Testing the Null Hypotheses

This week I took my final exams for the graduate courses I have been taking at a nearby university. For the past 11 weeks I have been deeply immersed in Psychopathology, Advanced Human Development and Probability and Statistics. My brain is SOOOO ready to be done with all that serious reading for a while. Right now all I want to do is pick up some fluff brain candy novel and go soak in a bubble bath.

In the classes that I teach, at the end of every term I talk about the concept of "Intellectual Bulimia"-- referring to the all too common practice of students cramming all the information they can into their brains and then regurgitating it back out for their instructors on final exams. With all the elegance of momma birds yacking up a meal into the beaks of their waiting offspring, these students dump their load in research papers and presentation portfolios, never to think on any of it again. I certainly saw some of that going on with my peers.

For me, however, there were several key concepts that were learned that will be kept and savored. One of those is the whole idea of “testing the Null Hypotheses”. When evaluating data, many times in comparing various averages or trends of numbers, it may APPEAR that there is a particular relationship between different variables. Yet when running the appropriate statistic (whether it be Z scores or T tests or Chi Squares or Anovas) it is possible to show if there is no significant relationship within a certain confidence level. I like that. It’s a way of saying “Liar, Liar Pants on Fire!” to sensational looking numbers that would lead us down the garden path of wrong assumptions.

I’ve heard it said that statistics can be sneaky and misleading. They are sort of like a bikini. What they reveal may be interesting, but what they cover up can be crucial.

So I liked learning more about statistics, how to read them, how to interpret them, etc so that I can assess information that I am exposed to with more critical thought and understanding than before. I had taken stats when I got my master’s degree in Sociology many moons ago. That time I was pretty much jumping through hoops for a grade. This time I was able to apply myself more and study it for real to get a firmer grasp on what it all meant. For a person who used to be downright phobic of anything involving math, I consider it a great achievement that I got as comfortable with the formulas as I did.

I just wish there were some clear statistics for knowing when things matter or don’t matter in LIFE.

For example: This afternoon on my lunch hour I ran to the store to pick up some pictures I had developed and get something to eat. When I came out, I discovered a note on my windshield saying “The wind blew my door open and hit your car. Call me if there are any damages.” Huh? So I got out to take a look. Sure enough, there is a rather noticeable dent in the right side panel. This is not some little pit door ding. This is a bona fide DENT. Ouch.

Now, here is my problem. Part of me thinks, “so what?” After all, it makes no difference whatsoever in the drivability of the vehicle. Caring about that dent feels awfully superficial. I cringe at the idea of spending a chunk of resources on getting it fixed, even if they are not my resources. There are so many better ways for resources to be spent in this world. It feels vain and shallow to have them directed to repairing a cosmetic problem on my car. So what if it is someone else who may have to pay. Is it even worth worrying about it??

Except I sort of do care. It’s my car and I don’t want it to be damaged. After many years of driving around in old beaters, I’m finally driving a car that I like and I try to take very good care of it. It’s not a FANCY car. But it is barely over a year old and is a good looking car. Now it is a good looking car with a dent. Also there is the whole consideration of re-sale value should I ever decide to move on to something else. So I’m pretty sure I’ll take it in and see about getting it fixed.

Yet I’ve been second guessing myself all afternoon about whether I’m ok with my own values on this… playing devil’s advocate till the cows come home.

In talking to other people here at work about it I’ve gotten the whole gambit of opinions, ranging from ABSOLUTELY get it fixed to who cares, it’s not all that bad. I don’t know what I think. I wish there was a simple statistic I could run that would tell me if this matters. Sadly, life is not that simple.