So are we learning anything this term, apart from how awesome Costa Rica is? The answer is yes, believe it or not. Biology and natural history is obviously a big part of what we’re doing, but lots of aha-moments also happen when we do stats.
In Monteverde we studied phytotelmata, which are little pools of water that form between the leaves of the bromeliad plant. These tiny little pools, as small as a few ml, host whole ecosystems. Who knew.
Some phytotelmata have more species of tiny animals than others, and we wanted to figure out what determines how many species can live in these. We guessed larger pools could host more species, and that turned out to be true, but more species are also to be found in larger plants, plants nearer to other plants, etc, etc, etc. This is where powerful stats come in: through a multivariate regression analysis we showed that all of these other factors actually aren’t relevant – they are just things that correlate with higher volume. For example, larger pools are to be found in larger plants. No-brainer.
There are lots of things in the world that correlate with each other. That doesn’t mean they’re relevant. Think, for example, of chocolate (I do, a lot). Does chocolate make you rich? Well, if you look at the stats, some of the countries in the world with the highest chocolate consumption (Switzerland, Sweden, USA), are also the richest. Or why not read my favorite xkcd?