Leven's Test Using R

 

Leven’s Test

Leven’s test is used to test the homogeneity of between variances within samples. Some statistical testing start by assuming that variances are equal among a certain samples or groups. For example, we need to test the homogeneity of variances of salaries within levels of education of the employees. The data are given below:

Salary	Level
21894.4	B
38539.2	B
38177.6	A
38738.4	B
36481.6	B
35226.4	A
15516	A
15516	A
30499.2	A
15180	A
38539.2	C
42812	C
47035.2	B
35506.4	B
31907.2	A
38743.2	B
33259.2	A
43174.4	B
38539.2	A
38901.6	B
35226.4	B
39984.8	A
49024	B
18020	A
15516	A
41507.2	A
37688.8	A
31346.4	B
15973.6	A
24338.4	A
28950.4	B
18020	A
24014.4	A
31346.4	A
24338.4	A
35226.4	A
28592.8	A
35226.4	A
48035.2	A
33744	A
33259.2	A
38177.6	C
36989.6	B
38901.6	A
38539.2	A
41035.2	B
35226.4	B
35226.4	A
40439.2	C
36989.6	C
24160	B
36140	A
15481.6	A
18020	A
19684.8	B
15481.6	A
33744	B
35226.4	A
38539.2	A
38743.2	B
38901.6	A
17192.8	A
30368.8	B
38539.2	A
25833.6	A
25833.6	A
28950.4	B
41793.6	C
38743.2	C
38539.2	B
31310.4	B

Using the Rcmdr, click on statistics, variances then Leven’s test. 

The results of the test are shown in the below figure. It can be seen that the F calculated value (0.006244) < F-Value (5.4764);thus, salaries are not homogenous between different educational levels of employees or there is a significant difference between variances according to educational levels.