Chapter 9: Inferences from Two Samples Section 9.1: Two Proportions Stat 50 Introduction Scenario: Think about whether you plan to vote in the next presidential election in 2020. Let's say I split the responses into two groups: Females and Males Notice: There are TWO proportions here: p1 = P2 = QUESTION: Is there a statistically significant difference? We don't care what the proportions actually are, we care about whether they're the or not. ** In this section (9.1), we will only deal with independent samples with variables. The Logic If p1 = P2 then P1 - P2 = So we will first collect and find and Then we will see if p1 - p2 is anywhere close to Recall the Logic: Say Ho: P1 = P2 and H1: P1 p2 Ho .P1 P2 Step 2: Level of Significance If it's not given, then use . Choice depends on seriousness of making Type error. Step 3: Test Statistic 2* = (P1 - P2) -(P, - P2) Xit*2 where the point estimate is p = - pq + pq n, + 12 n2