In: Statistics and Probability
Dr. Roh was interested in whether there was any disparity in sentencing based on the race of the defendant. He selected at random 19 drug convictions for cocaine distribution and completed the prison terms given to the 10 whites and 10 blacks sampled. The sentence lengths (in years) are shown below for the white and black offenders
White |
Black |
3 |
4 |
5 |
8 |
2 |
7 |
7 |
7 |
4 |
5 |
5 |
9 |
5 |
7 |
4 |
4 |
3 |
5 |
2 |
8 |
FYI,
White |
Black |
s2/1= 2.20 |
s2/2=2.84 |
solution:
Given that
n1 = 10
n2 = 10
x1-bar = 4
x2-bar = 6.4
s1 = √2.2 = 1.483239697
s2 = √2.84 = 1.685229955
(a) Hypotheses:
Ho: μ1 = μ2
Ha: μ1 ≠ μ2
Decision Rule:
α = 0.05
Degrees of freedom = 10 + 10 - 2 = 18
(b) Lower Critical t- score = -2.100922037
Upper Critical t- score = 2.100922037
Reject Ho if |t| > 2.100922037
(c) Test Statistic:
Pooled SD, s = √[{(n1 - 1) s1^2 + (n2 - 1) s2^2} / (n1 + n2 - 2)] = √(((10 - 1) * 1.48323969741913^2 + (10 - 1) * 1.68522995463527^2)/(10 + 10 -2)) = 1.587
SE = s * √{(1 /n1) + (1 /n2)} = 1.58745078663875 * √((1/10) + (1/10)) = 0.709929574
t = (x1-bar -x2-bar)/SE = -3.380617019
p- value = 0.003331454
(d) Decision (in terms of the hypotheses):
Since 3.380617019 > 2.100922037 we fail to reject Ho
Conclusion (in terms of the problem):
There is sufficient evidence of disparity in sentencing based on the race of the defendant.