## Monday, March 20, 2017

### hypothesis test for proportions, things to remember

Here's a few things to remember

Z scores for proportion
(p^ - p)/square root(p*(1-p)/n)

Z for difference of two proportion
(p1^ - p2^)/square root(p-bar(1-p-bar)/n1 + (p-bar(1-p-bar)/n2))
p-bar = (x1 +x2)/(n1 + n2)

Note that p-bar might also be noted at p-pooled

For hypotheses, remember that Ho always contains = and Ha contains <, > or "does not equal"
Confidence intervals for single proportion

p^ +/- Z*square root(p^(1-p^)/n)

For difference of two propotions
p1^ - p2^ +/- Z*square root(p1^(1-p1^)/n1 + p2^(1-p2^)/n2)

Z values for confidence intervals
90% = 1.645
95% = 1.96
98% = 2.33
99% = 2.575

You can also get these from Z chart

P-values are the value from the Z chart for corresponding Z score if Ha contains < and 1- value from the Z chart for corresponding Z score if Ha contains >. If Ha is "does not equal" you take 1 - value from the Z chart for corresponding Z score then multiply the result by 2.
You can also get p-values from Z scores using the link below.

http://www.socscistatistics.com/pvalues/normaldistribution.aspx