以下是引用philipfang在2006-9-10 18:35:00的发言:Suppose there are 100,000 non-doctor parents and 10% of them (10,000) have doctor children. Suppose again there are 10,000 doctor paretns and 15% (2000) have doctor children. Then both Choi and Hart's statements are true. This line of thinking can be dangerous. Suppose there is a similar question Instead of two people aguing. It's one person's argument goes like this "Clearly, all other factors being equal, children whose parents earned doctorates are more likely to earn a doctorate than children whose parents did not earn doctorates. This is because over 70 percent of all doctorate holders have a parent that also holds a doctorate." or vice versa "Recent study shows, all other factors being equal, children whose parents earned doctorates are more likely to earn a doctorate than children whose parents did not earn doctorates. Therefore, this study explains why over 70 percent of all doctorate holders in my university have a parent that also holds a doctorate." Now what's the flaw of this logic? Or harder, it can ask you to evaluate whether it is valid logic and why. Again the key here is not come up with a number, rather you have to see the statement about likely hood of a effect to a subject with a variable on and off is irrelevant to the statement about the sheer percentage of that effect in a population. Once one graps this. LSAT percent and number question will become easy. There are questions like 80% of percent airplane crash had one common thing assoicated with it, therefore that thing should be avoided to remdey potential crash. More premature babies are born to women under 40, therefore to the contray of common belief, woman under 40 are more like to have premature baby than over 40. on and on, same thing
[此贴子已经被作者于2006-9-13 11:17:15编辑过] |