April 7, 2012

Significance

Dear Guy in my Statistics Class, 


Okay, so this is slightly weird to be writing a public letter to you, but here goes: I like you. I think you're very attractive and, from what I've gathered in class, quite intelligent. Yet, I don't even know you. Frankly, I'm not even sure what your name is. But more importantly, I don't even know if you're attracted to men (which, I suppose is the important thing for me). Honestly, I'd just like to get to know you better; you seem like a really nice person, but given our zero previous interactions, I feel that trying to start up a conversation with you this late in the semester would be very awkward. I don't even know if you know who I am, but if you do, say hi to me, so that socially-awkward me doesn't have to and then further embarrass myself. 


My letter presents the following set of hypotheses.

H0: He is not attracted to men.
Ha: He is attracted to men.

In statistics, we can either accept the null hypothesis or reject it in favor of the alternative hypothesis. The objective is, based on statistically significant evidence, we would be able to arrive at the correct conclusion. However, it is very possible to come to a false positive or a false negative conclusion. In the case above, I risk making a type I error, which is the case where I reject the null (and conclude that he is attracted to men) when the reality is that the null is true (he isn't attracted to men). So let's say I were to act on this. The end result is pretty clear: I would just make a fool of myself. I would hope that this would be the last interaction I have with this guy, and then I would try and get over the trauma of what I did. I can also make a type II error, which is the situation where I fail to reject the null (and say that he is not attracted to men) when the reality is that the null is false (so he is attracted to men). So in this case, I would just go about my daily life and nothing would happen on my end, when there potentially could have been something, but I was too afraid and too cautious to start the conversation. I don't know how people discuss sexuality from a "tabula rasa" standpoint of first meeting a guy, but I am curious, if people want to provide insight for me, please leave a comment. My "gaydar" is and has been dysfunctional for about two years, so I can't use that, and even if I did have good gaydar, I don't know how I would then approach a total stranger in this current context. However, given no current methods to test these hypotheses without building up a lot of courage and engineering a conversation, I resort to extrapolation.

I've been faced with this situation many times before, except for the fact that I've known the people that I'm attracted to, and I've known that they are gay (okay, so it's the exact opposite?). And given this, I've never had the courage to ask somebody out. In terms of subtle flirting and "signs", I'm pretty good at that, but it's never a conscious choice and I never then ask the guy out. I often fear asking guys out because of the underlying context of dating and hypersexualization of gay relationships, but that's another post. I also just lack the self-confidence I need to ask somebody out. I don't know where everybody is getting their confidence from, but I would like some leftovers, if possible. So given these associations I have with dating and asking people out (and lack thereof), I correlate this to the situation above, and risk making a type II error, which seems like a better option to me than making a type I error.

Using prior evidence in order to motivate my future directions on relationships is perhaps not the healthiest option, but using a Bayesian framework, this seems to be the most natural approach. However, this is what I've been doing. I don't seem to understand how people can just make friends/relationships with strangers so quickly; perhaps it's just my introversion, but even in a natural setting for me such as the classroom, I've often met my best friends. I don't want to be shy Cameron who never meets other people, but I feel that is the only option for me to maintain my sanity.

And then there is the matter of what makes a result significant. What outcome would I want for the efforts of my experiment or trial to be worth the cost I put into it? If that is a relationship or just meeting a new person and making a new friend, then perhaps the ends justify the means. But how likely will this be the case?

2 comments:

  1. Love it and I feel ya. I meet a girl but it's too late in the semester to go past "hi" and "bye" and I'm too scared to anyways. You're not the only one...thanks for putting what so many of us awkward kids go through into words!

    ReplyDelete
  2. WELL PLAYED. Don't forget that your test is stronger when you give yourself more degrees of freedom. ^__~

    ReplyDelete