A discussion of our first experiment
Today was a good day for statistics and methodology questions, but not a lot of very pretty or exciting things – sorry in advance that the blog will be light on pictures (I’ll try to make up for it with the most punny titles I can think of). But we did talk a lot about R, statistics, experimental design, and more, so I’ll highlight the big points below.
The Mock Dataset
Over the weekend, Louie and Vincent had an extensive discussion about how to best design our first experiment. The result of this discussion is a handy script & R Markdown use a set of mocked-up data, to try and see which statistical tests would work best with the data and how exactly we’d want to collect our real data. It doesn’t go over everything, but it does explain some of the decisions they made and the code they chose, so it’s definitely worth checking out! You can find the markdown posted in the r-analysis channel on Slack, and the full script in the Drive, under Monarch Stripes > Landing experiments.
Choice or No-Choice: A Hard Choice to Make
To start us off this class, Vincent and Louie presented their mock dataset and gave their impressions of the data and analysis process. We discussed the pros and cons of the Choice and No-Choice experiments; namely, a no-choice experiment would need a lot more Tachinids to make a decent sample size, and we can say a lot with a choice experiment (especially if we’re low on flies). This raised another big concern: will we have enough flies? Louie is confident we’ll be able to breed plenty, so I’m content trusting him on that. Ideally, if we have enough Tachinids, we would do a choice and a no-choice experiment, but if we have to budget our flies, the class agreed that a choice experiment seemed the best way to go.
How Our Methods Earned Their Stripes
I’m reaching for whatever puns or idioms I can get, folks.
The next big topic of discussion was what pattern to compare to our striped pattern: black, white, or grey. Last time we’d discussed it, Louie had proposed doing our first choice experiment with a striped card and a black card, since the Tachinids are most likely attracted to black, and so we’d be most likely to see some kind of difference with that setup. But as we discussed today, if we only have enough flies to conduct one choice experiment (our worst-case scenario), we have a bit more room to say something meaningful if we compare a striped card to a grey card (since that lets us control for brightness). With that consensus reached, we moved on to another point of discussion…
The Life and Times of our Tachinid Flies
…How we intend on gathering our data from these experiments. Vincent initially proposed a fairly easy and straightforward method: watch the flies, and record their location every 3 minutes. While that’s definitely one of the easiest ways to do it, it does have its weaknesses, as it would be easy for these 3-minute slices to not represent the fly’s actual behavior. Louie proposed a more comprehensive version of this, essentially recording every time the fly changed its location, so we could know where it was at every second in a comprehensive timeline. We discussed the merits of this strategy a bit (being able to say where the fly is at any given time, and how much it moves around). We also discussed a more streamlined variation that loses some of that data: using two stopwatches and simply recording how much time the fly spends on each of the patterns. It seemed like the class was still undecided on which of these last two to do, so we can figure that out on Friday. Either way, though, we’ll need multiple stopwatches, so Eliza suggested using an online stopwatch app linked here (and also on Slack): http://stopwatch.online-timers.com/multiple-stopwatches
Odds and Ends
There were a few other topics discussed today that we never quite reached a resolution on. For example, everyone seems to have a slightly different idea of our actual experimental apparatus: is the fly tethered, or in a cage/box? How long is the tether/how large the box? How are we dealing with time spent flying/on walls? Clearly we still have a few details to iron out. But we’ll have a chance to lock in those details…
Alright, I gave up on the punny titles at this point, it’s hard to make a fly pun based on Friday. “Flyday”, if you will. The point is, we’ve got a plan for Friday!
-From ~1-3: Fly collection! Our biggest roadblock right now is making sure we have Tachinids. Since it’s supposed to be warm and sunny on Friday, we’ll spend the first two hours trying to fix that problem. The plan is to break into three groups – two large groups going to the student farms and the arboretum, and one smaller group checking out the Village Homes community gardens near Figgy Park. If we’re lucky, we’ll catch enough to start a breeding population, and then we’ll face the problem of too many flies, and figuring out what to do with them. But, one step at a time!
-From ~3-5: A practice run of our experiment! We’ll spend the first chunk of this time answering those questions we couldn’t today, and writing down solid, reproducible protocols. Then, if we have time, we can use some of the Tachinids we already have to do some preliminary trial runs of our experiment and work out the kinks.
The Fun Stuff
Louie has also extended the doodle poll, so be sure to chime in on your availability for the class potluck! Hopefully everyone can make it if we hold it a bit later on.
Thanks for making it through the wall of text all the way to the end! Your reward is a picture of Jess’s adorable rat, Itsy, who joined us in class today (and reminded me of how wonderful vertebrates are, after spending all of this time trying to catch flies and raise caterpillars).
Anywho, that’s all from me. See you on Friday, folks!