STAT 1000Q Lecture Notes - Lecture 28: Point Estimation, Standard Deviation, Sampling Distribution
STAT 1000Q verified notes
28/30View all
Document Summary
Stat 1000q lecture 28 - inference on difference of two population. Two populations are independent: population 1: true, unknown population proportion p1 of members with characteristic of interest (p 1 = x1/n1, population 2: true, unknown population proportion p2 of members with characteristic of interest (p 2 = x2/n2) = 1-cc t /2 = invnorm[(1+cc)/2] p 1 - p 2 = point estimate for p1 - p2: error term: The interval is all negative, so we are 95% con dent that p 1 - p 2 < 0 so p 1 < p 2 is 95% con dent that the proportion of. Liberal voters in ca is less than in ma. Hypothesis testing of difference of population proportions, p1 - p2: Test statistic: rejection criteria: p-value < : reject h0 p-value not < : fail to reject h0. Right tailed test: p-value = normalcdf(z, 1099) Left tailed test: p-value = normalcdf(-1099, z) Two-tailed test: p-value = 2 normalcdf(|z|, 1099)
