In: Statistics and Probability
For a sample of 900 police officers at a local police department, a researcher believes there is a relationship between “number of arrests per month” and “police use of force.” Using the following data, test the null hypothesis at the .01 level of significance that police use of force does not differ by the number of arrests per month that an officer makes. In so doing, identify: (1) the research and null hypothesis, (2) the critical value needed to reject the null, (3) the decision that you made upon analyzing the data, and (4) the conclusion you have drawn based on the decision you have made.
Number Of Arrests Per Month
Use of Force One Two Three Four or More Total
No Force 120 100 40 120 380
Force 120 140 100 160 520_
240 240 140 280 900
(1)
HA: Research Hypothesis: There is a relationship between “number of arrests per month” and “police use of force.”
H0: Null Hypothesis: There is no relationship between “number of arrests per month” and “police use of force.”
(2)
ndf = (r - 1) X (c - 1)
= (2 - 1) X (4 - 1)
= 3
= 0.01
From Table, critical value of = 11.3449
(3)
Observed Frequencies:
One | Two | Three | Four or more | Total | |
No force | 120 | 100 | 40 | 120 | 380 |
Force | 120 | 140 | 100 | 160 | 520 |
Total | 240 | 240 | 140 | 280 | 900 |
Expected Frequencies are calculated as follows:
Expected Frequencies:
One | Two | Three | Four or more | Total | |
No force | 240X380/900=101.33 | 240X380/900=101.33 | 140X380/900=59.11 | 280X380/900=118.22 | 380 |
Force | 240X520/900=138.67 | 240X520/900=138.67 | 140X520/900=80.89 | 280X520/900=161.78 | 520 |
Total | 240 | 240 | 140 | 280 | 900 |
Test Statistic is calculated as follows:
Observed (O) | Expected (E) | (O - E)2/E |
120 | 101.33 | 3.44 |
100 | 101.33 | 0.02 |
40 | 59.11 | 6.18 |
120 | 118.22 | 0.03 |
120 | 138.67 | 2.51 |
140 | 138.67 | 0.01 |
100 | 80.89 | 4.52 |
160 | 161.78 | 0.02 |
Total = | 16.72 |
Test Statistic = 16.72
Since calculated value of = 16.72 is greater than critical value of = 11.3449, the difference is significant. Reject null hypothesis.
(4)
Conclusion:
The data support the claim that there is a relationship between “number of arrests per month” and “police use of force.”