Chapter 24 ISCAM Chapter 5 Intro

COMPARING SEVERAL POPULATIONS, EXPLORING RELATIONSHIPS

The idea of comparing two groups has been a recurring theme throughout this course. In the previous chapters, you have been limited to exploring two groups at a time. You saw that often the same analysis techniques apply whether the data have been collected as independent random samples or from a randomized experiment, although this data collection distinction strongly influences the scope of conclusions that you can draw from the study. You will see a similar pattern in this chapter as you extend your analyses to exploring two or more groups. In particular, you will study a procedure for comparing a categorical response variable across several groups and a procedure for comparing a quantitative response variable across several groups. You will also study the important notion of association between variables, first with categorical variables and then for studies in which both variables are quantitative. In this latter case, you will also learn a new set of numerical and graphical summaries for describing these relationships.

Section 1: Two Categorical Variables

Investigation 5.1: Dr. Spock’s Trial - Chi-squared test for homogeneity of proportions

Investigation 5.1A: Newspaper Credibility Decline - Comparing distributions

Investigation 5.2: A moral tale - Randomized experiment

Investigation 5.3: Nightlights and Near sightedness (cont.) - Chi-squared test for association

Section 5.1 Summary

Section 2: Comparing Several Population Means

Investigation 5.4: Disability discrimination - Reasoning of ANOVA

Investigation 5.5: Restaurant spending and music - ANOVA practice

Applet Exploration: Exploring ANOVA

Section 5.2 Summary

Section 3: Two Quantitative Variables