This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Midterm 1. Consider a CRD on the quantitative factor X with levels 60, 70, 80, 90 and 100. We are interested in the linear contrast L = c 1 E ( Y  X = 60)+ .. . + c 5 E ( Y  X = 100) with contrast coefficients c = ( 4 , 2 , , 2 , 4 ). (a) Under the linear model E ( Y  X ) = β + β 1 X , show that a test of L = 0 is equivalent to a test of β 1 = 0. (b) Analyze the data in the pilot study below. Is the linear contrast L significant? Report on lack of fit as well. (c) Write the noncentrality parameter for the contrast L as a function of β 1 . As suming σ = 10, construct a power curve (as a function of n ) for β 1 = . 5. How many replications would we need to detect β 1 = . 5 with 80% power? X Y 60 27.2, 23.9, 18.5 70 32.3, 29.7, 1.3 80 40.2, 37.8, 19.6 90 19.4, 37.1, 32.7 100 36.0, 29.0, 37.1 2. Consider the following 2 4 unreplicated factorial design: Run Response Run Response (1) 7. 2 d 12.9 a 6.5 ad 19.4 b 6.2 bd 12.4 ab 4.6 abd 17.5 c 9.5 cd 16.2 ac 4.1 acd 18.9 bc 5.8 bcd...
View
Full
Document
This note was uploaded on 10/01/2009 for the course STAT stat706 taught by Professor Johnm.grego during the Spring '08 term at South Carolina.
 Spring '08
 JohnM.Grego

Click to edit the document details