1. Lecture 37 Section 14.4 Wed, Nov 29, 2006
Test of Homogeneity
Lecture 37
Section 14.4
Wed, Nov 29, 2006

2. Homogeneous Populations
Two distributions are called homogeneous if they exhibit the same proportions within the same categories.
For example, if two colleges’ student bodies are each 55% female and 45% male, then the distributions are homogeneous.

3. Example
Suppose a teacher teaches two sections of Statistics and uses two different teaching methods.
At the end of the semester, he gives both sections the same final exam and he compares the grade distributions.
He wants to know if the differences that he observes are significant.

4. Example Does there appear to be a difference?
Or are the two sets (plausibly) homogeneous? A B C D F
Method I 5 7 36 17
Method II 11 18

5. The Test of Homogeneity
The null hypothesis is that the populations are homogeneous.
The alternative hypothesis is that the populations are not homogeneous.
H0: The populations are homogeneous.
H1: The populations are not homogeneous.
Notice that H0 does not specify a distribution; it just says that whatever the distribution is, it is the same in all rows.

6. The Test Statistic
The test statistic is the chi-square statistic, computed as
The question now is, how do we compute the expected counts?

7. Expected Counts
Under the assumption of homogeneity (H0), the rows should exhibit the same proportions.
We can get the best estimate of those proportions by pooling the rows.
That is, add the rows (i.e., find the column totals), and then compute the column proportions from them.

8. Row and Column ProportionsB C D F
Method I 5 7 36 17
Method II 11 18

9. Row and Column ProportionsB C D F
Method I 5 7 36 17
Method II 11 18
Col Total 12 54 24

10. Row and Column ProportionsB C D F
Method I 5 7 36 17
Method II 11 18
Col Total 12 54 24
10%
15%
45%
20%

11. Expected Counts
Similarly, the columns should exhibit the same proportions, so we can get the best estimate by pooling the columns.
That is, add the columns (i.e., find the row totals), and then compute the row proportions from them.

12. Row and Column ProportionsB C D F
Method I 5 7 36 17
Method II 11 18
Col Total 12 54 24
10%
15%
45%
20%

13. Row and Column ProportionsB C D F
Row Total
Method I 5 7 36 17 72
Method II 11 18 48
Col Total 12 54 24
10%
15%
45%
20%

14. Row and Column ProportionsB C D F
Row Total
Method I 5 7 36 17 72
60%
Method II 11 18 48
40%
Col Total 12 54 24
10%
15%
45%
20%

15. Row and Column ProportionsB C D F
Row Total
Method I 5 7 36 17 72
60%
Method II 11 18 48
40%
Col Total 12 54 24
120
10%
15%
45%
20%
Grand
Total

16. Expected Counts
Now apply the appropriate row and column proportions to each cell to get the expected count.
Let’s use the upper-left cell as an example.
According to the row and column proportions, it should contain
60% of 10% of 120.
That is, the expected count is
0.60 0.10  120 = 7.2

17. Expected Counts
Notice that this can be obtained more quickly by the following formula.
In the upper-left cell, this formula produces
(72  12)/120 = 7.2

18. Expected Counts
Apply that formula to each cell to find the expected counts and add them to the table. A B C D F
Method I 5
(7.2) 7
(10.8) 36
(32.4) 17
(14.4)
Method II
(4.8) 11 18
(21.6)
(9.6)

19. The Test Statistic
Now compute 2 in the usual way.

20. df = (no. of rows – 1)  (no. of cols – 1).
Degrees of Freedom
The number of degrees of freedom is
df = (no. of rows – 1)  (no. of cols – 1).
In our example, df = (2 – 1)  (5 – 1) = 4.
To find the p-value, calculate
2cdf(7.2106, E99, 4) =
At the 5% level of significance, the differences are not statistically significant.

21. TI-83 – Test of Homogeneity
The tables in these examples are not lists, so we can’t use the lists in the TI-83.
Instead, the tables are matrices.
The TI-83 can handle matrices.

22. TI-83 – Test of Homogeneity
Enter the observed counts into a matrix.
Press MATRIX.
Select EDIT.
Use the arrow keys to select the matrix to edit, say [A].
Press ENTER to edit that matrix.
Enter the number of rows and columns. (Press ENTER to advance.)
Enter the observed counts in the cells.
Press 2nd Quit to exit the matrix editor.

23. TI-83 – Test of Homogeneity
Perform the test of homogeneity.
Select STATS > TESTS > 2-Test…
Press ENTER.
Enter the name of the matrix of observed counts.
Enter the name (e.g., [E]) of a matrix for the expected counts. These will be computed for you by the TI-83.
Select Calculate.

24. TI-83 – Test of Homogeneity
The window displays
The title “2-Test”.
The value of 2.
The p-value.
The number of degrees of freedom.
See the matrix of expected counts.
Press MATRIX.
Select matrix [E].
Press ENTER.

25. TI-83 – Example
Work the previous example, using the TI-83.