Analysis of Tire Experiment (Latin Square Design)
data tires;
input truck day brand mpg @@;
cards;
1 1 2 6.63 1 1 1 6.26 1 3 4 7.31 1 4 3 6.57
2 1 4 7.06 2 2 2 6.59 2 3 3 6.64 2 4 1 6.71
3 1 1 6.31 3 2 3 6.73 3 3 2 7.00 3 4 4 7.38
4 1 3 6.55 4 2 4 7.13 4 3 1 6.54 4 4 2 7.11
proc anova;
class truck day brand;
model mpg = brand truck day;
means brand/bon;
run;
Analysis of Variance Procedure
Class Level Information
Class Levels Values
TRUCK 4 1 2 3 4
DAY 4 1 2 3 4
BRAND 4 1 2 3 4
Number of observations in data set = 16
Dependent Variable: MPG
Sum of Mean
Source DF Squares Square F Value Pr > F
Model 9 1.63770000 0.18196667 11.77 0.0036
Error 6 0.09280000 0.01546667
Corrected Total 15 1.73050000
R-Square C.V. Root MSE MPG Mean
0.946374 1.833617 0.12437 6.78250
Source DF Anova SS Mean Square F Value Pr > F
BRAND 3 1.30705000 0.43568333 28.17 0.0006
TRUCK 3 0.06765000 0.02255000 1.46 0.3170
DAY 3 0.26300000 0.08766667 5.67 0.0348
Bonferroni (Dunn) T tests for variable: MPG
NOTE: This test controls the type I experimentwise error rate, but
generally has a higher type II error rate than REGWQ.
Alpha= 0.05 df= 6 MSE= 0.015467
Critical Value of T= 3.86
Minimum Significant Difference= 0.3397
Means with the same letter are not significantly different.
Bon Grouping Mean N BRAND
A 7.22000 4 4
B 6.83250 4 2
B
C B 6.62250 4 3
C
C 6.45500 4 1