statistics own me.

[alphatarget1]

wow, i don't know what I'm doing wrong, this sucks.

the proportion of adults living in a small town who are college graduates is estimated to be p = 0.6. to tes this hypothesis, a random sample of 200 adults is selected. the acceptance region is defined to be 110<x<130 where x is the number of college raudates in our sample. use the normal approximation.

evalute alpha (probability of a type I error).

well ok, i figured out the x bar (mean) to be =np which is (200*0.6) = 120 and my sigma to be (200*0.6*0.4)^(1/2) = 6.92 something. did a normal cumulative distribution from a really small number to 110, with the mean = 120 and st.dev = 6.92. i didn't get what the book got, must've done something wrong. any ideas?

 

[alphatarget1]

my head is about to explode.

 

[Mrfrog840]

what kind of math?

edit: stat math?

 

[Cawchy87]

First year stats it looks like. If this thread is around tomorrow I'll help.... too sleepy from Calculus tho.

 

[moomoo40moo]

Sorry, cant help.... http://forums.anandtech.com/messageview...atid=38&threadid=1755828&enterthread=y

 

[RGUN]

So is this hypothesis testing or do they just want to know the confidence interval?

So I get Z = 1.4434 and I forget what that makes alpha... something like 14.88%

 

[alphatarget1]

Originally posted by: RGUN
So is this hypothesis testing or do they just want to know the confidence interval?

So I get Z = 1.4434 and I forget what that makes alpha... something like 14.88%

hypothesis testing, but if i find the right Z value i should get the right answer.

i got the same Z too

 

[Skiguy411]

If I remembered how to do this stuff from last year I'd help, but alas I have forgotten.

 

[Instan00dles]

I suck at math

 

[RGUN]

If this is a hypothesis testing problem then dont you have two unknowns the way you word the question? 1 being whether or not the hypothesis is true and 2 being the testing interval?

When the test interval is ambiguous we have always been told to use alpha = 0.05, also, have you considered how being a one tailed or two tailed test might affect your values?

 

[Perknose]

statistics own me

In Russia . . .

 

[alphatarget1]

Originally posted by: RGUN
If this is a hypothesis testing problem then dont you have two unknowns the way you word the question? 1 being whether or not the hypothesis is true and 2 being the testing interval?

When the test interval is ambiguous we have always been told to use alpha = 0.05, also, have you considered how being a one tailed or two tailed test might affect your values?

we're trying to determine alpha...

 

[Hyudra]

i'm sure you have to go with one of the statistical tests to find alpha.

 

[Hyudra]

oh ya, you check if the p-value > or less then alpha... or maybe not..

anyways i would help, but I sold my book already.

 

[zerocool1]

wouldn't it be p-'alpha', on a two-tailed test?

 

[RGUN]

Speaking of getting owned by statistics... I kid you not this was a question from my statistics exam last week:

***** University orders 12 power supplies, 4 power supplies are defective, what are the chances of 2 or more defective power supplies in the sample.

That was it... there was nothing else, I told the prof 4 times we didnt have enough information to solve the problem but she kept saying I wasnt reading the question properly..... she doesnt speak english that well... and she writes like she speaks.