In: Statistics and Probability
A large national survey of American dietary habits showed a mean
calorie consumption of 2700 kcal and a standard deviation of 450
kcal among teenage boys. You are studying dietary habits of
children in your county to see if they differ from the national
norm.
(a) In a sample of 36 teenage boys, you find a mean consumption of
2620 kcal. At = 0.05, is this significant evidence that the mean
in your county differs from the national mean? Assume that the
standard deviation observed nationally can be used for .
(b) Using =0.05 and sample of size 36, what is the probability
that you will actually be able to detect a situation where your
county has a mean of only 2600 kcal? (That is, what is the power if
=2600?)
a)
Ho : µ = 2700
Ha : µ ╪ 2700
(Two tail test)
Level of Significance , α =
0.050
population std dev , σ =
450.0000
Sample Size , n = 36
Sample Mean, x̅ = 2620.0000
' ' '
Standard Error , SE = σ/√n = 450.0000 / √
36 = 75.0000
Z-test statistic= (x̅ - µ )/SE = ( 2620.000
- 2700 ) / 75.0000
= -1.067
p-Value = 0.2861 [ Excel
formula =NORMSDIST(z) ]
Decision: p-value>α, Do not reject null hypothesis
Conclusion: There is not enough evidence that the mean in your
county differs from the national mean
b)
true mean , µ = 2600
hypothesis mean, µo = 2700
significance level, α = 0.05
sample size, n = 36
std dev, σ = 450.0000
δ= µ - µo = -100
std error of mean, σx = σ/√n =
450.0000 / √ 36 =
75.00000
Zα/2 = ± 1.960 (two tailed
test)
We will fail to reject the null (commit a Type II error) if we get
a Z statistic between
-1.960
and 1.960
these Z-critical value corresponds to some X critical values ( X
critical), such that
-1.960 ≤(x̄ - µo)/σx≤ 1.960
2553.003 ≤ x̄ ≤ 2846.997
now, type II error is ,ß = P
( 2553.003 ≤ x̄ ≤
2846.997 )
Z = (x̄-true
mean)/σx
Z1 = (
2553.003 - 2600 ) /
75.00000 = -0.627
Z2 = (
2846.997 - 2600 ) /
75.00000 = 3.293
so, P( -0.627 ≤ Z
≤ 3.293 ) = P ( Z ≤
3.293 ) - P ( Z ≤ -0.627
)
= 1.000
- 0.265 = 0.7341 [
Excel function: =NORMSDIST(z) ]
power = 1 - ß = 0.2659