This Flo and Friends comic strip by Jenny Campbell offers an excellent example how switching the context of a statistical generalization can mislead people into accepting a false conclusion as factual.
Mr. O's statement is completely accurate within the context of human mortality statistics. Very few people do, in fact, die between 100 and 101 because very few people live to that age.
But Mr. O has switched contexts. Instead of a statistical generalization based on supporting data, he has used a subterfuge based on semantics.
In essence, Mr. O's words tell his young friend that "very few people die between 100 and 101" means that any individual who reaches that age is unlikely to die in that time span. He has done this as a joke, perhaps even as a means of getting his young friend to think about what he has just heard.
It is a classic example of linguistic bait and switch, and in this instance it is funny.
Such is not always the case.
Statistics present generalizations derived from examining data samples collected from large numbers of the members of specific groups. The larger the sample, the more reliable are predictions made about what individuals are likely to do when responding to issues important to the groups to which they belong.
Properly used, statistics allow us to move from the general to the specific and to know that doing so is a relatively safe bet because the generalization rests upon a sufficiently large, carefully collected and controlled data sample.
Statistics can also be misused to move from the specific to the general in order to condemn all of the members of a group based on the the actions of one or a few members of that group. This is a sucker's bet. It relies on the fact that humans will ignore logic and rational thought if they can be convinced that they are in danger.
This type of bait and switch is one of the favorite deceptions used by con men, ideologues, and even presidents who tweet alternative facts at three o'clock in the morning.
No comments:
Post a Comment