In: Statistics and Probability
In order to test whether camshafts are being manufactured to
specification a sample of n = 50 camshafts are selected at random.
The average value of the sample is calculated to be 4.38 mm and the
depths of the camshafts in the sample vary by a standard deviation
of s = 0.42 mm.
Test the hypotheses selected previously, by filling in the
blanks in the following:
The test statistic has value TS= .
Testing at significance level α = 0.01, the rejection region
is:
less than and greater than (2 dec
places).
Since the test statistic (is in/is not in) the
rejection region, there (is evidence/is no evidence) to
reject the null hypothesis, H 0.
There (is sufficient/is insufficient) evidence
to suggest that the average hardness depth, μ, is different to 4.5
mm.
Were any assumptions required in order for this
inference to be valid?
a: No - the Central Limit Theorem applies, which states the
sampling distribution is normal for any population
distribution.
b: Yes - the population distribution must be normally
distributed.
Insert your choice (a or b): .
Since the sample size > 30, we can use z- score even though the population standard deviation is not known
An estimate of the population mean is the sample mean = 4.38
Standard error = σ/√n = 0.42/√50 = 0.0594
The complete hypothesis test is given below:
Data:
n = 50
μ = 4.5
σ = 0.42
x-bar = 4.38
Hypotheses:
Ho: μ = 4.5
Ha: μ ≠ 4.5
Decision Rule:
α = 0.01
Lower Critical z- score = -2.575829304
Upper Critical z- score = 2.575829304
Reject Ho if |z| > 2.58
Test Statistic:
SE = s/√n = 0.42/√50 = 0.05939697
z = (x-bar - μ)/SE = (4.38 - 4.5)/0.05939696961967 = -2.020305089
p- value = 0.043351751
Decision (in terms of the hypotheses):
Since 2.020305089 < 2.575829304 we are in the rejection region. So we fail to reject Ho.
Conclusion (in terms of the problem):
There is no sufficient evidence that the mean depth is different from 4.5 mm
The central limit theorem applies since the sample size > 30