Question

In need of statistics help. The topic is Hypothesis testing for proportions.
What is the difference between a z-test and p-value?

Answer 1

@zepdrix

Answer 2

From what I'm reading, it's just the p value is the corresponding table result for the z score. yea?

Answer 3

Also, how low does the p value need to be in order to reject the null? 10%?

Answer 4

A `z-test` is a test using the standard normal (Z) distribution, which uses z-scores

If you have a left tailed test, then the `p-value` is equal to the area to the left of the z-score
If you have a right tailed test, then the p value is equal to the area to the right of the z-score
If the test is two-tailed, then you'll have to determine the area to the left of `-1*|z|` and to the right of `|z|` where `z` is the test statistic

the `p-value` helps you figure out if you reject the null Ho or not.

Rule:
If the p-value is less than alpha, you reject Ho. Otherwise, you fail to reject Ho

---------------------------------------------------------

In contrast, there is another way to do a hypothesis test and it's usually referred to the `classical approach` (or method). This is where you determine the critical values and that helps set up the rejection region

Answer 5

Sorry I guess I should simplify the two-tailed test rule

Simply find the area to the left of `-1*|z|` OR the area to the right of `|z|`. Whatever the result is, double it

Answer 6

this page has a good example

http://mips.stanford.edu/courses/stats_data_analsys/lesson_10/classical-P.html

Answer 7

so what if I don't have an alpha

Answer 8

if alpha isn't explicitly stated, then it's usually set to 0.05

this is just convention really.

Answer 9

I mean, we aren't doing level of confidence for this yet

Answer 10

this discussion confirms my last claim

http://www.analystforum.com/forums/cfa-forums/cfa-level-ii-forum/91308716

I'm sure your text also mentions it somewhere.

Answer 11

Yes, but not this chapter. I am unfortunately student teaching, and am forced to teach this even though I'm not familiar with it. This chapter is for hypothesis testing proportions.

Answer 12

So, do we just say so long as p is below .05 reject it?

Answer 13

or should I do .1? I have to give my students some general answer.

Answer 14

https://books.google.com/books?id=L0GiugxnVPwC&pg=PA115&dq=statistics+default+alpha+value&hl=en&sa=X&ved=0ahUKEwjaloWP9ZHLAhUH2yYKHXTvA9IQ6AEIHDAA#v=onepage&q=statistics%20default%20alpha%20value&f=false

this book says

`The business default level of significance is 5% (alpha = 0.05)`

scratch out the word `business` and the statement would still be valid since it applies to any branch of stats (not just business stats)

Answer 15

I'd go with 0.05

Answer 16

Alright, thank you. So if 1-ztest =p value, if p<.05 reject

Answer 17

unless your book explicitly says somewhere in it like "if alpha isn't stated, then use 0.1" to override what I posted

Answer 18

no, this book doesn't even say when to reject at all.

Answer 19

yes if `pvalue < 0.05` then reject Ho

this is assuming alpha isn't given

Answer 20

ok, and the 1 proportion z-test gives me 1-pvalue right?

Answer 21

I don't know what you mean by that

Answer 22

So the one-proportion z-test. That's how I find the p-value right?

Answer 23

yes you can use the one-proportion z-test to compute the z-score. Once you have the z-score, you would then use the Z distribution to find the p value

Answer 24

ok, that's what I was checking. Thank you jim

Answer 25

All of the tests involve you computing a standard T or Z score, which is then used to look up in a table of printed values to find the p value

Answer 26

no problem

Answer 27

I really hope I don't have to teach stuff I don't know again