This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Chapter 19: Testing for Differences among Three or More Groups Have learned to test hypotheses about two population means Sometimes want to compare 3 or more means Could use t tests to compare each pair of means But, increases likelihood of Type I Error Better way is to use Analysis of Variance (ANOVA) Uses many previous concepts Variance (Ch. 5) SS (Ch. 5) s 2 (unbiased estimate of σ 2 ; Ch. 12) df in variance (Ch. 12) Homogeneity of variance (Ch. 14) Type I Error (Ch. 13) Normally distributed sampling distributions (Ch. 14) OneWay ANOVA In oneway ANOVA, we have one IV with two or more conditions (e.g., “happy” , “sad” and “neutral” condition) k is number of conditions (levels) Null and Alternative Hypotheses ANOVA is an omnibus test (testing many different situations at once) H : μ A = μ B = μ C = … = μ k (all groups are equal) H A : At least two μs differ Note that we don’t specify which groups differ Because testing multiple groups, all H A ’s in ANOVA are nondirectional For Independent Groups Take 3 independent random samples and take the mean of each Even if population means are equal, there will be some variation in observed means across groups How much variation would be expected by chance alone? Within and Between Group Variation Partition (separate) variance of all scores into 2...
View
Full Document
 Spring '08
 Arthur
 Normal Distribution, Variance

Click to edit the document details