Skip to content

In defense of stories and classroom activities, from a resubmission letter from 1999

I was going through my files looking for some old data (which I still haven’t found!) and came across a letter from 1999 accompanying the submission of a revision of this article with Glickman.

Here’s a part of the letter, a response to some questions of one of the reviewers:

With regard to the comment that “You present absolutely no evidence that any of these demonstration methods is actually helpful. For at least a couple of these demonstrations you need to collect data to see if your tools are helping in understanding the concept. I will let you worry about how to measure this but this is a must”:

Of course, your statement is true, but consider the alternative, which is to do examples like this on the blackboard. We haven’t seen Moore & McCabe or Mosteller or anyone else conducting experiments to show that class-participation demos are _not_ better than straight lectures. And, given this state of uncertainty, we think that it’s useful to consider this alternative approach to teaching this material.

We agree that it would be a good idea for someone to collect data on the effectiveness of various teaching approaches. As all are well aware, this is a potentially huge research project. In the meantime, we think that presenting a bunch of demos in an easy-to-use format is potentially a major contribution. Our feeling is that a paper like this should have either (a) some really cool stuff that people can go out and use right away, or (b) some perhaps-boring stuff but with some evidence that it “works” (e.g., studies showing that students learn better when they work in groups). We think that there is room in the literature for papers like ours of type (a) and also other papers of type (b).

You might also notice that all the papers of the form, “A new proof of the central limit theorem” or whatever, never seem to have evidence of whether they are effective in class. Why? Because it seems evident that if such a new proof can increase statistical understanding, then it’s a good thing and can in some way be usefully integrated into a course. We think this is similar with the demos in our paper: they are ultimately about increasing understanding by focusing on the fact that statistics is, in reality, a participatory process with many actors. This is a deep truth which is obscured when a professor merely does blackboard material. (We have added this point in the conclusion to our article.)

. . .

Finally, the referee writes, “I think this paper needs more work so that it is not just a set of interesting stories.” Actually, I think that interesting stories (with useful directions) is not a bad thing. I wouldn’t want all the Teacher’s corner articles to be like that, but the occasional such article, if of high quality, is a contribution, I believe, in that people might actually read the article and use it to improve their teaching.

I continue to hold and express this pluralistic attitude toward research and publication.


  1. DS says:

    Psst, it looks like your RSS feed is broken. It shows the most recent post as being from last Saturday 9/26.

  2. John says:

    FYI: The blog migration seems to have broken your RSS feed. It hasn’t been updated since “An Exact Fishy Test” was published.

    • Andrew says:


      Thanks for commenting. I had the horrible feeling nobody had read this post, which makes me sad, because both the topic (classroom demonstrations) and the meta-topic (the value of stories) and the meta-meta-topic (pluralism in communication) are all important to me.

  3. Sean says:

    I thought it was interesting. I didn’t have much to add and am still trying to find some time to finish the paper. But, thanks for sharing.

  4. Keith O'Rourke says:

    The increasing availability of neat stuff worked out by others, will reduce the costs (and risks) of classroom demonstrations).

    For instance, this looks neat

    (I did something very similar using playing cards and dice but it was a bit awkward – coloured cubes that you stick together likely would be less awkward.)

    Lead in post by David Spiegelhalter at )

Leave a Reply to John