Wednesday, May 4, 2011

Causations, Links, and the Study of Mind

I have read previously how we have to be more careful about causations in our research. If A appears with B, it could be that (1)A caused B; (2)B caused A; (3)A and B are both caused by something else. Recently I experienced a real-life version of the warning. The story goes like this:

I was cooking in my kitchen. It was dinner time. I stpped on a small region of my wooden floors and felt that it sounded funny-it felt like there were something under it and it was puffy. An hour later my neighbor downstairs came knock on my door and reported she had water dripping from the kitchen ceiling. I went to see it-the region of her ceiling was all wet and there were indeed water dripping.

I went to the building manager and he told me to shut down the water in my house first to avoid more leaking--if assume there was leaking from under my wooden floor, maybe a pipe broke or something. I did shut down the water-by turning an old and dingy wheel. But then 10 mins later when I tried to turn it back on, there was no water. Not a drop in the entire house.

So let's sum up what's happened so far: (a)water leaking under my floor-->(b)I shut down the water supply-->(c)when I turn it back on there is no water anywhere(given that all other units have water running in their houses in my building). What could it be?It is intriguing to think of (a) to be the cause of (c) or of some sort of link, because when there was no (a) there was no (c) and the water was running smoothly. It also makes sense to think so because it would be kind of too easy to think that it could be something with the switch--that it happened to be the case that when I turned that wheel and shut off the switch for the water supply to my apt (something I would never do in a year since I've lived here), the switch happened to stop working so when I turn it on I don't get any water anymore. It seems very unlikly that when nobody turn that switch for years at all it's ok, but when I turn it this once--and when (a) happens that it actually is broken. Well, the result is, it turned out, that this is exactly the case--the switch was broken so when I turn it back on it was still stuck there so no water came(stuck from inside and looked perfectly alright on the outsied). As simple is it were, it was hard to believe at first.

And it was clear that event (a) and (c), although appeared in a series of event, were not related at all, in any fashion of (1)-(3)(in which case (3) is referring to single instead of plural), but actually are two completely independent things caused by two independent factors. Of course statistically speaking when you have large samples this may not happen. But it could also be the case where if we assume that the old and dingy switch at a lot of homes were not usually used for years, and then when an accident somewhere happened in the house, they turn-and the switch broke. In that case you may as well link (a) to (c) in a causation link (if, think in the study of minds), while in reality they are indeed two independent events caused by two independent factors-in this case, a broken pipe and a broken switch.

So, what have we learned from this?It seems like we should indeed be more careful while draw conclusions and generalizations about our experiments.

No comments:

Powered By Blogger