Question

Use PROC GLM in SAS to test whether the mean number of days of sick leave differs across 
bank branches.  Use alpha= 0.05, and justify your answer by referring to either a test statistic value 
and table value, or by referring to a P-value.

data banks;
input branch leave;
cards;
1 15
1 20
1 19
1 14
2 11
2 15
2 11
3 18
3 19
3 23
;
run;

Answer 1

Solution:

Treat branch as a grouping variable using class statement in proc glm

use model statement  

SAS Code:

data banks;
input branch leave;
cards;
1 15
1 20
1 19
1 14
2 11
2 15
2 11
3 18
3 19
3 23
;
run;

proc glm data = banks;
class branch;
model leave=branch ;
means branch;
run;

Output:

The GLM Procedure

Class Level Information
Class	Levels	Values
branch	3	1 2 3

Number of Observations Read	10
Number of Observations Used	10

The GLM Procedure

Dependent Variable: leave

Source	DF	Sum of Squares	Mean Square	F Value	Pr > F
Model	2	89.8333333	44.9166667	6.21	0.0282
Error	7	50.6666667	7.2380952
Corrected Total	9	140.5000000

R-Square	Coeff Var	Root MSE	leave Mean
0.639383	16.30528	2.690371	16.50000

Source	DF	Type I SS	Mean Square	F Value	Pr > F
branch	2	89.83333333	44.91666667	6.21	0.0282

Source	DF	Type III SS	Mean Square	F Value	Pr > F
branch	2	89.83333333	44.91666667	6.21	0.0282

The GLM Procedure

Level of branch	N	leave
		Mean	Std Dev
1	4	17.0000000	2.94392029
2	3	12.3333333	2.30940108
3	3	20.0000000	2.64575131

F test with F=6.21 and p=0.0282 confirms the suspicion,based on the plots and descriptive statisitcs,that there are differences
in the mean number of days of sick leave differs across bank branches.