2.6 Inferences on Proportions
2.6.1 Estimating Proportions
The proportion of the members of the population having the trait is:
Sample proportion, is an unbiased logical point estimator for $p$:
Confidence Interval on $p$
is approximately standard-normally distributed.
Then, it follows immediately that the following is a 100(1 - $alpha$)% confidence interval for p:
But we can’t use $p$ to estimate $p$. One solution is to replace $p$ by $hat{p}$:
Sample Size for Estimating $p$
An important question:
“How large a sample should be selected so that $hat{p}$ lies within a specified distance $d$ of $p$ with a stated degree of confidence?”
Sample size for estimating $p$, prior estimate available
given $d=z_{alpha / 2} sqrt{hat{p}(1-hat{p}) / n}$
Sample size for estimating $p$, no prior estimate available
will ensure $|p-hat{p}|<d$ with 100$(1-alpha) %$ confidence.
2.6.2 Testing Hypothesis On a Proportion
Let $X_{1}, ldots, X_{n}$ be a random sample of size $n$ from a Bernoulli distribution with parameter $p$ and let $hat{p}=overline{X}$ denote the sample mean. Then any test based on the statistic
is called a large-sample test for proportion.
We reject at significance level $alpha$
2.6.3 Comparing Two Hypothesis
Point estimator for $p_1 - p_2$:
Confidence Interval on $p_1 - p_2$:
which is valid for large sample sizes.
2.6.4 Test for comparing two proportions
The test
is called a large-sample test for differences in proportions.
We reject at significance level $alpha$
Pooled Test for Equality of Proportions:
Pooled estimator for $p$ when $p_1 = p_2$:
pooled large-sample test for equality of proportions:
We reject at significance level $alpha$
近期评论