OK, so I admit I occasionally wander around and fill random tests on the web as well. It seems a lot of people do, so what is the harm in admitting it?
This is not a rhetorical question. One might do a bit of social calculus: does filling this test say something about me? What will my colleagues think of me for admitting it? Will I go up or down on their esteem? By how many points? Will they forget quickly or will they remember this and tease me forever?
You get the drift. I’ve already made a major blunder and started a blog, so admitting stuff like that will not hurt me more. That is my reasoning. Of course, if you read in between the lines you are probably tabulating my anxiety and updating all the information that you thought you knew about me. This is what people do. All the time. Our brains do that. It is part of the game we play to survive, and since we benefit from social activities, these types of calculations and their outcome determine how we actually behave. In the end, they also determine how successful of an individual we become.
But each of us has a different calculator. So how much does it say about me, and how much does it say about you when you judge me (either way) based on this information?
In some sense this is a post about neurology and brain function. But not a single neuron has been dissected, nor a single brain function has been explained. This is just a teaser to make you think something different for a few seconds. Maybe you can make a whole afternoon worth of chatting based on this input.
As far as I can tell our brains are a reservoir of statistics and we perform a Bayesian inference calculation each time we think. We have a model, and we update it based on the new information. I will postpone a dissection of Bayesian probability for some other day, seeing as I didn’t bring my scalpels today. I inserted that stupid joke there on purpose to irritate some of you, so for those who are irritated go ahead and groan loudly (I’m sure you are reacting right now). I’d say about half of the people who read that line will groan internally and only one in a gazillion will groan out loud, except that I skewed the statistics by telling you, so I can’t believe it anymore either. That’s my guess.
The whole point of Bayesian inference is that we have a previous model (our bias) before we conclude something, and many questions about the a-priori chances of seeing something in a laboratory or concluding something depend on the models that one is using to describe them. So when people say there is a 60 percent chance of seeing supersymmetry at the LHC you are seeing their bias. The truth is we don’t know and the only true statement one can make is this: either we will see it or not, but we can not assign an objective probability to it. Fortunately subjective probabilities can be assigned, and this is what we do when we make those statements. When we give those probabilities, we are indicating how we would bet money on the outcomes of whatever question you might be posing. If you have a lot of information, you can place bets very cleverly. For example, I would bet all the money in the world to assert that the sun will rise tomorrow. I don’t think I could find anyone with any sense that would take that bet and if they did, I would not trust them. So in the end I wouldn’t end up making the bet. Not for that much money anyhow.
DISCLAIMER: For this post to be effective, it needs a fall guy. Seeing as it is impolite to do that to anyone else, I have to choose myself for the job. So you got it right: It’s all about me.