Chap13-Inference-about-Comparing-Two-Populations

Chap13-Inference-about-Comparing-Two-Populations - 1 Chap...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
1 CHAPTER 13: INFERENCE ABOUT TWO POPULATIONS Stats & Prob. for Bus. Mgmt (Stat1100) Jochem Chap 13: Comparing 2 Populations b Goals of Chapter 13 Be able to do inference* about… b 1. the difference between 2 populations means s Case1: σ and σ are unknown; it is assumed that σ = σ 2 1 2 1 2 s Case2: σ 1 and σ 2 are unknown; it is assumed that σ 1 σ 2 s Both for independent samples and matched pairs . b 2. the ratio of 2 population variances b 3. the difference between 2 population proportions * i.e. computing an interval estimate and testing a hypothesis Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b Recall from Chap 9 the sampling distribution of the difference between 2 means (slides 49-53): 3 s The difference between two RVs is … s normal if the two populations are normal. s approximately normal if the populations are non-normal and the sample sizes are large. Standard Error of the dif- ference between two means Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b And thus we have that. .. s the z-test statistic is: 4 Problem: Wee need opulation variances! s the interval estimator is: population variances! Solution: If population variances unknown, then (as in Chap 12) we replace population variances ( σ 1 and σ 2 ) with sample variances (s 1 and s 2 ) and use t-values instead of z- values. (forthcoming)
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b To use these formula we have to assume that we know the population standard deviations σ 1 and σ 2 . e remainder of this chapter assumes that the 2 5 b The remainder of this chapter assumes that the 2 populations are:* s either normally distributed, or s non-normal but we have a large sample size. * If this is not satisfied, we can use a nonparametric test – the “Wilcoxon rank sum test”. We won’t cover this method in this course, but you find it in Chapter 19 of your book. Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b When we replace σ 1 2 with s 1 2 and σ 2 2 with s 2 2 we can have 2 cases: s Case 1: We can assume that σ 1 = σ 2 6 s Case 2: We can assume that σ 1 ≠ σ 2 b For each case, the book gives us a test-statistic and an estimator. s (Note: we get more accurate estimates when σ 1 = σ 2 .) Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b Test Statistic for μ 1 - μ 2 when σ 1 2 = σ 2 2 : s The t-test statistic is: 7 s where is the “pooled variance estimator” , which is the weighted average of the two sample variances (with the degrees of freedom of both samples used as the weights): This formula is called the “equal-variances test statistic.” Chap 13: Comparing 2 Populations b 1A. Difference b/w 2 means (indep. samples) b Confidence Interval Estimator for μ 1 - μ 2 when σ 1 2 = σ 2 2 : s And subsequently: 8 This formula is called the “equal-variances confidence interval estimator.” Example
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 03/26/2012 for the course STAT 1100 taught by Professor Chiappetta during the Spring '08 term at Pittsburgh.

Page1 / 16

Chap13-Inference-about-Comparing-Two-Populations - 1 Chap...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online