IE 6200 Solutions: Practice Problems .

1. State the decision rule that would be used to test the following hypotheses. Evaluate the appropriate test
statistic and state your conclusion.

(a) H0 : µ = 120 versus H1 : µ < 120; ȳ = 114.2, n = 25, σ = 18, α = 0.08.

ȳ − 120
Solution: Reject H0 if 18 ≤ −1.41; z = −1.61; reject H0 .
√
25
(b) H0 : µ = 42.9 versus H1 : µ 6= 42.9; ȳ = 45.1, n = 16, σ = 3.2, α = 0.01.
ȳ − 42.9
Solution: Reject H0 if 3.2 is either ≤ −2.58 or z ≥ 2.58; z = 2.75 reject H0 .
√
16
(c) H0 : µ = 14.2 versus H1 : µ > 14.2; ȳ = 15.8, n = 9, σ = 4.1, α = 0.13.
ȳ − 14.2
Solution: Reject H0 if 4.1 ≥ 1.13; z = 1.17; reject H0 .
√
9

2. A random sample of size 16 is drawn from a normal distribution having s = 6.0 for the purpose of testing
H0 : µ = 30 versus H1 : µ 6= 30. The experimenter chooses to define the critical region C to be the set
of sample means lying in the interval (29.9, 30.1). What level of significance does the test have? Why is
(29.9, 30.1) a poor choice for the critical region? What range of ȳ values should comprise C, assuming the
same α is to be used?
 
29.9 − 30 Ȳ − 30 30.1 − 30
Solution: By definition, α = P (29.9 ≤ Ȳ ≤ 30.1 | H0 is true) = P 6.0 ≤ 6.0 ≤ 6.0 =
√ √ √
16 16 16
P (−0.07 ≤ Z ≤ 0.07 = 0.056. The interval (29.9, 30.1) is a poor choice for C because it rejects H0 for the
ȳ values that are most compatible with H0 (that is, closest to µ0 = 30). Since the alternative is two-sided,
H0 should be rejected if ȳ is either ≤ 30 − 1.91 √6.0
16
= 27.1 or ≥ 30 + 1.91 √6.0
16
= 32.9.

3. Recall the breath analyzers described in Example 4.3.5. The following are thirty blood alcohol determinations
made by Analyzer GTE-10, a three-year-old

(a) If µ denotes the true average reading that Analyzer GTE-10 would give for a person whose blood
alcohol concentration is 12.6%, test

H0 : µ = 12.6 versus H1 : µ 6= 12.6.

at the α = 0.05 level of significance. Assume that σ = 0.4. Would you recommend that the machine
be readjusted?
ȳ − 12.6
Solution: H0 should be rejected if 0.4 is either ≤ −1.96 or z ≥ 1.96. But ȳ = 12.76 and z = 2.19,
√
30
suggesting that the machine should be readjusted.
(b) What statistical assumptions are implicit in the hypothesis test done in part (a)? Is there any reason
to suspect that those assumptions may not be satisfied?
Solution: The test assumes that the yi0 s constitute a random sample from a normal distribution.
Graphed a histogram of the 30yi0 s shows a mostly bell-shaped pattern. There is no reason to suspect
that the normality assumption is not being met.
4. Suppose H0 : p = 0.45 is to be tested against H1 : p > 0.45 at the α = 0.14 level of significance, where
p = P (ith trial ends in success). If the sample size is 200, what is the smallest number of successes that will
cause H0 to be rejected?
k − 200(0.45)
Solution: The null hypothesis would be rejected if z = p ≥ 1.08 (= z0.14 ). For that to
p 200(0.45)(0.55)
happen, k ≥ 200(0.45) + 1.08 120(0.65)(0.35) = 98.

5. Among the early attempts to revisit the death postponement theory introduced in Case Study 6.3.2 was an
examination of the birth dates and death dates of 348 U.S. celebrities (134). It was found that 16 of those
individuals had died in the month preceding their birth month. Set up and test the appropriate H0 against
a one-sided H1 . Use the 0.05 level of significance.
1 1
Solution: Let p = P (person dies if month preceding birth mouth). Test H0 : p = 12 versus H1 : p < 12 .
1
16 − 348( 12 )
Given that α = 0.05, H0 should be rejected if z ≤ −1.64. In this case, z = q = −2.52, which
1
348( 12 )( 11
12 )
suggests that people do not necessarily die randomly with respect to the month in which they were born.
More specifically, there appears to be a tendency to “postpone” dying until next birthday has passed.

6. (a) Find P (T < 2.365) when ν = 7.

Solution: Looking for 2,365 along row ν = 7 for t-distribution when you find the correct column, look
at α value at the top. α = 0.025
2.365 = t0.025 ⇒ P (T > 2.365) = 0.025 ⇒ P (T < 2.365) = 1 − P (T > 2.365) = 1 − 0.025 = 0.975.
(b) Find P (T > 1.318) when ν = 24.
Solution: α = 0.1 ⇒ P (T > 1.318) = 0.1
(c) Find P (−1.356 < T < 2.179) when ν = 12.
Solution: P (T > 2.179) = 0.025 (row ν = 12) ⇒ P (T < 2.179) = 0.975
P (T < −1.356) can’t find directly in the table, so we use symmetry P (T < −1.356) = P (T > 1.356) =
0.1 ⇒ P (−1.356 < T < 2.179) = P (T < 2.179) − P (T < −1.356) = 0.975 − 0.1 = 0.875
(d) Find P (T > −2.567) when ν = 17.
Solution: P (T < −2.567) = P (T > 2.567) = 0.1 ⇒ P (T > −2.567) = 1 − P (T < −2.567) =
1 − 0.01 = 0.99

7. An electrical rm manufactures light bulbs that have a life span that is approximately normally distributed.
The population standard deviation is not known. A sample of 30 bulbs are found to have an average life
span of 800 hours and a sample standard deviation of 45 hours.

(a) Find a 90% condence interval for the population mean.

Solution: Population variance σ 2 unknown ⇒ use t-distribution (Ok. since population distribution
is approximately normal)

s = 35, n = 64, x̄ = 500, n = 25 → ν = 24, 1 − α = 0.95 ⇒ α = 0.05

2-sided confidence interval t α2 = t0.025 = 2.064 row nu = 24.

s s
x̄ − t α2 √ < µ < x̄ + t α2 √
n n
35 35
500 − 2.064 √ < µ < 500 + 2.064 √
25 25
485.552 < µ < 415.448
(b) Would a 99% condence interval computed from the same sample be wider or narrower than the condence
interval found in part (a)?
Solution: Wider
(c) Find a 85% condence lower bound for the population mean
Solution: 1 − α = 0.9 ⇒ α = 0.1 ⇒ tα = t0.1 = 1.318 row ν = 24
s 35
µ > x̄ − t0.01 √ = 500 − 1.318 √
n 25
µ > 490.774.