![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
And now for a little bit of math...
Sep. 11th, 2007 12:54 amLet's talk about risk. In most of the decisions we make, there is a certain amount risk involved in making the wrong choice. Some decisions have such negative possible risk-associated outcome that we don't make that decision at all, regardless of the actual frequency with which the bad outcome occurs.
A few years ago in a class I was delivering in Australia, one of my students was about to travel to India. I asked him whether he planned to bring food with him. The question was, should he bring food to eat while he was there, or should he brave the local fare? So, let's say he braves the local food, and it was the wrong decision; the food poisoning could vary from mildly uncomfortable, through requiring a hospital and not being able to work, all the way to life threatening. The possibility of food poisoning is what we call "beta" risk. And what about if he brought his own food, ate his own food, and it was the wrong decision; it would be rude to the locals, and he could miss out on some really great chow. This is what we call "alpha" risk. The obvious way to eliminate the possibility of beta risk entirely was to make himself susceptible to alpha risk. ...After class he went out and bought food. I think I scared him badly.
Another common example of risk is the decision made by fire alarms on ceilings all over the world -- should it signal an alarm, or not? If it signals an alarm and it is the wrong decision, this would be alpha risk; firemen would come, people would be angry, there would be much hoo ha. So let's say it does not signal an alarm. This is, of course, it's usual state. And let's say failing to signal the alarm is the wrong decision; so a fire happens and no one is alerted until they encounter the heat and flames for themselves, possibly people die. The beta risk here is potentially deadly, while the alpha risk is merely annoying. The way to eliminate beta risk? The alarm should go off all the time. Obviously, this is not practical, so the designers of fire alarms must balance their small machine between these two risks.
Doctors make decisions with your health all the time. If they decide you are sick when you really are not, this alpha risk could lead to additional painful tests, humiliating speeches to your family, and possibly spending a fair amount of money. If they decide you are not sick when you really are... The health consequences depend on the severity of the illness in question. I heard a few years back that the tests for HIV had a known low beta risk and very high alpha risk. We call this "often has false positives." Doctors deliberately increased the alpha risk so that the beta risk would be lower.
Our government is getting in the act of choosing which risk to minimize. Previously, when someone suspected of criminal activity was taken into custody, they were "innocent until proven guilty." This attitude had a low beta risk (chance of incarcerating an innocent) and a higher alpha risk (chance of letting a guilty person go). As a nation, we felt that this was the way it should be. However, our current administration seems to feel strongly that driving alpha risk down, down, down is the goal. Not surprisingly, this means that the beta risk has shot drastically up.
A few years ago in a class I was delivering in Australia, one of my students was about to travel to India. I asked him whether he planned to bring food with him. The question was, should he bring food to eat while he was there, or should he brave the local fare? So, let's say he braves the local food, and it was the wrong decision; the food poisoning could vary from mildly uncomfortable, through requiring a hospital and not being able to work, all the way to life threatening. The possibility of food poisoning is what we call "beta" risk. And what about if he brought his own food, ate his own food, and it was the wrong decision; it would be rude to the locals, and he could miss out on some really great chow. This is what we call "alpha" risk. The obvious way to eliminate the possibility of beta risk entirely was to make himself susceptible to alpha risk. ...After class he went out and bought food. I think I scared him badly.
Another common example of risk is the decision made by fire alarms on ceilings all over the world -- should it signal an alarm, or not? If it signals an alarm and it is the wrong decision, this would be alpha risk; firemen would come, people would be angry, there would be much hoo ha. So let's say it does not signal an alarm. This is, of course, it's usual state. And let's say failing to signal the alarm is the wrong decision; so a fire happens and no one is alerted until they encounter the heat and flames for themselves, possibly people die. The beta risk here is potentially deadly, while the alpha risk is merely annoying. The way to eliminate beta risk? The alarm should go off all the time. Obviously, this is not practical, so the designers of fire alarms must balance their small machine between these two risks.
Doctors make decisions with your health all the time. If they decide you are sick when you really are not, this alpha risk could lead to additional painful tests, humiliating speeches to your family, and possibly spending a fair amount of money. If they decide you are not sick when you really are... The health consequences depend on the severity of the illness in question. I heard a few years back that the tests for HIV had a known low beta risk and very high alpha risk. We call this "often has false positives." Doctors deliberately increased the alpha risk so that the beta risk would be lower.
Our government is getting in the act of choosing which risk to minimize. Previously, when someone suspected of criminal activity was taken into custody, they were "innocent until proven guilty." This attitude had a low beta risk (chance of incarcerating an innocent) and a higher alpha risk (chance of letting a guilty person go). As a nation, we felt that this was the way it should be. However, our current administration seems to feel strongly that driving alpha risk down, down, down is the goal. Not surprisingly, this means that the beta risk has shot drastically up.
