Programming in R Workshop Post-mortem

30 January 2018

The workshop outline can be found here

I advertised this workshop as appropriate for anyone to attend, even (and especially) if they hadn’t done any programming before – I wanted this to be useful to people who had only ever used functions in Excel, say. My goal here was to get people familiar with the very idea of typing in commands to an editor/REPL, and running them to get some result. I also wanted learners to see RStudio in action, get a feel for what it’s like to drive it.

This session was intended as gentle introduction, to warm people to the very foundational ideas of “code”, syntax, and data. And, because we are moving at breakneck speed (the next session is on using dplyr in R for data wrangling, this session was important for everyone who wanted to come to future sessions to attend. For that reason, I took a poll on Facebook and selected the 2 times that covered the greatest number of learners, and ran the workshop twice.

Now, in addition to newbies, I also know there are several learners with at least some programming experience, so in the first session I ran, I organized it as a “choose your own adventure”: either start with some very very basic Intro R exercises from DataCamp if you are a beginner, or if you have some experience start with Software/Data Carpentry lessons depending on your learning style. I provided an opinionated guide to each of these options.

Bad idea. Everyone chose to start with the DataCamp exercises, and I suspect that’s because it’s the first thing on the of options and the other options seem equally good places to start – that is, I probably didn’t explain the options very well (or have good reasons). In any case, everyone dutifully hammered away at the DataCamp exercises and by the 1.5h mark it seemed had exhausted their supply of dopamine, and weren’t able to concentrate on the next parts of the workshop plan. The good thing about doing the DataCamp intro R lesson exercises first is that they are short, self-contained, and give you immediate feedback when you solve them. And that’s also their downside: it’s monotonous, without much context or application, and can be too much of a good thing.

The second time I ran the workshop I told learners directly to start with the DataCamp exercises, but only stay on them for at most 30 minutes before moving to the Software Carpentry lesson. That way, learners got a feel the editor/console/REPL process but had enough concentration left to explore RStudio and sink their teeth into some promising exercises.

So, takeaways:

Be clear about what learners should do, and limit the drill exercises. Perhaps in other circumstances (a longer workshop? more experienced learners) the Choose Your Own Adventure technique could work, but in a short, 2 hour, workshop with beginners it fell flat.
Stress having a break. I learned this after the first session, and have continued to make a point telling people when an hour is up, and giving them permission to take a break.
2 hours is not really enough time to do this well, but I knew that. I tried to make it very clear, and will continue to do so, that these workshops, and this entire series is only meant show learners what is possible and give them a chance to play: not a place to become an expert.
Running two sessions of the workshop provided a great opportunity to rethink and rework my lesson plan, and it wasn’t a heck of a lot more work to do. Some people came out for both sessions also.