In: Statistics and Probability
(Round all intermediate calculations to at least 4 decimal places.) Consider the following measures based on independently drawn samples from normally distributed populations: Use Table 4. |
Sample 1: s21s12 = 250, and n1 = 16 |
Sample 2: s22s22 = 231, and n2 = 11 |
a. |
Construct the 90% interval estimate for the ratio of the population variances. (Round "F" value and final answers to 2 decimal places.) |
Confidence interval | to |
b. |
Using the confidence interval from Part a, test if the ratio of the population variances differs from one at the 10% significance level. Explain. |
The 90% confidence interval (Click to select)containsdoes not contain the value 1. Thus, we (Click to select)cannotcanconclude that the population variances differ at the 10% significance level. |
Solution
Back-up Theory
Given X ~ N(μ1, σ12), Y ~ N(μ2, σ22), n1 ≠ n2
100(1 - α) % Confidence Interval for σ12/ σ22 is:
{(s12/ s22)/ (Fn1 – 1, n2 – 1, α/2), (s12/ s22) /(Fn1 – 1, n2 – 1, 1 - α/2)} where s1, s2 are respective sample standard deviations; Fn1 – 1, n2 – 1,α/2 and Fn1 – 1, n2 – 1, 1 -α/2 are respectively upper and lower (α/2) percent point of F-distribution with degrees of freedom (n1 – 1) and (n2 – 1); and n1 , n2 are sample sizes.
Now to work out the solution,
Part (a)
Here α = 0.1, s12 = 250, s22 = 231, n1 = 16, n2 = 11.
Substituting these values,
90% Confidence Interval for ratio of population variances: 0.38 to 2.75 Answer
Details of calculations
R = (s12/ s22); F1 = (Fn1 – 1, n2 – 1, α/2); F2 = (Fn1 – 1, n2 – 1, 1 - α/2) |
n1 |
16 |
n2 |
11 |
s1^2 |
250 |
S2^2 |
231 |
α |
0.1 |
R |
1.082251 |
F1 |
2.845017 |
F2 |
0.393125 |
LB |
0.380402 |
UB |
2.752942 |
n1 - 1 |
15 |
n2 - 1 |
10 |
α/2 |
0.05 |
1 - (α/2) |
0.95 |
Part (b)
The 90% Confidence Interval, 0.38 to 2.75 does contain the value 1. Thus, we cannot conclude that the population variances differ. Answer
DONE