p-value

#1
Can someone please help me understand what this means in plain english?

The p-value = 2 x [1-phi(8.8)] < .001, and the results are highly significant.

Am I multiplying 8.8 times phi? Do I get phi from a table?
 
#2
Seems to me, you're working with a two-tailed hypothesis test. phi is the cumulative density function of whatever distribution you're working with. 8.8 is the test statistic.

It isn't "8.8 times phi"
 
Last edited:

hlsmith

Omega Contributor
#3
kdsan,

It would help if you posted the source or a link to the original content.




Buckeye,


Have you read Fermat's Enigma. Its a great book describing the problem and Andrew Wiles' saga to answer it.
 
#4
I haven't had the chance to read it. But, that quote has been a running joke in a few of my classes when we don't know how to answer a question. I find it funny in that way.
 

hlsmith

Omega Contributor
#6
I wasn't too familiar with using the Z-test for comparing proportions. So I looked it up, you find the p-value for the test statistic 8.8 on the standard normal table and multiply it by 2 if you were running a two-tailed test (i.e., Ho: p1 = p2; Ha: p1 not = p2, no directionality on difference).


So you look 8.8 on the standard normal table and minus it from 1 since it in on the left tail. You get 2(1 - 0.99997) for the pvalue.