OUTLINE FOR CHAPTER
10 IN THE 6^{th} EDITION
Inferences Concerning One or
Two Means
 Introduction
 A needed shift in inferential focus: from correlations to means
 Reasons why multiple chapters are needed to consider inferences
on means
 Inferences Concerning a Single Mean
 The inferential purpose
 Interval Estimation
 Tests concerning a null hypothesis
 The null hypothesis
 The two most popular test procedures: ttests and ztests
 Degrees of freedom (df )
 Inferences Concerning Two Means
 Independent vs. correlated samples
 The basic distinction
 Three settings that produce correlated samples
 The sample sizes, if different, provide a signal as to the
kind of samples
 Terminology
 The inferential purpose
 Setting up and testing a null hypothesis
 The general and typical forms of the tested Ho
 Three settings that produce correlated samples
 The three most popular test procedures: t, z, and F
 Interval estimation with two means
 Multiple Dependent Variables
 Results presented within passages of text or in tables
 The Bonferroni adjustment technique
 Effect Size Assessment and Power Analyses
 Do impressive plevels signify important findings?
 Options for checking to see if a statistically significant result
has practical significance too
 Various ways to estimate effect size
 Criteria for "small," "medium," and "large" estimated effect sizes
 Performing a power analysis before data are collected
 Underlying Assumptions
 The four main assumptions . . . and the two discussed most by
applied researchers
 Testing the assumptions (and options if an assumption seems untenable)
 Equal sample sizes and the notion of "robustness"
 Comments
 A nonsignificant result doesnąt mean that the null hypothesis
is true
 Overlapping distributions
 The typical use of ttests
 Practical significance versus statistical significance
 Type I and Type II errors
