Posted: 4/27/2005 8:39:05 AM EDT
|
We need to analyse some statistical data as part of a lab final. Since there wasn't anything in our book on this (because it's not a stats class, and up to this point everything was just done on MS Excel) but we were given a few practice problems to test our ability to do it before going into the final lab. Here's the problem: Use the given degree of confidence and sample data to construct a confidence interval of the population proportion p. n = 176 x = 112 95 % A) .515 < p <.627 B) .502 < p < .640 C) .516 < p < .626 D) .501 < p < .641 You don't have to solve it for me, I just need to know how to go about doing so. Thanks a lot. ETA: I know how to do this with p and q, I think... or when the mean is known, but what is x? It's not /X (mean) right? |
.
I don't have the lab yet. These are just practice problems so that we can practice what we need to do for the lab. It must be a standard normal distribution, but it doesn't say. |
No, he specified in the original problem that 'p' is the population proportion |
Assuming a normal distribution, 95% CI means Z = 1.96 So, I reasoned that the confidence interval equals 112 ------ +/- Z*Stderr(proportion) 176 where the standard error of a porportion equals sqrt( (p*(1-p)) / n ) but since we don't know p (population proportion) we use the sample proportion (112/176) (same as in the above equation which is basically p(samp) +/- Z*stderr(prop) |