Question 1

If you were to have simply guessed the answer for each of the two previous questions, what is the mathematical probability that you would have gotten them both right?

Accepted Answer

A)   one in twelve
B)   one in sixty-four
C)   one in sixteen
D)   one in four C

Question 2

If a researcher reduces the likelihood of making a Type I error by using a .01 rejection level rather than the traditional .05 level, she:

Accepted Answer

A)   also reduces the likelihood of commiting a Type II error.
B)   increases the statistical power of the data analysis.
C)   increases the likelihood of commiting a Type II error.
D)   eliminates the possibility that sampling error might produce a relationship between variables within a research sample. C

Question 3

Effect size refers to:

Accepted Answer

A)   how strongly the values of one variable affect the values of another variable.
B)   the mathematical probability that a relationship between variables within a sample may have been caused by sampling error.
C)   the likelihood that design flaws may have produced a relationship between variables within a research sample.
D)   the ability of a statistical analysis to detect a true relationship between variables. A

Question 4

What factor that influences the statistical power of an analysis of the relationship between variables can be most easily controlled?

Accepted Answer

A)   sample size
B)   the actual strength of the relationship between variables
C)   the rejection level that should be used
D)   the amount of variability in the values of the variables

Question 5

A "true relationship" between variables is one that exists:

Accepted Answer

A)   within a research sample.
B)   because of rival hypotheses.
C)   within the population from which a research sample is drawn.
D)   because of sampling bias

Question 6

The natural tendency of research samples to differ from the population from which they were drawn and from each other is known as:

Accepted Answer

A)   sampling bias
B)   random error
C)   sampling error
D)   design flaw

Question 7

The term "Type II error" refers to

Accepted Answer

A)   mistakenly concluding that variables are unrelated when a true relationship exists.
B)   mistakenly concluding that there is a true relationship when variables are really unrelated.
C)   failing two recognize how sampling bias may have affected the research findings.
D)   using a non-directional research hypothesis when a directional one would have been justified.

Question 8

A "null finding" that two variables believed to be related are really unrelated:

Accepted Answer

A)   is of no value.
B)   is less valuable than if the variables had been found to be related.
C)   can be just as valuable as the finding of a relationship between the variables.
D)   suggests that the research was methodologically flawed in some way.

Question 9

When social workers tend to answer questions about social issues based on what they think they should believe about them (rather than what they really believe) this is an example of:

Accepted Answer

A)   systematic error
B)   random error
C)   social desirability bias
D)   cultural bias

Question 10

A statistically significant relationship between variables within a research sample is one in which:

Accepted Answer

A)   The relationship must be a true one.
B)   The relationship cannot possibly have been caused by sampling error.
C)   The relationship cannot have been caused by research design flaws, sampling bias, or other variables.
D)   The mathematical probability that sampling error could have caused the relationship is very small.

Question 11

A statistically significant relationship between variables within a research sample may not be:

Accepted Answer

A)   valuable to the social work practitioner.
B)   surprising or previously unknown.
C)   very strong.
D)   all of the above

Question 12

The "null hypothesis" states that:

Accepted Answer

A)   a relationship between variables within a research sample is a true relationship.
B)   a relationship between variables will not exist within a research sample.
C)   a relationship between variables within a research sample was caused by sampling bias.
D)   a relationship between variables within a research sample was caused by sampling error.

Question 13

A null research hypothesis predicts that:

Accepted Answer

A)   variables will be found to be related, but in an unknown direction.
B)   variables will found to be unrelated.
C)   variables will be found to be related and in a certain direction.
D)   variables will not be found to be related in a research sample, but they are related in the population from which it was drawn.

Question 14

A p value is the mathematical probability that a relationship between variables within a research sample could have been produced by:

Accepted Answer

A)   unethical research methods.
B)   sampling error.
C)   sampling bias.
D)   some other variable or variables.

Question 15

When variables are related within data drawn from a research sample, the relationship may have been caused by:

Accepted Answer

A)   design flaws
B)   rival hypotheses
C)   sampling bias
D)   all of the above

Exam 5: Hypothesis Testing: Basic Principles