Does Correlation Always Mean Causation?
Here are some statistical findings of Dr. Donald A. Redelmeier:
- Talking on a cellphone while driving is as dangerous as driving while intoxicated.
- About 25 more people die in crashes on presidential Election Days in the United States than the norm, which he attributes to increased traffic, rushed drivers and unfamiliar routes.
- He also discovered a 41 percent relative increase in fatalities on Super Bowl Sunday, which he attributed to a combination of fatigue, distraction and alcohol.
- Changing lanes doesn’t get you there sooner, but it increases the chances of collision about threefold.
- When physicians make a medical decision for a single hypothetical patient, they favor more expensive treatments than when making a decision for a group of hypothetical patients with similar symptoms.
- Medical school class presidents died an average 2.5 years earlier than others in the same classes.
Cell phones are as dangerous as driving drunk? Hard to believe.
The lane changing result could have been predicted using efficient market theory. There is usually not a free lunch on the highway any more than there is a free lunch in the stock market.
Eating while driving is even worth than talking on a cell phone while driving. Texting is worse yet. For that matter, taking the old sedating antihistamines like Benadryl is about as bad as driving at the limit of drinking and driving.
Some of this is testing without theory, and therefore, meaningless. A lot of things are correlated, but not causal. The sun rises after the cock crows, but not because of it.