Sleuth_04

Sleuth_04 - Alternatives to the t-Tools Stat 511 Chapter 4...

Info iconThis preview shows pages 1–10. Sign up to view the full content.

View Full Document Right Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Alternatives to the t-Tools Stat 511 Chapter 4 Case Study 1: O-Ring Failures Observational study Highly unbalanced Far from normal Apparently unequal standard deviations and larger sample is associated with smaller standard deviation Case Study 2: Cognitive Load Randomized experiment Students not randomly selected Highly skewed distributions Censored data Case Study 1: Assumptions What assumptions are violated? normality equal standard deviation Will the log transformation help? Case Study 2: Assumptions What assumptions are violated? normality data are censored Will the log transformation help? e Rank-Sum Test e rank-sum test has several names Wilcoxon test Mann-Whitney Developed in the 1940s by Frank Wilcoxon (a emist, not a statistician) A nonparametric or distribution free tool No speci f c distributional assumptions required Tests the null hypothesis of identical population distributions e Rank-Sum Test Implied assumption: populations have the same shape and spread Performs nearly as well as the two-sample t-test when the data are normal and considerably better when there are extreme outliers Resistant (because it is based on ranks) Can deal with some kinds of censored data Disadvantages Con f dence intervals di cult to compute by hand (and many statistics pa ages dont do them either) its a trial-and-error calculation (see pp. 93-94) Does not extend easily to more complex situations. e Rank-Sum Test: Calculate T 1. List all observations from both samples in increasing order (Display 4.5 page 91)....
View Full Document

Page1 / 26

Sleuth_04 - Alternatives to the t-Tools Stat 511 Chapter 4...

This preview shows document pages 1 - 10. Sign up to view the full document.

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