Name:__________________________________ Date:_______________ AP Statistics

My Quiz Review for Tests of Significance

1. According to the National Federation of the Blind, the number of non-institutionalized, male or female, all
ages, with all education levels in the United States reported to have a visual disability in 2014 was 2.2%. A
survey was conducted from a random sample of 500 American Indians found there were 19 with a visual
disability. Is the American Indian population different from that of the nation? Use α = .05.
Let ρ = proportion of American Indians who have visual disability. H0: ρ = 0.022 Ha: ρ ≠ 0.022
Procedure: One sample z-test for proportion.
Random: Stated. Independence/10%: Safe to assume there are more than 500(10) American Indians.
Large count: np0 = 500(0.022) = 11 ≥ 10 and n(1-p0) = 500(1-0.022) = 489 ≥ 10. p̂ =19/500 = 0.038.
pˆ − p 0 0.038 − 0.022
z= = = 2.439
p 0(1 − p 0) 0.022(1 − 0.022) For two-tail test, P(Z< -2.439 OR Z > 2.439) = 0.0147251
n 500
Because the P-value (0.0147) is less than α (0.05), we reject H0. There is convincing evidence to believe the
true proportion of American Indians who have visual disabilities is different from 0.022.
(a) Assuming the null hypothesis is true, what is the standard error of sample proportions in this
problem? Show your work.
p (1 − p ) 0.022 (1 − 0.022 )
SE = = = 0.00656
n 500
(b) Suppose the true proportion of the American Indian population with visual disabilities is ρ = 2.7%.
When α = 0.05, the power of the test is 0.721. Interpret this value.
Given that ρ = 2.7%, there is a probability of 0.721 of finding convincing evidence that the true proportion of
the American Indian population who are blind is not equal to 2.2%.
(c) What is β for this test? Interpret this value in context.
β = P(Type II error) = 1 – Power = 1 – 0.721 = 0.279. There is a 0.0279 probability of not finding
convincing evidence to believe ρ ≠ 2.2% when in reality ρ ≠ 2.2%.
2. A company that manufactures appliances claims that the mean life span of its microwave ovens is more than
10 years. You want to test this claim. A random sample of 21 microwave ovens and their life spans (in
years) is listed in the stem-and-leaf plot below. Can you support the company’s claim? α = 0.10
Mean Life Span Let µ = the mean life span of microwave ovens. H0: µ = 10 Ha: µ > 10
α = 0.10
7 12 Procedure: One sample t-test for mean
8 36 Random: Stated in stem.
9 346 Independence/10%: Safe to assume there are more than 21(10) =210 microwaves.
10 23457 Normal/Large sample: Because we do not know the shape of the population and
11 489 the sample size is small (n < 30), we look at the graph. The stem-and-leaf plot does
12 258 not show strong skewness or outliers. So, it is reasonable to use t-procedures.
13 27 x = 10.7 sx = 2.01
14 1
x −  10.7 − 10
Key 7 | 1 means 7.1 years. t = = = 1.652
s 2.01
n 21
Degrees of freedom = 21-1 = 20. P(t > 1.652) = 0.057. Because the P-value (0.057) is less than α (0.10), we
reject H0. There is convincing evidence to believe that the life span is greater than 10 years.
(a) How do you interpret the P-value?
Assuming that the mean life span of a microwave is µ= 10 years, there is a probability of p = 0.057 of getting
a sample mean of x = 10.7 years or more by random chance alone.
(b) What type of error might you have made? Explain the mistake in context.
We might have made a Type I error. A type I error occurs when we believe that the life span of microwaves
is greater than 10 years when, if fact, the lifespan is 10 years. P(Type I error) = α = 0.10.
3. The speed of a personal computer is measured by its clock speed, which indicates how fast its
central processing unit (CPU) can process data. A computer manufacturer now sells a desktop
computer whose nominal operating speed is 900 MHz. The marketing ploy is that 900 MHz is a
conservative estimate; the computer actually exceeds this speed. Define the parameter, state the
hypotheses, describe type I and type II errors and discuss the possible consequences of each type of
error.
µ = Average speed of the desktop computers (MHz).
H0: µ = 900 Ha: µ > 900

A Type I error occurs when we believe that the average operating speed is µ > 900 when, in fact, µ = 900.
A consequence is that customers might be disappointed.

A Type II error occurs when we believe that the average speed is µ = 900 when, in fact, µ > 900.
A consequence is that the company may lose potential customers because it didn’t market higher speeds.

4. In testing the hypotheses Ho:  = 0 vs Ha:  > 0, if we reject Ho at  = 0.05, then we

(a) can also reject Ho at  = 0.01. (b) can also reject Ho at  = 0.10. (correct)
(c) have a p-value that is greater than 0.05. (d) have a calculator with low batteries.

5. Using the following hypotheses, which choice below best describes a Type II error?
Ho: High school students get plenty of sleep. Ha: High school students do not get enough sleep.
(a) You conclude that high school students need more sleep but they really get plenty of sleep.
(b) You conclude that high school students get plenty of sleep but they really don’t need more.
(c) You conclude that high school students get plenty of sleep but they really do not sleep enough.
(d) You conclude that high school students do not get enough sleep because you are tired today.

6. Which of the following is not a condition for performing inference about a population mean µ?
(a) Inference is based on n independent measurements or the population is at least 10 times as large
as the sample.
(b) The population distribution is approximately normal or the sample size is large (n≥30).
(c) To use a z test, we must know the population standard deviation σ.
(d) The data are obtained from a random sample from the population of interest.
(e) Both np and n(1 – p) are 10 or greater. (correct)

7. Bags of a certain brand of tortilla chips claim to have a net weight of 14 ounces. A representative of a
consumer advocate group wishes to see whether there is any evidence that the mean net weight is less
than advertised. (Let Ho: µ = 14, Ha: µ < 14) To test the claim, he selects 16 bags at random and
determines the net weight of each. He finds the sample mean to be 13.82 ounces and the sample standard
deviation to be 0.24 ounces. P(t < -3) = 0.004468
(a) We would reject Ho at significance of 0.10 but not at 0.05.
(b) We would reject Ho at significance of 0.05 but not at 0.025.
(c) We would reject Ho at significance of 0.025 but not at 0.01.
(d) We would reject Ho at significance of 0.01. (correct)
(e) We would fail to reject Ho at the α = 0.10 level.

8. A random sample of 42 students was asked if they spent too much time procrastinating studying for tests.
The resulting proportion was 2.1 standard deviations below the null hypothesis. Assuming this was a
one-sided test (p<po), which of the following represents the range of the p-value to this inference
procedure? Normalcdf(lower: -1000, upper: -2.1, σ= 1, µ= 0) reveals P(Z < -2.1) = 0.0179
(a) .005<p-value<.01 (d) .025<p-value<.05
(b) .01<p-value<.02 (e) 0.05<p-value<.10
(c) .02<p-value<.025
9. Suppose you want to test H0: ρ = 0.7 against Ha: ρ > 0.7. A sample produces z = 2.01 for the value of the
test statistic. What is the P-value?
Normalcdf(lower: 2.01, upper: ∞, σ= 1, µ= 0) reveals p = 0.022. So, P(Z > 2.01) = 0.022

10. In the problem above, suppose Ha: ρ ≠ 0.7. Now, what is the P-value?
Because the z-distribution is symmetric, P(Z < -2.01 OR Z > 2.01) = 2∙(0.22) = 0.044.

11. Suppose you want to test H0: µ = 5 against Ha: µ < 5. A sample of size n = 30 produces t = -1.87 as the
value of the test statistic. What is the P-value?
Degrees of freedom = df = 30 - 1 = 29.
tcdf(lower: -∞; upper: -1.87, df: 29) reveals p = 0.0358.
So, P(t < -1.87) = 0.0358
12. In the problem above, suppose Ha: µ ≠ 5. Now, what is the P-value?
We need to double the probability from the previous question because the area from the calculator (and
table) is for a one-tailed test, and we are doing a two-tailed test here.
P(t < -1.87 OR t > 1.87) = 2∙(0.0358) = 0.0716.

13. In a test of H0: µ = 8 against Ha: µ > 8, you take a sample of size n = 25 from a Normal population. What
values of t are significant at α = 0.05?
Degrees of freedom = df = 25 – 1 = 24.
InverseT(Area to right=0.05, df = 24) reveals t = 1.71.
P(t > 1.71) ≈ 0.05 → t ≥ 1.71.
14. In the problem above, suppose that Ha: µ ≠ 8. Now, what values of t are significant at α = 0.05?
InverseT(Area to right=0.025, df = 24) = 2.06.
Because the t-distribution is symmetric, P(t < -2.06 OR t > 2.06) ≈ 0.05. So, t ≤ -2.06 OR t ≥ 2.06
15. Suppose you want to test H0: ρ = 0.3 against Ha: ρ > 0.3. Suppose the sample size is n = 50 and the
number of successes is x = 23. Find z and p.

p̂ = 23/50 = 0.46.
pˆ − p 0 0.46 − 0.3
z= = = 2.47
p 0(1 − p 0) 0.3(1 − 0.3)
n 50
Normalcdf(lower: 2.47, upper: ∞, σ= 1, µ= 0) reveals that p = 0.0068. → P(Z > 2.47) = 0.0068.
16. Suppose you want to test H0: µ = 23 against Ha: µ ≠ 23. In a sample of size n = 30, x =24.2 and sx = 5.7.
Find t and p.
x −  24.2 − 23
t= = = 1.15
s 5.7
n 30
tcdf(lower: -1.15; upper: 1.15, df: 29) reveals that P(-1.15 < t < 1.15) = 0.74.
So, P(t < -1.15 OR t > 1.15) = 1 – 0.74 = 0.26.
17. A test to screen for a serious but curable disease is similar to hypothesis testing, with a null hypothesis
of no disease, and an alternative hypothesis of disease. If the null hypothesis is rejected, treatment will
be given. Otherwise, it will not. Assuming the treatment does not have serious side effects, in this
scenario it is better to increase the probability of:
(a) making a Type 1 error, providing treatment when it is not needed. (correct)
(b) making a Type 1 error, not providing treatment when it is needed.
(c) making a Type 2 error, providing treatment when it is not needed.
(d) making a Type 2 error, not providing treatment when it is needed.

18. A random sample of 25 college males was obtained and each was asked to report their actual height
and what they wished as their ideal height. A 95% confidence interval for µ = average difference
between their ideal and actual heights was 0.8" to 2.2". Based on this interval, which one of the null
hypotheses below (versus a two-sided alternative) can be rejected?
(a) H0: µd = 0.5 (correct)
(b) H0: µd = 1.0
(c) H0: µd = 1.5
(d) H0: µd = 2.0

19. What is the power of a test?

(a) A strong force
(b) The probability that we correctly reject a false null. (correct)
(c) The probability that we correctly accept a true null.
(d) How tough it is.

20. Which of the following best describes a type II error?

(a) The null is true, but we mistakenly reject it.
(b) The null is false and we reject it.
(c) The null is false, but we fail to reject it. (correct)
(d) The null is true but we fail to reject it.

21. The probability of a type II error is

(a) alpha (c) beta
(b) sigma (d) 1-beta

22. A lower alpha level will increase a test's

(a) chance of type I error
(b) chance of type II error
(c) power

23. A new medicine is actually better than placebo, and your team failed to reject the hypothesis that it was
just as good as placebo
(a) type I error
(b) type II error (correct)
(c) power
(d) no error here

24. A new medicine is actually no better than placebo, but your team concluded it was effective.
(a) type I error (correct)
(b) type II error
(c) power
(d) no error here.