Tag Archives: statistics

Sex, Elevators, and Correlation

There’s a math lesson somewhere in here. Read on.

Sex

I don’t remember how old I was when I first wondered, “Where do babies come from?”, but I was young enough that people thought a good test of my arithmetic skills was to ask me, “How old are you?”, but old enough that even now, I have memories of this but no memory of how I answered. I knew then that sex and babies had something to do with each other, even if I didn’t know what sex was.

But how did I know at that age, before we had to awkwardly sit through a sex education class, that sex and babies were related? I knew it the same way all kids knew anything about such a taboo topic: a little bit of gossip, a few bits of misinformation from the “older” kids, and a whole lot of TV — and by that I mean sitcoms, the news, and commercials, nothing rated above PG-13.

I remembered something about abstinence and the life consequences of teen pregnancy and that “sex” was always central to these topics. And of course, abstinence was “not having sex” which meant “not getting pregnant”. So of course, by youth logic, “having sex” meant “getting pregnant”.

Now, dear readers, let’s for the next few paragraphs forget everything we know about human procreation and pretend we are this kid who is confronted with this topic for the first time, is too shy / scared to ask anyone about it, and is attempting to come to some logical conclusion about “Where do babies come from?”.

Here are the main “facts” that I had as a child:

  • If you had sex at a young age, then you could get pregnant at a young age. And having a child as a teenager was not good because a child was a lot of responsibility and that you would have to drop out of school. And if you dropped out of school your life was over.
  • There was something mystical about virginity and losing it; and fathers were always worried about their daughters on prom night.
  • You shouldn’t have sex before marriage because you could get pregnant and that your child would be a bastard. And being a bastard child was bad. “Bastard” was also a bad word and if you said it, you would get punished.
  • You should only have sex with the person you love.
  • A lot of people get married and then have kids. Some people have lots of kids, some people have just one kid.
  • A lot of people get married in their twenties and thirties and sometimes the woman is pregnant during the wedding.
  • A lot of people get married in their twenties and thirties and some couples “wait” to have a child and others have one within the first year or so of marriage.
  • On a TV show, the woman would excitedly announce to her boyfriend / fiance / husband, “Guess what?! I’m pregnant!” and for some reason the male in receipt of this news was almost always flabbergasted and went into a panic. And this was funny or dramatic depending on the show. Very rarely did the male react with equal joy, but when he did it was mostly because the couple had “waited so long” to have a child.
  • On a TV show or otherwise, there would be the dramatic question of “Who is the baby’s father?”. It was never, “Who is the baby’s mother?”; everyone knew who the mother was, but the father was sometimes a mystery.
  • On a TV show, the boyfriend / fiance / husband would deny that the newborn child was his and accuse some other male of being the father. But the mother would never / could never deny that the newborn was hers.
  • Sometimes adults gossiped about the 40-something-year old woman who got pregnant. And the gossip was typically of the form, “Wow, that’s really late.” or “I can’t believe she got pregnant at her age.”.

All this information is “normal” for a kid to hear. A question though is, “What wack-a-doodle conclusion does a kid come to when working off this information?”. In my case, I came to the following conclusions, which weren’t ridiculous:

  • Babies come from females.
  • A male and a female have to have sex in order for there to be a baby.

The wack-a-doodle (but “logically” sound!) conclusions that I came to were … and remember I was a kid! Probably less than ten years old! 😀

  • People have sex only once in their life.
  • Once a female has sex she can become pregnant randomly throughout her life!
  • Once a female has sex she can become pregnant a random number of times throughout her life!

Those last three conclusions explained it all!

  • Why else would the guy be surprised about his partner’s pregnancy? If sex and babies were directly related it shouldn’t have been a surprise! But of course he was surprised because he wasn’t expecting. He was probably thinking that she would get pregnant at some point, but not now. He knew he couldn’t predict it, but it was such “bad” timing for her to be pregnant now.
  • Of course some couples were excited about the pregnancy, they had “waited” for so long! That also explains why sometimes people pray for a child — that’s probably what the couple meant when they said, “They were trying to have a child.”.
  • Of course having sex at a young age is “bad”! That’s why some families are really big! Because the couple had sex at a young age and the woman got randomly pregnant several times. That’s why it was advisable to “wait” to have sex because then you have pretty good odds of not having a lot of kids.
  • It makes sense that the father could be a mystery because the woman could only have sex once and whenever that was, the guy isn’t in the picture now.
  • Of course it’s a surprise that someone in her forties got pregnant! It was random and she just randomly got pregnant “late”. That’s why the adults would comment on the age.

As ridiculous as this sounds, this is a completely consistent and plausible explanation given the information that I had as a kid for how sex and babies are related. My theory explained all the data / information that I had.

So as a broader question, do we “invent a theory to fit the data”? I think we do (see what I did there?). This is dangerous if we don’t stick to a scientific process — namely to discard the theory once it no longer fits the (new) data and to search for a new, more comprehensive theory. When it comes to the hard sciences, we tend to work in a mostly controlled and controllable environment with repeatably testable experiments. When it comes to research in other areas like, say, education, how often do we go back and test those results from 50 years ago? 30 years ago? 10 years ago? Do we treat them as mostly immutable “hard science results”? From what sample size do we extrapolate results that lead to “education reform”?

Elevators

A few years ago, I had to routinely go to a client’s office. The drill was the same. I would arrive at the office, press the buzzer, announce myself, get buzzed in, proceed to the elevator, press the button to go to the second floor, arrive at the second floor, be greeted, and settle into the office. And I did this probably a good two dozen times. Same pattern, same results with the only nuance that sometimes I’d have to hold down the second floor button before the elevator got moving, but whatever, probably a loose connection.

Normally, I would head out to grab some lunch and come back following the same routine of getting buzzed in. But one day, we decided to have a “power lunch”. My client and I went to a nearby sandwich place and we “did business over lunch”. Then we headed back. Of course, this time we didn’t have to get buzzed in, my client had the access card.

When we got in the elevator, I pressed the second floor button, but once again it seemed to not respond. So I held the button down. Strangely the elevator still didn’t respond. All the while, my client was searching through his pockets and much to my chagrin, he pulls out the elevator key, inserts it and abracadabra, the elevator begins to move.

WHAT!?!

Going Up

Well, it turns out that those two dozen times prior, my pressing the second floor button did nothing. It never did anything. Someone on the second floor called for the elevator once I got buzzed in. They would often wait about 20-30 seconds before requesting the elevator to come to the second floor because that’s about how long it would take me to get from the main door to the elevator. That’s why I had to “hold down the button” sometimes — I just got in a little sooner than the person on the second floor made the call request.

This is ridiculous and this is serious. It’s ridiculous because I thought that my action of pressing the button made the elevator go to the second floor. It’s serious because of the “indisputable” correlation in the sequence of events that had absolutely nothing to do with the causation (press button for second floor, arrive on second floor).

This also raises a plethora of questions with the biggest question in my mind as “How many things go on in our lives, in our world, in our little tiny speck of the universe, where we truly believe we are controlling a certain sequence of events when in reality there is an unknown external entity ‘allowing’ it to happen?”. Also, “How much do we simply chalk up to ‘variance in the data’ when in reality it’s another set of dimensions we simply do not understand nor consider and the model we have to explain the data is just coincidentally a ‘good fit’?”; “Do we even understand the dimensionality of the problems we are considering?”. In the elevator story, the variance was that sometimes I had to hold down the second floor button. It was unexplained behavior and I merrily chalked it up to “must be a loose connection”, which if I had paused to think about it, would have made little sense.

Sex, Elevators, and Correlation

When you saw the title of this article did you think it had something to do with some correlation study about sex in elevators? If so, what kind of social conditioning enters into our evaluation process? And how quickly can we identify that what is happening contrary to our expectation is not actually anomalous but rather a false sense of anomaly caused by a weird bias of our own culture?

If you didn’t think this article would have anything to do with sex in elevators, what in the world did you think it would be about?


New to this blog? Check out the “About The Blog” page and come say “Hello” to @shahlock