20 June 2010

Graph presentation and the Duverney Standard

We all ask our class to label their experimental graphs clearly. That seems like such an easy-to-follow instruction. But remember, students come into our class with widely varying lab experiences. Some of their previous teachers have probably not cared about whether axes have correct units, let alone labels. Other teachers might have been almost obsessive-compulsive about the format of titles and labels.

Remember that high school students generally view laboratory work as merely a set of steps to be followed, at least at first. They don’t understand why clear communication of results is necessary. After all, they think, the teacher told us exactly what procedure to use, what to graph, and possibly even what our graph should look like. So why wouldn’t she understand our labelless, unitless graph?

Such reasoning is rationale #1 for allowing, encouraging, and requiring some freedom and independence in laboratory exercises. Students must see that they are undertaking an active investigation, not just carrying out a rote process. But I digress. That’s a topic for another post.

I’ve noticed that, no matter how intense my instruction, constructive criticism, red ink, and general nagging, my students’ graphs continue to look like crap… except for folks in the research class. They usually have an epiphany after their first informal presentation to their classmates, who ask pointed questions about the experiment. Suddenly, the presenters see the confusion generated by an unlabeled graph. They look at me sheepishly while I give them the “I-told-you-so-for-two-years” look.

Taking a cue from my research course, I’ve had some success improving graph appearance in first-year courses by clearly elucidating the intended audience for the graph. It’s not good enough that they themselves understand what they mean; it’s not even good enough if they think that I should know what they mean. Their graph should be clear enough that another student at the same level of physics should be able to tell exactly what they did just by reading their graph.

I called this the “Duverney Standard,” named after a popular student who graduated the year before. I expect that Mr. Duverney, despite his ignorance of a particular laboratory activity, should be able to read a graph alone and interpret the experimental results for himself. (Of course, this year the standard will acquire a new name. Perhaps I could sell the naming rights to a graduate for a donation.)

Now, I’m not particularly OCD about graph presentation. After all, Mr. Duverney is known to be a smart gentleman. If a student chooses to use a clear title in lieu of verbose axis labels, fine. If the axis labels are clear, then a title isn’t necessary. The standard is simple and non-negotiable. When axes are labeled merely m (kg) and t (s), I take off significant credit, and simply write “Duverney.” If a student asks about the lost points, I ask whether Mr. Duverney, having looked at the graph, could reasonably reproduce the experiment. That’s usually a sufficient explanation to extract the sheepish “oh” from the student.

Now, Defcon II in the argument would be to scan a copy of the offending graph and send it to Mr. Duverney for comment. There is no doubt in my class’s collective minds that I am willing and able to carry out such a threat. Thus, I’ve never had to actually do so.




1 comment:

  1. One thing I've found that helps is to ask students to interpret someone else's graph. You can find a bunch of student work for science lab (of varying quality) in a variety of subjects here: http://investigationexchange.org

    ReplyDelete