week13 - Lecture 32 Nancy Pfenning Stats 1000 Chapter 16:...

Info iconThis preview shows pages 1–3. 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
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Lecture 32 Nancy Pfenning Stats 1000 Chapter 16: Analysis of Variance Example Suppose your instructor administers 3 different forms of a final exam. When scores are posted, you see the observed mean scores for those 3 different forms82, 66, and 60are not the same. Is this due to chance variation, or do the 3 exams not share the same level of difficulty? Version 1: 65, 73, 78, 79, 86, 93, 100 Version 2: 39, 58, 63, 67, 69, 74, 92 Version 3: 39, 52, 62, 64, 66, 77 [Note: if there were only 2 different exams, a two-sample t test could be used.] It would be unrealistic to expect 3 identical sample mean scores, even if the exams were all equally difficult: in other words, theres bound to be some variation among means in the 3 groups. Also, of course there will be variation of scores within each group. If the ratio of variation among groups to variation within groups is large enough, we will have evidence that the population mean scores for the 3 groups actually differ. Picture two possible configurations for the data, both of which represent 3 data sets with means 82, 66, and 60. Thus, variation among means (from the overall mean of 70) would be the same for both configurations. 1. There could be large within-group variation, such as we see in Exams 1a, 2a, and 3a, in which case the ratio among to within is small; in this case, the data could be coming from populations that share the same mean. 2. There could be small within-group variation, such as we see in Exams 1b, 2b, and 3b, in which case the ratio among to within is large; in this case, the population means probably differ. 141 Note that sample size, although not yet mentioned, plays a role, too. If there were 100 students in each group, we would set more store by the differences than if there were only 2 students in each group....
View Full Document

This note was uploaded on 02/15/2012 for the course STAT 1000 taught by Professor Taeyoungpark during the Fall '06 term at Pittsburgh.

Page1 / 4

week13 - Lecture 32 Nancy Pfenning Stats 1000 Chapter 16:...

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