Thread: Stat 100
View Single Post
Old 07-12-2008, 12:04 PM   #55
Alex
.
 
Join Date: Feb 2005
Posts: 13,354
Alex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of coolAlex is the epitome of cool
Sorry, BTD, I suspect this is going to confuse you more but now GD and I hash it out and I don't know any other way without getting explicit.


GD: I thought about that, but the resulting difference doesn't make a huge difference until 1/100th of a percent even if pool depletion is taken into account.

299/636 x 337/635 x 298/634 = 0.1172727 = 11.72727%

Not taking pool depletion into account it is:

299/636 x 337/636 x 299/636 = 0.1171119 = 11.71119%

Using BTD two-digit rounding results in:

.47 x .53 x .47 = 0.117077 = 11.7077% (ignoring the issue of significant digits).

However, the question, as set up does not draw the three students from the pool of survey respondents so it isn't actually possible to take pool depletion into account. The survey is a 636 student subset. The three students in question are drawn from the entire pool of Penn State students and the question does not give the information on that pool size.

But that is the hard part of seeing questions out of context of the class. I don't know what level of imprecision is acceptable. So if I were doing this homework I'd probably present the second equation above. But I don't think the first one (which is what I think GD is going for) is justified. Though it would be if the three students were drawn from the population of survey respondents.
Alex is offline   Submit to Quotes Reply With Quote