external to their heads. So far, the only thing I’ve thought of is to have them create a fill-in-the-blank worksheet for each lab (like an income tax form), and turn in the blank worksheet and try filling out each other’s worksheets.  If they get in the habit of writing down the steps as steps, it may help them be able to reconstruct their work when they convert it into full sentences for the final reports. It may be too late for me to do anything formal this year (only 2.5 weeks left), but I’ll suggest it to the students anyway.

The advice I’d give to Greg Jacobs is to leave the “clear, coherent paragraph” until later in the quarter—get them to create worksheets first.

I’d welcome any suggestions from my blog readers on ways that I can get students to learn to organize their thoughts in a way that they can present them coherently to others. Block diagrams alone don’t seem to be enough, and vague things like “mind maps” are likely to do more harm than good.


  1. Well said. Teaching how to explain a calculation process in words is a difficult, year-long process. I love your points about the simultaneous benefit and futility of “scaffolding” the steps to a solution. I’ve had a lot of feedback from teachers I trust telling me to just throw the class into the “clear, coherent paragraph.” Sure, I’m going to have to give all kinds of constructive feedback. I’ll see how it goes. I certainly don’t expect beauty on day 1; but I’ve been convinced to show my hand up front, so that students see what they’re expected to do right away, even if they can’t yet meet my full expectations.

    Comment by Greg Jacobs — 2014 May 21 @ 09:34 | Reply

  2. In my household of engineers (plus an AP Physics student), we always start with a diagram and a list of knowns. I don’t think any of us know how we’re going to solve something until we draw it – we need to translate from words to a physical representation before we can even start to think. It *does* drive my HS junior a little batty when she asks for help and the first thing either parent does is redraw the problem. In my view a “clear, coherent paragraph” detailing solution method is an end product, one that is not always needed and not something that one would produce at the start.

    I had two thoughts about the needs for your class. First, it sounds like the students have not been taught how to keep an engineering design notebook. I also wonder if a specification, as opposed to a worksheet, might be a useful format for organization of their thoughts. You could give them a basic template so they’d have an idea of what matters. I’m thinking here of some of the informal design specs I’ve used that might start with goal/objective, followed by sections that make sense for your purpose. Some of mine might have sections for parameters, variables, components, metrics, required input interface, outputs, diagrams, test plan, implementation plan. Format varies depending on what seems needed for the project at hand, but should contain the information that needs to be shared if there are multiple people.

    Comment by kcab — 2014 May 21 @ 11:22 | Reply

    • Their “projects” here are one-week projects, so there isn’t much time for long plans. The design reports I’m getting from the students have the right structure and all the needed parts, so I don’t think that a template for the reports would change anything.

      It is true that they have no idea how to keep an engineering notebook, but I’m not sure I can be a big help to them there, as my own “notebooks” are not a good model. I tend to do mainly computational projects and keep my notes in a README file for each project. This blog is as close as I come to an engineering notebook for hardware projects. I do schematics on scraps of paper then redo them in SchemeIt (usually, but not always, before I try wiring them). My calculations for the parameters are usually on the same scrap of paper, so not much better than what the students are doing, though I can reproduce my work more reliably than the students seem to be able to.

      I do a fair amount of “tinkering” rather than engineering by design for the hardware circuits, and my class this year may have leaned too heavily towards design computations. I may want to have them do a little more tinkering next year (adjusting gain to the right level, for example) and reverse-engineering to figure out how big the input signals are, rather than trying to compute the signal sizes and design the appropriate gain.

      I’m beginning to get them to understand and use block diagrams, and I have gotten through that they need to have a schematic before I’ll help them in lab. I think I could add (next year) the expectation that they have made a worksheet for their computations.

      Comment by gasstationwithoutpumps — 2014 May 21 @ 12:26 | Reply

  3. Issues like this run across the curriculum. Math students that cannot or will not show their work, be the solution right or wrong, programming students that will not work out algorithms on paper before trying to type code, students not being able to take understandable notes during a chemistry experiment and so on. The students will always argue that since they got the solution all the other stuff is not needed. I am thinking it is a philosophy that needs to be changed, not a ability that needs to be taught. I have tried to address the issue by discussing a problem in class before trying to solve it. kcab’s “redraw the problem” is exactly what I am talking about. My trig students refused to draw pictures before trying to solve a problem. I started grading the pictures. High schools seem to love 4 years of required English literature. It would be nice if they bagged one year of Shakespeare for a year of technical writing.

    Comment by gflint — 2014 May 22 @ 07:33 | Reply

    • While I would love it if high schools replaced one year of “literature analysis” with one year of technical writing, I wouldn’t want Shakespeare to be what they cut. Actually, what I would like is for English classes to spend more time reading and less time on literary analysis. Writing classes should be entirely separate from literature classes—the two skills are almost completely unrelated and it is only historical accident that English teachers are expected to teach both. (My son has gotten much better writing instruction from history teachers than from English teachers, perhaps because the history papers had more content, and writing devoid of content is useless exercise.)

      I think that a large part of the problem seen across the curriculum is indeed “answer-getting”, and I’ve talked with my students about that (and about “imposter syndrome” and other meta-cognitive things). I’m not sure that I can change a mindset developed over 14 years in 10 weeks, though.

      Comment by gasstationwithoutpumps — 2014 May 22 @ 08:23 | Reply

      • We do 4 years of Shakespeare. Our kids come out excellent writers in English but it just does not seem to transfer over to the sciences or math. The “we should not have to do English in science/math” seems to be a mind set.

        Comment by gflint — 2014 May 22 @ 11:48 | Reply

        • I would love to see “writing” and “English” completely decoupled. The English teachers do a poor job of teaching writing anyway. Many colleges have decoupled them, though there is still an excess of English majors becoming writing instructors, and propagating poor writing styles that don’t work outside English classes.

          Comment by gasstationwithoutpumps — 2014 May 22 @ 19:29 | Reply

          • Yes!!! I have argued this for years. “Writing” is not the same thing as “literary appreciation”. They should be separate courses.

            Comment by Bonnie — 2014 May 23 @ 10:24 | Reply

  4. As a practicing engineer (network, not EE) I find this problem carries into the professional domain. While much of the network design and implementation work is templated, which basically means a PSM (pink, shaved monkey) that can read can accomplish the work (and there are PSMs in other countries who command much smaller salaries than my stateside compatriots…) , what separates the mediocre engineering troubleshooter from the exceptional is detailed notes.

    what is the problem statement?
    what are the hypothesis that might explain the problem statement, i.e. what fault domains would potentially account for the problem at hand?
    what tests do I propose to prove/disprove the hypothesis and what are the outcomes of each test?
    Did the test prove or disprove a hypothesis or were the results ambiguous and thus the test inconclusive?

    Only if there are clear notes about your train of thoughts, the tests and the outcomes can you make substantial progress when there are complicated problems, especially with problems where there are multiple faults (those problems were two or more things need to break before the problem manifests itself aka “planetary alignment” or the “bullet through the swiss cheese” types of problems )

    Only with clear notes can you stay on track when troubleshooting, i.e, we finished testing scenario # 7 at 3:42 AM and the test disproved the hypothesis whereas we never did finish testing scenarios # 3 & 5 because we got distracted and went down a rabbit hole with scenario # 4.

    And finally, and to point, only with the detailed notes, along with log files, screen dumps, etc, are you able to make heads or tails of what actually happened when the team reassembles to perform Root Cause Analysis a couple of days later, maybe after everyone has had two nights of +6 hours of sleep.

    While this does not specifically address the issues found in the lab, it does speak to the discipline needed to be an outstanding engineer in the field.

    BTW, my note have minute-by-minute time stamps in the margins, file names for any log files taken, who said what when, etc, etc. Often times when there are debates about what happened and when, a scanned page of my notes settles the debate, and I get to write those notes as I see fit while I am taking them.

    Really, really enjoy your blog…

    Comment by Richard Hardenstein — 2014 May 24 @ 18:07 | Reply

  5. Thanks—it is good to hear confirmation from a working engineer. As an engineering professor, I’m often a little afraid that ivory-tower thinking has crept into my course and curriculum designs, so it is good to hear that one of my objectives meets with approval from working engineers.

    Comment by gasstationwithoutpumps — 2014 May 24 @ 18:13 | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply to kcab Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: