OR: But Perhaps not the One You Think
America is living in the middle of an information age. As a market analyst I see numbers thrown around from the moment I get up until the moment I go to bed (please call before 2 AM). We try to make sense of the numbers by using statistics to plot graphs and charts so that we can "wade" through the numbers without becoming "numb" and often this is a very useful tool for making sense of the overload. However, we often try to apply reasoning to numbers that is not valid and we now are in the midst of an election season replete with numbers and inferences that do not hold up to scrutiny.
Recently there was a blog that discussed terrorism and it's implications for the Muslim faith. When I got to read it there were any number of replies so the discussion was well underway. But the sense I made of the article was that Muslims are involved in terrorism (statistically) and that it "might" be partially to do with the religion itself (conclusion).
This article is not meant to be a reply to that blog (and there are many other articles that could be pointed out) but there were several fundamental problems with the writer's analysis. First, statistics on terrorism prior to 1990 are essentially worthless because the definition used to define terrorism has changed over time. So, for example, if white supremicists bomb a black church in Georgia in 1965 and they were Baptists it might not have been counted at that time, might be now (with our heightened awareness to terrorism).
Second of all, using statistics drawn by inference violates several principles of statistics. For example, Douglas Adams in Hitchhikers Guide to the Galaxy notes that there are 0 people (as in none) in the galaxy. Statistically that would be correct and the argument is as follows (paraphrasing):
There are perhaps 20,000 planets in the galaxy.
There is no life on any of them other than our own (that we know about).
Therefore since this is the only planet with life, that is the statistical anomaly.
So the overwhelming statistical analysis says that life cannot exist here either.
So, there is no life on earth. (Sorry for any inconvenience). Please pick up a boarding pass at the nearest
non-existent space facility.
What this means is that when we take evidence and try to apply it to populations by inference it fails the scrutiny of the method in the first place. The statistics suggest terrorism increased over time. That's as far as it can be taken.
When you take statistics and expand the reasoning a new phenomenom occurs. This is the basis for bias and prejudice. Most people won't question the statistics (that's why we have them in the first place) and so we naturally accept the conclusions as real. But in statistics you assign causes based on a thorough examination of the issues involved. Terrorism, for example, occurs because people are (in no particular order):
1) Poor
2) Uneducated
3) Educated to believe in something more important than preserving life and property
4) Told to do it by others (including parents)
5) Hopeless
6) Having fun (the old revenge is sweet issue)
7) Looking for respect amongst their peers
8) Angry with society, your parents, or bosses (especially at the post office)
9-1000) Hundreds more to choose from!!
And yes.....religion. The Ku Klux Klan, the John Birch Society, David Koresh and the Branch Davidians, Tim McVeigh and Terry Nichols all had some ties with radical religious ideas.
BUT.....
That is not a reflection of the faith portrayed by the religion but rather the interpretation of faith made by a few (statistically possible in ALL religions) of the followers.
What you end up with by inference is a prejudice and profiling of those who "look like" they might be a particular faith. We end up with a disillusionment there is a "plot", a conspiracy and that we are at risk because those d*** ____________s are out to destroy all we hold dear!
Now, does that mean there aren't any terrorists? No, I never said that. You see, statistics applies to groups but not necessarily to individuals. For example, if you did a statistical study of traffic on a particular highway (those annoying speed bands they put across the road) the police can track how fast people were going and when. This means they can put a patrol car there at a particular time to catch the speeders. However, this doesn't mean if you pass by that spot when the police are there you will be speeding.
We could go into the causes of that (because assigning causes is what statistics does, in part) but we will forgo your serious lack of attention to traffic patterns when everyone else is going 75 and you are going 60 (62.5 in Canada). The point is that a person is not a group and using inference to think this guy (or girl) "looks" like anything is wrong.
So what can you, the reader, do to make sense of statistics (in the season of....)? First, consider the source and the way the statistics are derived. Second, consider the causes of the results before making judgments about them. Third, don't interpolate (draw conclusions) that weren't within the scope of the statistics in the first place. This is where the danger lies for most of us, to assume that statistics "point" to something and then act as if they justify our actions.
Now about that "life on earth" anomaly.