CNN reports that almost 10% of drivers would fail a sample state licensing exam created by GMAC Insurance. While my wife claims that I’m bitter about getter a lower score than she did I am mostly disappointed at the poor excuse for “reporting” on the part of CNN. Instead of doing some critical analysis of the test, CNN regurgitates the statistical manipulation as provided by whoever fed them the story. Taking the test immediately reveals a number of fallacies.
The article claims that, based on a 70% passing score, 1 in 11 drivers would fail the simulated driving test. While many states use similar ratios, written tests are primarily used for permits. Such tests are also designed for people who have just studied a state specific driver’s manual. CNN also fails to mention questions that may have state specific answers and is not critical of questions that are poorly worded. Here are a few that may be easier to answer if one were to study an accompanying manual (spoilers):
“1. Where should you park when you need help after your tire suddenly deflates while driving on a highway?”
Great question, but the “correct” answer (”b. Off the pavement”) doesn’t apply to many interstate highways that don’t have an “off the pavement” because there is a guardrail. A better “all-purpose” answer might be “c. Where your car will be visible for 200 feet from the front.” If I one were really being critical one might point out that the answer does not qualify whether he disable vehicle should be visible 200 feet from the front of the car or from the front of a car approaching from behind. In addition to not specifying whether the highway has a shoulder (which is on the pavement and the correct place to park a disable vehicle), the question also does not specify whether the highway is divided, in which case the correct place to park is on the left shoulder if the car is in the leftmost lane when the tire suddenly deflates.
“8. Turn your front wheels toward the curb when you are parked _____.”
If you take this test in San Francisco, you would (correctly) be torn between “a. Facing uphill” and “c. Facing downhill,” since this is required in both cases.
“10. If you have trouble seeing other vehicles because of dust, precipitation, or smoke blowing across the roadway, you should drive slower and turn on your_______:”
My wife and I disputed our answers to this question (she agreed with the test writers). During the day it is safer to use fog lights or parking lights during obstructed visibility. The reason is that water particles and sand in the air reflect your headlights, in effect reducing visibility. There is nothing wrong with using headlights during precipitation. At night you have your headlights on in any case. My wife chose headlights because she said that running lights and parking lights may not be the same thing. Right or wrong, it is a poorly worded question that depends on the specifics of the situation and the features available on the vehicle that you are driving.
“11. If your vehicle starts to hydroplane, you should:”
This is another question that depends on the vehicle. The “correct” answer is “c. Slow down gradually by easing the gas, and not applying the brakes” but this is not true if you can effectively pump the breaks or have anti-lock breaks which can provide some increased friction.
“12. When a car with bright headlights comes toward you at night, you should:”
The answers to this question again had my wife and me in confused consensus (and again she guessed right).
“a. Move toward the right edge of your lane
b. Look above the oncoming headlights
c. Look below the oncoming headlights
d. Look toward the right edge of your lane”
We both ruled out b and c as ineffective but while my wife decided that choice a. could lead to you driving off the road, I reasoned that in choice d. it would be dangerous to take your eyes off of the road.
The simulated test might be a wonderful bit of trivia but relatively useless for anything else. With a spread of 3 questions (15% of 20 questions) between “smartest” and “dumbest” it is meaningless to try and draw any practical conclusions. The questions were not validated (there is no proven correlation between a slightly lower score on this test and worse driving). Answers to some of the questions are ambiguous, depending on unspecified conditions. Finally, these types of tests are designed for beginner drivers who have studied an accompanying manual from which the questions (in particular the wordings) are derived.
What might be a practical method of checking driving skills across the country? The way every state actually does its licensing (not permitting): an actual driving test administered by trained instructors.