Re:Use Only as Directed
by belorn on Monday September 23, 2013 @05:35AM
(#6429796)
When using statistics, it needs to be done with more care. This is a typical example where the article fails to support its main argument because of not using statistics correctly:
> The amount of acetaminophen these patients had ingested was close to the recommended daily dose of 4 grams. The median was 6 grams per day — a surprise, because the toxic dose was thought to be between 10 and 15 grams, Erush said. She also found that most of these patients had other risk factors, such as chronic alcohol use.
That is like saying: The median speed in fatal accidents are 10% over the speed limit, with most involved drivers having other risk factors such as chronic alcohol use.
Without separating out the factors, we can't say how much higher risk of an fatal accident a 10% over the speed limit might cause. The information is simply not there. I would have liked that the article told me about the increased risk that 6 grams per day might give by itself. Combining 6 grams with alcohol is a bad idea, and is as obvious as being drunk and speeding. The article really shouldn't focus on those kind of scenarios, and it weakens the main argument they are trying to make.
> The amount of acetaminophen these patients had ingested was close to the recommended daily dose of 4 grams. The median was 6 grams per day — a surprise, because the toxic dose was thought to be between 10 and 15 grams, Erush said. She also found that most of these patients had other risk factors, such as chronic alcohol use.
That is like saying: The median speed in fatal accidents are 10% over the speed limit, with most involved drivers having other risk factors such as chronic alcohol use.
Without separating out the factors, we can't say how much higher risk of an fatal accident a 10% over the speed limit might cause. The information is simply not there. I would have liked that the article told me about the increased risk that 6 grams per day might give by itself. Combining 6 grams with alcohol is a bad idea, and is as obvious as being drunk and speeding. The article really shouldn't focus on those kind of scenarios, and it weakens the main argument they are trying to make.