|
The sample has mean ¯ x = 1.243 and standard deviation s = 0.448.
a. We wish to test
H 0 : µ = 1,
H 1 : µ > 1.
The test statistic has value
t 0 =
¯ x − 1
s/ √ n
=
1.243 − 1
0.448/ √ 20
= 2.43.
Because t 0 < t 0.01,19 = 2.539, we fail to reject H 0 at the 0.01 significance level. So there is
insufficient evidence, at the α = 0.01 level, to conclude that the population mean expense
ratio for large-cap growth mutual funds exceeds 1%.
b. Since we failed to reject H 0 in part (a) and in fact µ = 1.33 > 1, we have committed a type II
error.
c. By the formula on page 40 of Lecture 6,
β = Φ
?
z α −
δ √ n
σ
?
= Φ
?
2.33 −
(1.33 − 1) √ 20
0.5
?
= Φ(−0.62) = 0.2676.
So the power of the test is 1 − β = 0.7324.
|
|