Deck 6: Sampling Distributions, Rejection Regions, and Statistical Test Selection

Question 1

Why are non-parametric tests often well-suited for use in social work research?

A) Sample sizes are often small.
B) Samples may not be considered random.
C) The dependent variable is often considered to be only nominal or ordinal level.
D) all of the above

Answer 1

all of the above

Question 2

In a sampling distribution, the formula for the standard error of the mean is

A) the standard deviation of the population divided by the mean of the population.
B) the standard deviation of the population divided by the sample size.
C) the standard deviation of the population divided by the square root of the sample size.
D) the standard deviation of the population divided by the square root of the mean of the population.

Answer 2

the standard deviation of the population divided by the square root of the sample size.

Question 3

A researcher operationalizes the variable "marital status" as whether a client reports that he or she is married or single. He hopes to study the relationship between gender and marital status among clients in a correctional facility. What options for the selection of a test of statistical inference are available?

A) Only a parametric test can be used.
B) Only a non-parametric test can be used.
C) Either parametric or non-parametric tests can be used if the research sample is a random one.
D) Parametric tests can be used, but only if the size of the research sample is large.

Answer 3

Only a non-parametric test can be used.

Question 4

Which statement about sampling distributions is correct?

A) There is a different sampling distribution for every sample size.
B) The mean of a sampling distribution is the same as the mean of the population from which samples are drawn.
C) Sampling distributions consist of a distribution of the means of an "infinite number of samples of a given size."
D) all of the above

Question 5

What is a requirement for use of parametric tests?

A) One or more variables should be interval level or ratio level.
B) Dependent variables should be regarded as normally distributed.
C) Cases should be selected independently of each other.
D) all of the above

Question 6

How do parametric tests differ from non-parametric tests?

A) They are less powerful.
B) They require more assumptions for their use.
C) They are less difficult to compute.
D) all of the above

Question 7

In which sample size is there likely to be the greatest amount of sampling error?

A) 10
B) 25
C) 30
D) 100

Question 8

How can a researcher make a non-parametric test more powerful (less likely to miss a true relationship between variables)?

A) Use a more rigorous rejection level such as .01 or .005.
B) Use a larger research sample.
C) Select the rejection region only after the results of the test are known.
D) all of the above

Question 9

When a statistical test produces a very low p-value such as .01

A) it is probably safe to reject the null hypothesis.
B) there is support for a one-tailed or two-tailed research hypothesis.
C) there is a lack of support for a null research hypothesis.
D) all of the above

Question 10

Confidence intervals for very large samples are likely to be

A) the same as for smaller samples.
B) smaller than for smaller samples.
C) larger than for smaller samples.
D) less accurate than for smaller samples.

Question 11

Statistical tests of inference

A) all produce a p-value.
B) tell us how safe it would be if we were to reject the null hypothesis.
C) are different mathematical formulae designed for use under different conditions.
D) all of the above

Question 12

When using a rejection level of .05, the rejection regions for a two-tailed research hypothesis lie beyond z = or + or-

A) 1.65
B) 1.96
C) 2.58
D) 3.00

Question 13

The central limit theorem states that, if an infinite number of relatively large (over 30) samples were to be drawn from a population in which a variable is somewhat positively skewed, the sampling distribution for those samples would be

A) equally positively skewed.
B) more positively skewed than within the population.
C) a normal distribution.
D) negatively skewed.

Question 14

In parameter estimation, the range in which a population parameter is believed to fall is known as a

A) standard error of the mean.
B) confidence interval.
C) confidence level.
D) true mean.

Question 15

When a statistical test is described as "robust," it means that the test

A) is easy to compute without the help of a computer.
B) is unlikely to result in a Type II error.
C) requires the use of all actual case values for its computation
D) still produces accurate results if not all of its assumptions for use can be met.