Deck 16: Inferential Statistics

Question 1

The symbol σ (the lower case Greek letter sigma) is usually used in statistics to refer to the:

A) sample variance.
B) population mean.
C) sample mean.
D) population standard deviation.

Answer 1

D

Question 2

The population mean is often shown with the lowercase Greek letter:

A) α.
B) μ.
C) π.
D) σ.

Answer 2

B

Question 3

Which of the following will increase the width of the confidence interval?

A) Raising the mean
B) Lowering the mean
C) Increasing the sample size
D) Decreasing the sample size

Answer 3

C

Question 4

Statisticians often construct an interval around their point estimates. There are a few ways this can be done. The most common way is called:

A) confidence intervals.
B) p-values.
C) maximum likelihood.
D) effect range intervals.

Question 5

Suppose you conducted a study comparing the means of two groups. You conduct a group t-test and observed p < .05. You can conclude:

A) the study worked.
B) there is a large difference in the means between the groups.
C) which group, in the population, likely has the higher mean.
D) the study failed.

Question 6

Which of the following is an often used formula for the standard error of the mean?

A) Variance/standard deviation
B) Variance/sample size
C) Standard deviation/square root of sample size
D) Sample size/variance

Question 7

One method to calculate a 95% confidence interval is: to sample, with replacement, from the original variable and calculate the statistic; to repeat this thousands of times; and to take the middle 95% of these replicate statistics. This is called:

A) ANOVA.
B) bootstrapping.
C) correlation.
D) Efronian super-inference.

Question 8

You have collected data from 16 people on a health anxiety scale. The mean is 100 and the standard deviation is 40. What is the 95% confidence interval (to two decimal places):

A) (80.00, 120.00).
B) (78.69, 121.31).
C) (98.42, 101.58).
D) (94.70, 105.30).

Question 9

If you are testing a null hypothesis, and incorrectly fail to reject it, when it is false, this is called an error of the:

A) first kind: a Type 1 error.
B) second kind: a Type 2 error.
C) third kind: a Type 3 error.
D) fourth kind: a Type 4 error.

Question 10

The statistical concept of power is most appropriately described as:

A) the likelihood of rejecting a null hypothesis when the true effect is of some specified size.
B) the likelihood of rejecting any null hypothesis.
C) the likelihood of correctly not rejecting a true null hypothesis.
D) the likelihood that your observed sample effect size is in the same direction as the population effect size.

Question 11

You have conducted a t-test. The result is: t(37) = 2.46, p = .02. What can you conclude?

A) The hypothesis that the result is important is confirmed
B) The hypothesis that result is large is confirmed
C) The hypothesis that is effect in the population is zero is rejected
D) The result does not allow any hypothesis to be confirmed or rejected

Question 12

Which of the following is a common formula for the one-sample t-test?

A) (mean1 - mean2)*standard error
B) mean/standard error
C) mean/standard deviation
D) standard deviation/mean

Question 13

Which of the following is a reason that you might choose to do a Mann-Whitney test rather than a t-test?

A) Your sample size is large.
B) You are not sure about how your data are distributed.
C) The standard deviation of your data is large.
D) You have more than two groups.

Question 14

Suppose that you have collected data from two groups of people and want to compare the means of these. Which of the following tests is most appropriate (assuming the data meet assumptions)?

A) Group t-test
B) Paired t-test
C) χ2 test
D) Correlation

Question 15

You are reading an article where the author allocated people to groups and compared means with a Oneway ANOVA. The author reports: F(3,36) = 6.23, p = .05. How many groups did the study have, and how many participants in total?

A) 3 and 36
B) 4 and 36
C) 3 and 40
D) 4 and 40

Question 16

An odds ratio is used:

A) instead of a t-test when one group is much larger than another.
B) to measure the association between two binary variables
C) to test the hypothesis that two binary variables are associated.
D) instead of an ANOVA when there are multiple groups.

Question 17

You have conducted a study to examine the association between math scores and generalized anxiety. Your result is: r = .11, n=1,000, p < .001. Which of the following is the most appropriate interpretation?

A) The association in the population is positive, but small.
B) The correlation is too small to be interpreted.
C) The correlation is highly significant and therefore important.
D) You can reject the null hypothesis that mean of the math scores is the same as the mean of the generalized anxiety score because it is statistically significant.

Question 18

You collected data from one group of students using a well-known state anxiety scale. You then administer a less well-known state depression scale. You are interested in how scores on these scales are associated. Which of the following tests is most appropriate?

A) Group t-test
B) Paired t-test
C) χ2 test
D) Correlation

Question 19

You have conducted a study with twelve people. Each person gave two blood samples, one before watching Friday the 13th part 14, and one after watching it. You measure cortisol levels. You subtract these scores (after level minus before level) to measure the effect of the movie on cortisol level. The mean difference is +4 micrograms per deciliter. The difference has a standard deviation of 10. You conduct a one-sample (or paired) t-test. What t value do you get (to two decimal places)?

A) 4.38
B) 4.80
C) 11.55
D) 1.39

Question 20

For the above example, what is the correct degrees of freedom to use to look up in a t-test?

A) 10
B) 11
C) 12
D) Use a z table or an infinite number for a t table