Deck 6: Using Multiple Sample Means to Talk About Populations: T-Tests and Anova

Question 1

We find that 47% of 500 Democrats are against a certain policy. Which is the best statistical procedure to use?

A) a chi-square test
B) a confidence interval
C) a t-test
D) ANOVA

Answer 1

B
Explanation:We have the proportion for a single group and a sample size.

Question 2

If the population mean for men is the same as the population mean for women, which of the following is least likely:

A) the mean of a sample drawn from the population of men will be much higher than the mean of a sample drawn from the population of women
B) the mean of a sample drawn from the population of men will be a bit higher than the mean of a sample drawn from the population of women
C) the mean of a sample drawn from the population of women will be a bit higher than the mean of a sample drawn from the population of men
D) the mean of a sample drawn from the population of women will be equal or nearly equal to the mean of a sample drawn from the population of men

Answer 2

A

Question 3

In a sampling distribution of sample mean differences, each "building block" represents:

A) a person
B) a sample mean
C) the difference between a sample mean and the population mean
D) the difference between two sample means

Answer 3

D

Question 4

We want to see if Democrats and Republicans differ in income (measured in dollars). Which is the best statistical procedure to use?

A) a chi-square test
B) a confidence interval
C) a t-test
D) ANOVA

Question 5

We run a t-test that elicits a conclusion of p<.01. Now, we draw samples that have twice the variation, but the rest of the statistics stay the same. Based on these new findings, which of the following cannot be true?

A) p>.05
B) p<.05
C) p<.01
D) p<.001

Question 6

We run a t-test that elicits a conclusion of p<.05. Now, we draw samples twice as large and the rest of the statistics stay the same. Based on these new findings, which of the following cannot be true?

A) p>.05
B) p<.05
C) p<.01
D) p<.001

Question 7

We want to see if Democrats, Independents, and Republicans differ in income (measured in dollars). Which is the best statistical procedure to use?

A) a chi-square test
B) a confidence interval
C) a t-test
D) ANOVA

Question 8

Which of these is NOT a calculation used in ANOVA?

A) within-group sum of squares
B) between-group sum of squares
C) across-group sum of squares
D) all of these calculations are part of ANOVA

Question 9

Which of these is most similar to calculating the variance?

A) calculating the between groups sum of squares
B) calculating the within groups sum of squares
C) calculating the F-value
D) calculating the t-value

Question 10

We have statistics for a variable for six samples: Catholic men, Jewish men, Muslim men, Catholic women, Jewish women, and Muslim women. We run an ANOVA for the male groups, and find that p<.01, meaning that we can claim that the male means are not the same. If each of the women's groups has twice the variation on the variable as each corresponding male group, which of the following cannot be true?

A) p>.05
B) p<.05
C) p<.01
D) p<.001

Question 11

In the example about housework attitudes and housework behavior, which of the following best captures the major finding?

A) Men's behavior seems to be affected by their attitudes, but women's attitudes do not seem to be affected by their attitudes.
B) Women's attitudes seem to be affected by their attitudes, but men's attitudes do not seem to be affected by their attitudes.
C) Both men's and women's behavior seem to be affected their attitudes.
D) Neither men's nor women's behavior seems to be affected by their attitudes.

Question 12

According to the textbook, how are chi-square and ANOVA similar to one another?

A) both set up hypothetical situations to which you compare your real data
B) both use ratio-level variables
C) both use ordinal-level variables
D) both have the goal of achieving statistical significance by having the real data be as similar as possible to the hypothetical data

Question 13

We want to see if Democrats, Independents, and Republicans differ in income (measured as low, medium, and high). Which is the best statistical procedures to use?

A) a chi-square test
B) a confidence interval
C) a t-test
D) ANOVA

Question 14

In the example about sleep problems and hours spent on email, the primary problem with the sneaky researcher's ANOVA results was that:

A) there were only five cases in each group
B) three of the four groups had very similar means
C) the data were based on a non-random sample
D) he was comparing only two groups

Question 15

In Willer's research on overdoing gender, which of the following what NOT a measure that he used?

A) support for the war in Iraq
B) hostility toward homosexuality
C) unwillingness to show pain
D) desirability of SUV

Question 16

In Willer's research on gender threat, who responded to the gender threat at a statistically significant level?

A) men
B) women
C) both men and women
D) neither men nor women

Question 17

Which of the following is a way the textbook critiques Willer's experiment?

A) the sample size wasn't large enough
B) there were more men in the sample than women
C) the subjects were not briefed about the point of the survey afterwards
D) the subjects were not drawn randomly from the university's list of students

Question 18

What research method does Abascal use in her article on race relations?

A) secondary analysis of others' survey data
B) analysis of survey data she herself collected
C) content analysis of television police dramas
D) experimentation

Question 19

What statistical procedure does Abascal use in her article on race relations?

A) chi-square
B) confidence intervals
C) t-tests
D) ANOVA

Question 20

How many groups did Abascal use in her article on race relations?

A) 2
B) 4
C) 6
D) 8

Question 21

Barker conducted his research on dogs at which workplace?

A) Google
B) Replacements Ltd.
C) Boeing
D) Pets at Play

Question 22

Barker used which statistical technique in his research on dogs at work?

A) chi-square
B) confidence intervals
C) t-tests
D) ANOVA

Question 23

Barker found that the employees who usually brought their dogs to work but during the period of his study left their dogs at home experienced:

A) declining stress levels through the work day
B) increasing stress levels through the work day
C) declining productivity throughout the day
D) increasing productivity throughout the day

Question 24

The textbook suggests, in trimming down the t-test SPSS output, to delete which test results:

A) Levene's Test for Equality of Variances
B) Latour's Test for Mean Differences
C) LaVonne's Test for Mean Equality
D) Lefleur's Test for the Standard Error

Question 25

The textbook suggests, when running an ANOVA, to ask SPSS to give you:

A) the corresponding t-test results
B) the descriptives for the variables
C) the reliability of the variables
D) the corresponding chi-square test results

Question 26

A common belief is that gay men realize their same-sex attraction earlier in life than lesbian women do. In the 2008GSS, they asked those attracted to their same sex at which age they were first attracted, and here are the results:
Run a t-test using these sample results and fully interpret your findings.

Question 27

Catholics historically have been stereotyped as having a bar-going culture, more so than Protestants. Here are some statistics from GSS2008:
So GSS2008 Catholics do go to bars more frequently than Protestants, but run a t-test to see if this is a significant difference.

Question 28

Interested in whether having young children causes one to go the public library more often, we run a t-test using the following statistics:
This is not a statistically significant difference. Assuming all the other results remain the same, find the smallest possible mean difference that WOULD give you a statistically significant difference.

Question 29

If our two variables are race and income, describe a research question that would lead you to use a t-test, and another research question that would lead you to use ANOVA.

Question 30

If our two variables are race (whites and blacks) and religiosity, describe a research question that would lead you to use a t-test, and another research question that would lead you to use chi-square.

Question 31

Explain both mathematically and conceptually, why having greater variation around the sample means in a t-test makes in more difficult to achieve statistical significance.

Question 32

With regard to ANOVA, describe the connections among the amount of variation in our samples, the within groups sum of squares, the F-value, and the p-value. Start your answer with the phrase: "As the variation in our samples increases…."

Question 33

Explain how Willer uses t-tests to study the effects of gender threat. Among which group are there statistically significant differences?

Question 34

In Abascal's study of race relations, one comparison of black allocators produced a statistically significant difference, while another comparison of black allocators did not. Describe each of these comparisons, and what Abascal says these findings imply about race relations.

Question 35

In his study of dogs at work, why was Barker concerned that there might have been social desirability bias among his subjects? In your answer, be sure to be clear that you know what social desirability bias is.