Saturday, May 25, 2013

A Dose of Skepticism (Can be a Coping Skill).

I debated whether to split this up into two parts but in the end decided against it, so this will be a lengthy post.  I hope everyone enjoys it enough to stick around!

Many of you, like me, will first come across scientific research findings in popular media.  I really don't like the hype that studies can get simply because they are about controversial, socially relevant, or otherwise important issues.  Then there's the way that researchers/media will, for publicity's sake, draw spectacular conclusions based on insubstantial evidence or using biased arguments.  Then there's how people take those conclusions and get ever more angry with one another as logical fallacies and peudoscience* supply the front for a perfect storm of anger, blame, smugness, egomania and all-around emotional chaos.
Pictured: Everyone on the internet.

Here are some more conspicuous problems that I would like to point out:

Correlation does not imply causation.

Two things seem related, so one has to be caused by the other, doesn't it?  Well... no.  One might cause the other, but some other variable might actually be the cause of either or both.  It's why it is so difficult to say things like "video games cause real-life violence."  They might, or already violent people might gravitate toward certain video games.  Or they might be one factor among many.  Or they might not be a factor at all, but so many people play video games that it's unlikely to find a truly violent person who doesn't.  I'm not saying any of those are correct, but they're all possibilities.  Be critical if you see an article that doesn't at least account for those possibilities.

In certain circumstances, confusing correlation with causation can go beyond conveying misleading information and create harmful perceptions.   Think about it this way: a headline reads "People Under 6'5" More Intelligent."  The article says that on a certain graduate-level math exam, researchers found that exam takers over 6'5" didn't do as well as those under 6'5".  We can conclude that it must be all that extra height, can't we? It was researched, so it must have at least a grain of truth to it, right? ... And now a stereotype exists that taller people are unintelligent.  It was so easy to leave out or downplay the part where those over 6'5" who participated in the study were more likely to be on a professional sports team and have left college early.

The study also concluded that Einstein would be more likely 
to do well on the test than the general public.

Quotable does not mean unquestionable.

People love to quote others, especially if the quote comes from someone famous or influential.  It doesn't matter if the people they quote are/were actually kind or good or right or making any sense, because what they really love is quoting others who agree with them.  Yes, many quotes are hilarious or interesting or thought-provoking and deserve recognition, and some state facts or are almost universally acknowledged to be true, like a man in possession of great fortune being in want of a wife. However, your favorite public figure saying something doesn't make it true.

And according to the internet, these two men said everything.

Having a higher degree doesn't make someone right all the time.  Having done extensive research in a certain field doesn't bestow All The Knowledge.  Being well-known does not grant someone wisdom.  I am half convinced that if Benjamin Franklin had said, "Beer is proof that God loves us and wants us to be happy while drinking under the bright noonday olive green sky," people would be looking at you funny and calmly quoting Benjamin Franklin as you pointed to the sky screaming, "It's blue!  Blue, I tell you!"

I have an opinion, and studies will prove that I'm right.

Quoting only people who agree with an opinion is part of a larger issue known as the confirmation bias.    This occurs when researchers look for and use only data that supports their hypotheses.  Let's say that Jamie believes that dark-haired short women with small waists are most attractive.  Jamie googles "short dark hair small waist perfect" and finds tons of studies saying... you guessed it... that a short, dark-haired woman with a small waist is the "ideal" woman.  Jamie cites those studies and then gives friends and family a Y/N questionnaire that asks, "Do you think dark-haired women are attractive?" "Do you think women with small waists are attractive?" "Do you think shorter women are attractive?" Jamie sees that over half of participants answer Y, and suddenly we have:

(Now I'm finally free to put down women who don't meet my standards, because science.)

... Only it's not good science.  Looking up preferred characteristics in conjunction with the word "perfect" skewed results in favor of Jamie's opinion from the start.  Then the questionnaire won't allow people to say whether they find other qualities more personally appealing.  Jamie also gave the questionnaire to a group of people who are more likely to hold similar opinions.  These are only a few of the problems.  While most research isn't this blatantly biased, the confirmation bias certainly exists.  That it's usually well-disguised only makes it more problematic.

Statistics and numbers can be manipulated.

Hark, a quote!  The one I referenced briefly in the post Intolerance Isn't Awesome, "There are three kinds of lies:  lies, d***ed lies, and statistics," was popularized by Mark Twain.  I'm going to show why I think this quote is relevant, even if it is perhaps hyperbolic.  First, I don't believe all statistics are really lies.  "Statistically significant" in research is important.  But it needs to be taken in context, because numbers can be used like this:

"I don't eat vegetables, because statistically many American women who eat vegetables aren't expected to live past age 80."
Would those be the same women who statistically aren't expected to live past 80 anyway?

Apple Juice: Made with 100% real fruit juice.  

It might be the juice of freshly picked apples with and nothing else, or it might be "100% real" juice from three different fruits and added sugar.

People who drink lemonade 12%** more likely to die in summer months.
People who drink lemonade might be outside more during summer and nearer to water sources, where they would naturally be more likely to drown than those who stay inside.  Death by Lemonade would be a misnomer unless the deaths were due to citrus allergies or choking on seeds.

Lemons don't kill people.  
Kindergartners trying to make an extra $5 on weekends in July kill people.

Remember that it matters who does the research and who pays for it as well.  A large painting company probably won't publish research that says fumes from their paint have caused a significant number of illnesses.

They'll spin you right 'round, baby. 

There are news channels and publications with obvious political leanings and agendas, and they can put "spin" on stories like nobody's business. Sometimes they're even forthright about it.  Unfortunately, most of the time reporters don't come out and say, "We are trying to get you to think in a particular way about this," but do just that.  If you ever see facts "floating" in articles that seem jarring in the flow of the writing, or if you don't see citations, or if you notice that a strangely high number of comments agree with what seem like extreme conclusions, proceed with caution.  In the media, not giving all the facts or giving some facts out of context isn't the same as lying... it's journalism. I don't see this kind of insidious reporting disappearing any time soon, so I believe that people need to be able to recognize it.

We <3 straw men and ad hominem attacks.

"How can you talk if you haven't got a brain?"
"Some people without brains do an awful lot of talking, don't they?"
-The Wizard of Oz***

Mary says "I think we should raise the legal BAC to 0.09."

Dave can disagree with that statement if he wants.  He can talk about what a BAC of 0.09 does to a person physiologically.  He can discuss the probability of accidents occurring with a BAC of 0.09 or below.  What he does instead is say:

Wrong. Having no consequences for driving impaired will make people do it more often, and then it the roads will naturally become more dangerous.

This is a straw man argument.  It makes the other person's opinion open to or easier to attack by distorting it.  It builds up a straw man- a sort of scapegoat opinion- in order to burn it down.  Mary didn't say that there shouldn't be consequences for impaired driving or that the road wouldn't become more dangerous without those consequences.  All Mary said was that she thinks there ought to be a change in the legal definition of "drunk."  Dave is more interested in getting others on his side.  If Mary points this out, though, there is always the possibility of this occurring:

Why should people care what some stupid ugly troll thinks?

Ouch.  ... Wait.  Aside from "attractive" being subjective.... how are Mary's looks relevant?  And does using the label "troll" make her a troll?  (A: They aren't, and no).  This is an example of the logical fallacy argumentum ad hominem (lit. "to the man argument").  It means avoiding the issue at hand and going straight for personal insults and character attacks to discredit someone.  Politics is infamous for these "smear campaigns."  Let's not talk about how to fix a problem or why a particular plan won't work.   Let's focus on the fact that the other guy is wealthy, or poor, or black, or white, or wears glasses, or is a loud breather, or has one foot larger than the other.  And then let's make it seem like that matters.  Instead of appealing to intelligence, logic, or common sense, we'll darn well appeal to peoples' prejudices.  It's easier and still effective.****


So when looking at new information, question what you read.  Do the legwork to see if there is more information out there.  Look at citations.  See whether research has been replicated; repetition seems boring, but it lends a lot more credibility.  Don't assume that numbers don't lie.  And if you're easily lured in by headlines or can't look away from the virtual train wrecks that are Public Comment Sections, keep the above in mind.  They've helped me stay afloat in a sea of professional manipulators, and I hope they can help you.

*Note: If you read Intolerance Isn't Awesome (link in post), you will realize that by "pseudoscience" I'm NOT referencing religion in any way, shape or form.  I do not equate fake science and faith.
**Yes, I pulled this number out of my magic hat.  I was looking for a rabbit, but I suppose you take what you can get.
***Help, I can't stop the quotes!  Depending on context, this could actually be an ad hominem attack by Scarecrow if he used that line in a debate...
****This is a call to everyone who uses ad hominem attacks or goes along with ones the media perpetuates (let's be honest; we all do it at times): Argumentum ad hominem is the lazy way out.  It is an indication that you don't have a better argument.  If you feel superior afterward, all it means is that you feel good about putting others down, not that you really are superior.  Think about what that says about smear tactics.  They- I'm going to use "they" for anyone who uses smear tactics- get you to fall for inferior arguments by appealing to that little part of your brain that says, "Of course I'm better than he is/she is/they are. That's a given."  They believe that you aren't savvy enough to think any further than that.  And if you go along with them, you're going along with those who either don't want to or can't come up with a good argument.  Do they sound like people you can trust?  Respect?  Admire?  

No comments:

Post a Comment