Sybarite: People aren't very good at making decisions that involve risk. Many people are afraid of airplanes, although accidents are extremely rare

I've always thought a lot of the fear of flying had to do with degrees of control. Sure, there are a ton of factors while driving that you can't do anything about, but for the most part you feel in charge of your own fate. Flying on the other hand removes you from the equation entirely, forcing you to rely on a whole slew of people you don't know to do their jobs right and keep you from plowing into the ground. Granted you surrender control on something like a bus too, but if the bus driver is incompetent or the maintenance crew didn't do their jobs right your life still probably isn't going to end in a ball of flame.

This is pretty much how I feel about flying as well. I do fly, but not that often. I know my chances of dying on a plane are way less than the chance I take behind the wheel every day. But as you said, I "feel" like I at least have some control over the situation when I'm driving (even though I don't have control over other people on the road). And don't even get me started on how I feel when I'm sitting in the passenger seat and my wife is driving. Let's just say that I'd rather be on a plane.

My 4th grade teacher used to say, "Practice makes permanent."

People are bad at probability math in general, we just don't seem to be hardwired to come to the correct answers even when given the data we need. There are a lot of reasons for this. Others are correct when they note that people are willing to accept more risk in situations they themselves control (iirc, it's about an order of magnitude more risk they will tolerate). Another reason people are so bad is that we have problems dealing with small numbers or very infrequent events. People can't really understand what it means if I give them an actual risk, telling someone the chance of dying in an airplane is 0.0000001% really just tells them that it is "unlikely" and will not "feel" any different that if I told them the risk of death was 0.000001% or 0.^A% (although these three risks are all 1 full order of magnitude, or a factor of 10, off from each other). Even if I try to use numbers that are visually more understandable (saying "once in ten thousand years" or use scientific notation), these numbers are hard to really conceptualize properly.

Finally, risk is the combination of both the likelihood of an event and the consequences of that event. People can very easily over emphasize one half of that equation. An easy example would be modern nuclear power: the consequences of a plant failure can be enormous, but the likelihood given current designs and safety requirements make this probability vanishingly small (but never zero, you can engineer to minimize risk, but never to eliminate it...another concept people seem to have problems with). People look at the potential consequences (radiation/fallout in the event of a meltdown) and immediately will fight any plans to put such a plant into service near them. Even though the risk may be small, they have completely ignored the likelihood of an accident occurring and focused only on the consequences. This often seems to be more prevalent when talking to people about sudden catastrophic events.

/Does risk analysis for a living
//getting a kick and all that jazz

The test they describe does not seem to be valid. They only offer you a bigger reward for choosing a riskier option. I don't see where a person has an actual loss at risk. Why not say fark it, I'll take a chance for a big win versus a little win if the loss is zero either way.

Oliver Twisted: The test they describe does not seem to be valid. They only offer you a bigger reward for choosing a riskier option. I don't see where a person has an actual loss at risk. Why not say fark it, I'll take a chance for a big win versus a little win if the loss is zero either way.

I agree. the researcher should also consider that waiting minutes to die horribly, falling from thousands of feet is more emotionally wrenching that dying instantly in a car wreck.