Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
Difference in Proportion
Message posted by Paula (via 205.188.193.188) on July 8, 2001 at 5:26 PM (ET)
Please Bail Me OUT!
Determine whether the difference in the proportion of people in each group who identify themselves as having alcohol abuse problems is statistically significant.( at .05)
PO DEN
SA N=30 N=24
#w f=6 f=8
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: Difference in Proportion
Message posted by JG (via 128.8.23.198) on July 9, 2001 at 12:24 AM (ET)
Z=1.06 and p-value=.145
Re: Difference in Proportion
Message posted by Paul (via 208.238.117.10) on July 13, 2001 at 5:35 PM (ET)
Your p value is greater than your alpha setting of .05. Thus, accept the null hypothesis...there is no statistical difference.
Re: Difference in Proportion
Message posted by Jack Tomsky (via 208.249.113.130) on July 13, 2001 at 6:58 PM (ET)
The exact test for the equality of two binomial proportions is Fisher's test, which is based on the hypergeometric distribution. (The only inexact aspect to this test is due to the discreteness of the observed counts.)Since your hypothesis is expressed as a two-sided hypothesis, the p-value is given by
p = 2*min[Prob(X<=8), Prob(X>=8)]
=2*min[0.9225, 0.2121]
=0.4242,
where the probabilities are the hypergeometric probabilities of the second sample.
Since p>0.05, the null hypothesis is accepted.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.