Top Banner
25-Apr-00 AGR206 1 ANCOVA One continuous Y variable. One or more continuous X variables (covariates). One or more class variables (treatments). Model: (Y-GM Y ) = (Y-Y’) + B (X-GM X ) + Y = j + B (X- GM X ) +
10

ANCOVA

Feb 13, 2016

Download

Documents

truda

ANCOVA. One continuous Y variable. One or more continuous X variables (covariates). One or more class variables (treatments). Model: (Y-GM Y ) = (Y-Y’) + B (X-GM X ) + e Y = m j + B (X- GM X ) + e. Use of ANCOVA. Increase sensitivity of ANOVA. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: ANCOVA

25-Apr-00 AGR206 1

ANCOVA One continuous Y variable. One or more continuous X variables

(covariates). One or more class variables

(treatments). Model:

(Y-GMY) = (Y-Y’) + B (X-GMX) +

Y = j + B (X- GMX) +

Page 2: ANCOVA

25-Apr-00 AGR206 2

Use of ANCOVA Increase sensitivity of ANOVA. Adjust the treatment means to

what they would be if all were at the average value of X.

In MANOVA (not considered any further).

Page 3: ANCOVA

25-Apr-00 AGR206 3

Assumptions Normal, independent and equal-

variance errors. X is not affected by treatments. Relationship between Y and X’s does

not need to be linear. Homogeneity of slopes. The model

that relates Y to X’s must be the same for all groups (except for intercept).

Page 4: ANCOVA

25-Apr-00 AGR206 4

Example: 3 treatments

Y

20

25

30

35

40

45

50

15 20 25 30 35X

Previous sales

New sales

Page 5: ANCOVA

25-Apr-00 AGR206 5

Uncorrected ANOVAAnalysis of VarianceSourceModelErrorC Total

DF 2

12 14

Sum of Squares 338.80000 307.60000 646.40000

Mean Square 169.400 25.633 46.171

F Ratio 6.6086Prob>F 0.0116

Means for Oneway AnovaLevel123

Number 5 5 5

Mean 38.2000 36.0000 27.2000

Std Error 2.2642 2.2642 2.2642

Std Error uses a pooled estimate of error variance

Page 6: ANCOVA

25-Apr-00 AGR206 6

ANCOVAEffect TestSourceXtrtX*trt

Nparm 1 2 2

DF 1 2 2

Sum of Squares 243.14124

1.26328 7.05047

F Ratio 69.4230 0.1803 1.0065

Prob>F <.0001 0.8379 0.4032

Effect TestSourceXtrt

Nparm 1 2

DF 1 2

Sum of Squares 269.02869 417.15091

F Ratio 76.7232 59.4828

Prob>F <.0001 <.0001

Least Squares MeansLevel123

Least Sq Mean 39.81740696 34.74201681 26.84057623

Std Error0.85755067960.84966045380.8384392070

Mean 38.2000 36.0000 27.2000

Page 7: ANCOVA

25-Apr-00 AGR206 7

HW06 Solution

Page 8: ANCOVA

25-Apr-00 AGR206 8

HW06 Solution

Page 9: ANCOVA

25-Apr-00 AGR206 9

HW06 Solution

Page 10: ANCOVA

25-Apr-00 AGR206 10

HW06 Solution