Deck 7: Hypothesis Testing

Question 1

It is impossible to prove something in statistics because ______.

A) statistics don't really tell us anything
B) statistics are inherently probabilistic
C) statistics are frequently misinterpreted
D) none of these

Answer 1

B

Question 2

What is the following? H₀

A) null hypothesis
B) alternative hypothesis
C) nil hypothesis
D) none of these

Answer 2

A

Question 3

How would your pronounce the following? H₁: µ ≠ 8

A) the null hypothesis is that "mu is not equal to 8"
B) the alternative hypothesis is that "mu is not equal to 8"
C) either of these
D) none of these

Answer 3

B

Question 4

When we reject the null hypothesis we ______.

A) conclude that we have found statistical significance
B) conclude that the null hypothesis is improbable
C) both a and b
D) none of these

Question 5

When we retain the null hypothesis we conclude ______.

A) the null hypothesis, although unlikely, is plausible
B) the alternative hypothesis is more probable than the null hypothesis
C) the null hypothesis is more plausible than the alternative hypothesis
D) we do not have enough evidence to reject the alternative

Question 6

When we make a conclusion that we have found statistical significance, we ______.

A) reject the null
B) reject the alternative
C) retain the null
D) retain the alternative

Question 7

The first step of the hypothesis testing process is ______.

A) state the hypotheses
B) conduct the statistical test
C) state the research question
D) derive a decision rule

Question 8

Research questions should ______.

A) be testable
B) be phrased in term of operational definitions
C) ask if there is no difference between variables
D) none of these

Question 9

Which of these is a valid research question?

A) "Do men and women like our product equally?"
B) "Do West Coast organizations sell as much as East Coast organizations?"
C) "Is sales performance higher after we provide training?"
D) all of these

Question 10

A null hypothesis must at a minimum contain ______.

A) the parameter to be tested
B) the theoretical relationship between the parameter and test value
C) a value to be tested against the parameter
D) all of these

Question 11

Which of these is an example of an appropriate stated null hypothesis?

A) H₁: µ = 0
B) H₀: µ > 90
C) H₀: µ ≠ 52
D) none of these

Question 12

When testing a hypothesis about µ, you will most likely compare a ______ to a hypothetical test value.

A) µ
B) x̄
C) $\sigma$
D) X

Question 13

Consider the following research question: Do men prefer our product more than women? This would be addressed with a ______.

A) two-tailed test
B) one-tailed test
C) both a and b
D) none of these

Question 14

Consider the following research question: Are customer complaint levels different this year than last year? This would be addressed with a ______.

A) two-tailed test
B) one-tailed test
C) both a and b
D) none of these

Question 15

Which of the following could appear in a null hypothesis?

A) =
B) >
C) <
D) none of these

Question 16

Which of the following could appear in an alternative hypothesis?

A) =
B) ≥
C) ≤
D) ≠

Question 17

A two-tailed test always ______.

A) is non-directional
B) uses ≠ in the alternative hypothesis
C) both a and b
D) none of these

Question 18

For the following H₀, indicate the most likely H₁: H₀: µ = 1

A) H₁: µ ≥ 1
B) H₁: µ < 1
C) H₁: µ ≠ 1
D) H₁: µ ≤ 1

Question 19

The most common significance level is ______.

A) 1%
B) 5%
C) 10%
D) 99%

Question 20

When alpha is smaller ______.

A) a larger sample size is required to reject the null
B) a smaller sample size is required to reject the null
C) beta is larger
D) beta is smaller

Question 21

When alpha decreases ______.

A) the probability of a Type I error increases
B) the probability of a Type I error decreases
C) the probability of a Type II error increases
D) the probability of a Type I error decreases

Question 22

The region is rejection is defined by ______.

A) $\beta$
B) P
C) $\alpha$
D) none of these

Question 23

If H₀ is true, when $\alpha$ = .05 it is ______ than when $\alpha$ = .01

A) more likely that you will falsely reject the alternative
B) less likely that you will falsely reject the alternative
C) more likely that you will falsely reject the null
D) less likely that you will falsely reject the null

Question 24

To minimize Type II error, you can ______.

A) decrease $\alpha$
B) increase n
C) decrease p
D) increase µ

Question 25

Assume that you have a fair coin. You flip this coin 10 times and conclude that it is unfair based upon a statistical test. You have ______.

A) made a correct decision
B) committed a Type II error
C) committed a Type I error
D) none of these

Question 26

Assume that you have an unfair coin. You flip this coin 10 times and conclude that it is fair based upon a statistical test. You have ______.

A) made a correct decision
B) committed a Type II error
C) committed a Type I error
D) none of the

Question 27

If $\alpha$ = .05 and p = .02 ______.

A) you should reject the null
B) you have found statistical significance
C) it is unlikely that your result is due to chance alone
D) all of these

Question 28

If $\alpha$ = .05 and p > .05 ______.

A) the test statistic was beyond the critical value
B) you have committed a Type I error
C) you should reject the null
D) none of these

Question 29

Which of the following could be a correctly reported test of the statistic "q"?

A) q = 16, p = .02
B) q > 43, p = .17
C) q = 19, p ≤ .87
D) none of these

Question 30

If $\alpha$ = .1 and p = .8, we would state in our formal conclusion ______.

A) "Retain the null"
B) "Retain the null and accept the alternative"
C) "Reject the null and accept the alternative"
D) "Reject the alternative and accept the null"

Question 31

If H₀ is true, it is possible to commit a Type II error

Question 32

In a particular measurement situation, it is possible to commit both a Type I and a Type II error.

Question 33

$\alpha$ and $\beta$ should be set before data is collected.

Question 34

p-values range from -1 to +1.

Question 35

If you flip a coin 30 times, and it is always heads, that coin is definitely unfair.

Question 36

µ ≠ 10 is a valid null hypothesis.

Question 37

All possible values for a statistical test are described by either the null or alternative hypothesis.

Question 38

When we retain the null hypothesis, we accept the alternative hypothesis.

Question 39

When p = .52, we have found statistical significance.

Question 40

"Are our sales numbers the same from year to year?" is a valid research question.

Question 41

The region is rejection is defined by $\alpha$ .

Question 42

Changing $\alpha$ does not effect $\beta$ .

Question 43

If the null contains ≤, the alternative must contain the ≥.

Question 44

It is more difficult to reject the null with a 1% significance level than a 5% significance level.

Question 45

A researcher that rejects the null when they should have retained the null has committed a Type I error.

Question 46

We might report p = .02 when reporting the significance by hand, but we would be more likely to report p < .05 when reporting the significance using a computer.

Question 47

"Reject the null and accept the alternative" is a valid conclusion.

Question 48

If your research question states "Did those that were trained have higher job performance than those who were not trained?", your alternative hypothesis is likely to contain >.

Question 49

In a non-directional test, there are two critical values.

Question 50

Describe the difference between a Type I and Type II error. When do these occur?

Question 51

A friend asks if he should set alpha at .05 or .01. Which do you recommend and why?

Question 52

Given a particular research question and a particular study conducted to test that question, is it possible to commit either a Type I or a Type II error? Are both possible, is only one possible, or is neither possible? Why?

Question 53

What does p < .05 mean? Explain how you would interpret this.