In: Statistics and Probability
A company operates three machines during three shifts each day. From production records, the data in the table below were collected. At the .05 level of significance test to determine if the number of breakdowns is independent of the shift.
Machine | ||||
Shift | A | B | C | |
1 | Observed | 46 | 11 | 13 |
Expected | 40.27933 | 14.07821 | 15.64246 | |
2 | Observed | 37 | 10 | 11 |
Expected | 33.3743 | 11.6648 | 12.96089 | |
3 | Observed | 20 | 15 | 16 |
Expected | 29.34637 | 10.25698 | 11.39665 |
Solution:
Given:
Level of significance = 0.05
Machine | ||||
---|---|---|---|---|
Shift | A | B | C | |
1 | Observed | 46 | 11 | 13 |
Expected | 40.27933 | 14.07821 | 15.64246 | |
2 | Observed | 37 | 10 | 11 |
Expected | 33.3743 | 11.6648 | 12.96089 | |
3 | Observed | 20 | 15 | 16 |
Expected | 29.34637 | 10.25698 | 11.39665 |
We have to test if the number of breakdowns is independent of the shift or not.
Hypothesis :
H0: the number of breakdowns is independent of the shift
Vs
H1: the number of breakdowns is NOT independent of the shift.
Test statistic:
Formula for Chi square independence
Where
Oij = Observed frequencies for ith row and jth column.
Eij = Expected frequencies for ith row and jth column.
Thus we need to make following table:
Oij | Eij | Oij2/Eij |
46 | 40.27933 | 52.5331 |
11 | 14.07821 | 8.5948 |
13 | 15.64246 | 10.8039 |
37 | 33.3743 | 41.0196 |
10 | 11.6648 | 8.5728 |
11 | 12.96089 | 9.3358 |
20 | 29.34637 | 13.6303 |
15 | 10.25698 | 21.9363 |
16 | 11.39665 | 22.4627 |
N = 179 |
Thus
P-value:
Use following Excel command:
=CHISQ.DIST.RT( X2 , df )
where
df = ( R - 1) X (C - 1)
R = Number of Rows = 3
C = Number of Columns = 3
thus
df = ( 3-1) X ( 3-1)
df = 2 X 2
df = 4
thus
=CHISQ.DIST.RT(9.8894,4)
=0.0423
Thus P-value = 0.0423
Decision Rule:
Reject H0, if P-value < 0.05 level of significance, otherwise we
fail to reject H0
Since P-value = 0.0423 < 0.05 level of significance, we reject H0.
That is we reject claim: the number of breakdowns is independent of the shift
Thus correct answer is:
A. Yes, you can reject the claim that the number of breakdowns is independent of the shift because the p-value = 0.0423