Comments on the Presentations

Since we did not have time for in-class debriefing, I thought I'd pass along some of my general thoughts on the lightning presentations.

Why this assignment?

As I explained when I first talked about the projects, I think it's important for you to be able to talk about statistics. I hope that having to summarize information for your colleagues and present it in a comprehensible form was a useful exercise. (I also think that most Grinnell students can use a bit more practice in having to talk to a group of people. For example, some of you still need to work on making eye contact and speaking loudly enough.)

Since I very much care that your poster projects go well, a second purpose of this project was to give you experience with the data set. Most of you seem to have spent some time looking at the various variables. As importantly, most of you seem to have learned a bit more R so that you could analyze your particular variables. I hope that learning will be helpful as you work on the analysis of data for your poster.

Finally, I expected that you would learn a bit by watching each other present.

Although you were generally dealing with similar kinds of data, you found very different ways of approaching the data. So, one thing you may have learned were other ways to think about similar data. (It struck me that, at least early on, people were making notes about what groups were saying and thinking about how they could fit those things into their own presentations.)

One interesting thing (at least to me) is how different groups dealt with non-binary categorical data. A number of groups ended up taking the non-binary data and combining values to make it binary. I think most of you will have noticed that different groups did better and worse jobs of describing the original data and the technique they used for combining values. The groups also presented the binary variables in different ways: Some starting with the "raw" data and then refining into binary variables, others just starting with the binary form, still exploring how different ways of dividing the values lead to different conclusions. I hope you'll use those kinds of things as examples of better/clearer ways to do similar activities.

I note that some groups expressed relative risk in very precise terms (e.g., the relative risk is 2.3124) while others expressed relative risk in very general terms (e.g., men were about twice as likely as women to ...).

I also hope that you learned the value of including clear textual information in your presentations. For example, while many of your titles were amusing, in the end, a number of comments came back to the wording of the question in the original survey. (You may want to look at that instrument again in preparation for your posters.)

I would encourage you to think a bit about the following questions:

• Which presentations did I find most successful?
• Why did I find those presentations successful?
• Which presentations did I find least successful?
• Why did I find those presentations less successful?
• If my group could do the presentation again (using the same variable), what would we change?

Some groups made some interesting statistical mistakes. I'll include notes on those mistakes in my evaluations of the projects, so that those who made them don't make them again. For a modicum of extra credit, you can bring to class a list of the statistical mistakes you observed. (Yes, you may discuss this question with your colleagues.)

I do plan to give a similar exercise the next time I teach 115. I envision a few changes:

• It will be earlier in the semester, almost immediately after the descriptive statistics section. (I had trouble getting the data earlier this semester.)
• There will be additional instructions on using R to build an appropriate bar graph. (I'll provide those instructions in the next few days in case you need them.)
• I'll bring a timer to class, and stop people after three minutes, whether or not they are done.
• I'll use more levels in the rubric, since there were clearly different levels of quality.

I think a number of you learned that it's a bad idea to try to put things together the night before the presentation is due. Please keep that in mind as you work on your posters.

I would also appreciate hearing if you have any suggestions or thoughts on what went well or poorly on this activity.