I taught a ~50 minute session covering the R package ggplot2 at the Software Carpentry boot camp for the Data Science for the Social Good program. I passed out note cards, and at the end of the session had the participants write down one thing I did well and one thing that I need to improve. I had a high response rate; 24 of ~30 participants provided feedback.
How I taught the lesson:
I wrote an R script that walked exactly through the lesson. For the presentation, I had the R script open in RStudio (and so did the students), and we walked through it together. I had exercises interspersed throughout the lesson, which they worked on in pairs. Some of the exercises were simply to adjust some of the parameters, but others required them to search through the online documentation themselves to find the right command. Lastly, I also made a concept map for ggplot2. I handed out paper copies of the concept map (they also had the digital copy), and referred to it throughout the lesson by bringing it up on my screen.
What I did well:
- The exercises. The students really enjoyed working together and putting their knowledge to the test immediately after I introduced a specific topic. I have used exercises interspersed throughout the lesson in the past, and this was great confirmation that this is a good strategy.
- The organized and well-documented script. Many students felt that the script helped them follow the structure of the lecture and that it would be a good resource in the future.
What I did not do well:
- My public speaking voice. I mumble when I speak casually. Also, I do not have a booming voice.
- Too fast paced. Many students felt like the lesson was far too fast to absorb all the material. This was somewhat out of my control because the time allotted for my lesson was shortened since previous sections went over. However, it is important for me to remember that even straightforward concepts will take a little time to absorb since it is the first time students are being exposed to them, and that I should always go slower if possible. One good suggestion one of the students gave me was to incorporate more live coding into my presentation since that would have forced me to go slower. In the future, I want to try a mix of documented and live coding to see how that goes.
- Ask leading questions. Only one student pointed this out, but I found it to be extremely insightful. When I ask a question that I do not think should be difficult and don’t get a quick response, I get flustered since I do not want to spend much time on a simple “are you following” type question. S/he suggested that I ask leading questions since it was not always obvious to them the exact answer I was looking for.
Was the concept map useful?
The concept map was the main new teaching tool that I implemented in this lesson after having learned about them in the instructor training. I found that creating the concept map was an extremely beneficial method for organizing my thoughts and planning my lesson. When I read Hadley Wickham’s book on ggplot2, I was often confused at first with all the vocabulary since it was unclear how it all fit together. Creating the concept map forced me to solidify my understanding. Thus, as an effective way to plan a lesson, I think the concept map was definitely useful.
However, the student response to the concept map was a mixed bag. I wanted the students to have a bird’s eye view of the all the components so that would make more sense as we walked through each component in a linear fashion through the lesson. Five students mentioned the concept map in their feedback:
Great conceptual graphic.
Awesome graphic page…
Concept map is confusing!
Concept map-not sure how that fits in.
It is too quick to really understand the separate characteristics on the concept map.
So I think part of the trouble with the concept map had to do with the limited time of the lesson and the high-level nature of the concept map. ggplot2 takes some time to get used to, so I guess I shouldn’t be so surprised that the students felt frustrated that they could not fully comprehend the concept map.
Take-away
I need to talk slower and louder. Also, I need to do a better job gauging how much information I can fit into one lesson, i.e. cover fewer nodes on the concept map.
And if anyone has time to check out my R script and/or concept map, I would appreciate any feedback.