Deck 14: Single-N Experiments and Quasi-Experiments

Question 1

Which is NOT a prerequisite for inferring causality?

A) establishing covariation
B) establishing temporal precedence
C) establishing that the observed relationship is not due to nontreatment variables
D) random assignment

Answer 1

random assignment

Question 2

Which of the following is NOT essential for inferring causality?

A) Changes in one variable are accompanied by changes in another variable.
B) The cause must come before the effect.
C) All variables other than the cause and the effect must be kept constant.
D) None of the above: all are necessary for inferring causality.

Answer 2

All variables other than the cause and the effect must be kept constant.

Question 3

Which is a prerequisite for inferring causality?

A) establishing covariation
B) establishing temporal precedence
C) establishing that the observed relationship is not spurious
D) all of the above

Answer 3

all of the above

Question 4

If the score in the treatment condition is higher than the score in the no-treatment condition, ______ has been established.

A) covariation
B) temporal precedence
C) spuriousness
D) causality

Question 5

In a simple experiment, a statistically significant result has traditionally been interpreted as meaning that

A) temporal precedence has been established
B) a stable baseline has been established
C) there is less than a 5% chance that the relationship between the independent and dependent variable is spurious
D) there is less than a 5% chance that temporal precedence has been violated

Question 6

Establishing a stable baseline in a single-n design supports the idea that ____ has been approximated.

A) statistical significance
B) temporal precedence
C) covariation
D) keeping all relevant nontreatment variables constant

Question 7

Which of the following are obstacles to establishing a stable baseline?

A) the inability to control all relevant variables
B) studying a highly complex organism
C) studying a subject in a highly complex environment
D) all of the above

Question 8

In an A-B design, the "B" stands for

A) baseline measurement
B) stable baseline measurement
C) post-treatment measurement
D) dependent variable

Question 9

In a highly controlled A-B design, which of the following is NOT a serious threat to internal validity?

A) history
B) maturation
C) testing
D) none of the above; all are serious threats

Question 10

The ABA design

A) is a single n design
B) is a between-subjects design
C) is also called a reversal design
D) a and c

Question 11

The reversal design is superior to the A-B design because it

A) is less likely to mistake maturation effects for treatment effects
B) is less likely to mistake testing effects for treatment effects
C) both a and b
D) neither a nor b

Question 12

According to the text, the reversal design is superior to the A-B design because it

A) is less likely to mistake history effects for treatment effects
B) is less likely to mistake testing effects for treatment effects
C) both a and b
D) neither a nor b

Question 13

According to the text, the reversal design is superior to the A-B design because it

A) is less likely to mistake history effects for treatment effects
B) is less likely to mistake maturation effects for treatment effects
C) both a and b
D) neither a nor b

Question 14

The single n design which entails switching back and forth between the experimental and control condition is referred to as the _______ design.

A) non-equivalent control
B) reversal
C) multiple-baseline
D) time-series

Question 15

The reversal design is superior to the A-B design in its ability to

A) handle testing effects
B) handle maturation effects
C) handle carry-over effects
D) both a and b

Question 16

In a reversal design, behavior may fail to return to baseline levels after the treatment is withdrawn. This could be due to

A) carry-over effects
B) maturation effects
C) testing effects
D) all of the above

Question 17

By changing an A-B design to a A-B-A-B-A-B design, you probably do all of the following EXCEPT

A) decrease the risk of carry-over effects
B) rule out cyclical maturation effects
C) rule out cyclical practice effects
D) increase your study's internal validity

Question 18

Psychophysical researchers do all the following EXCEPT

A) collect more than one judgment per stimulus
B) minimize the total number of judgments that participants make
C) randomize the order of stimuli
D) counterbalance the order of stimuli

Question 19

In a multiple baseline design, you can be sure that the treatment caused the behavior change if

A) all behaviors change when the treatment is introduced
B) behavior returns to the original baseline after the treatment is withdrawn
C) only the treated behavior changes
D) treatment carry-over effects are limited

Question 20

Which of the following is something a within-subjects researcher would be much more likely to do than a single-n researcher?

A) Randomly assign participants to different orders.
B) Use placebo treatments to reduce sensitization.
C) Use pre-experimental practice trials to reduce practice effects.
D) a and b

Question 21

The strategies of the within-subjects experimenter and the single-n experimenter are most similar when it comes to

A) construct validity
B) external validity
C) internal validity
D) ethics

Question 22

Single n designs are useful when

A) you are interested in the effects of a treatment on one particular individual
B) you are interested in studying what you believe to be a universal process
C) both a and b
D) neither a nor b

Question 23

The value of quasi-experimental designs is that they

A) always have better external validity than a true experiment would
B) are more convincing than true experiments
C) often can be used to study problems that you could not study using randomized experiments
D) all of the above

Question 24

According to Cook and Campbell, which of the following do quasi-experiments lack?

A) experimental units
B) outcome measures
C) random assignment
D) treatments

Question 25

What term refers to the fact that extraordinary or unusual events are likely to be followed by more ordinary events?

A) history
B) maturation
C) mortality
D) regression

Question 26

Which threat to internal validity is not a threat to the pretest-posttest design?

A) history
B) maturation
C) selection
D) testing

Question 27

If no subjects dropped out of a pretest-posttest study, one threat to validity that you wouldn't have to worry about is:

A) history
B) maturation
C) mortality
D) testing

Question 28

If, in a pretest-posttest design, you administered the pretest and the posttest exactly the same way, you shouldn't have to worry about

A) history
B) instrumentation
C) mortality
D) testing

Question 29

In a pretest-posttest design, regression is an unlikely explanation for the results if

A) your measurements are completely free of random error.
B) subjects' scores on the pretest were not extreme
C) subjects' scores on the posttest were more extreme than their pretest scores
D) all of the above

Question 30

People do better on a task because of practice on that task. However, the researcher mistakes this practice effect for a treatment effect. The research's internal validity is poor due to

A) history
B) instrumentation
C) mortality
D) testing

Question 31

Which of the following will NOT cause a change in the participant?

A) history
B) instrumentation
C) maturation
D) testing

Question 32

Which of the following will cause a participant's scores to change, even though the participant hasn't actually changed?

A) instrumentation
B) statistical regression
C) both a and b
D) neither a nor b

Question 33

Studying participants under highly controlled conditions is most likely to reduce the threat of

A) history
B) maturation
C) testing
D) regression

Question 34

Conducting the study in a short period of time should reduce all of the following threats EXCEPT

A) history
B) maturation
C) mortality
D) regression

Question 35

Using a reliable measure would be most effective against which threat?

A) instrumentation
B) regression
C) selection
D) testing

Question 36

The major advantage of the time-series design over the pretest-posttest design is

A) instrumentation is less of a problem
B) history is less of a problem
C) the effects of several potential threats to validity can be estimated
D) the effects of several potential threats to validity can be controlled

Question 37

Which of the following poses a problem for the time-series design?

A) inconsistent instrumentation effects
B) inconsistent mortality effects
C) inconsistent testing effects
D) all of the above

Question 38

Which of the following is considered the time-series design's Achilles' heel?

A) history
B) testing
C) mortality
D) regression

Question 39

The primary difference between time-series designs and single-n designs is that single-n designs do a better job at

A) establishing temporal precedence
B) establishing covariation
C) keeping irrelevant variables under control
D) all of the above

Question 40

When a researcher later obtains a second group of participants which is similar to the first group, but has not experienced the treatment like the first group had done, she has a

A) simple experiment
B) multiple-baseline design
C) time-series design
D) non-equivalent control group design

Question 41

The major threat to the validity of the non-equivalent control group design is

A) history
B) maturation
C) mortality
D) selection

Question 42

Two researchers do a non-equivalent control group study. Researcher A matches the groups on key variables, Researcher B does not. All of the following are true EXCEPT

A) Researcher A has reduced the threat of selection
B) Researcher A is vulnerable to selection by maturation interactions
C) Researcher A is more vulnerable to testing effects
D) Researcher A is more vulnerable to regression effects

Question 43

When comparing a disadvantaged group that gets the treatment to an advantaged group that gets no-treatment, all of the following are true EXCEPT

A) Regression effects may make the treatment seem more effective than it is
B) Selection by maturation effects may make the treatment seem less effective than it is
C) The most convincing evidence for the treatment's effect is if the treatment group scores better than the no-treatment group
D) Matching will not make the groups equivalent.

Question 44

In order to establish that a treatment causes an effect, a researcher demonstrate that spuriousness is present.

Question 45

In order to establish covariation a researcher will compare the behaviors of the treatment and nontreatment groups once the treatment has been administered.

Question 46

Temporal precedence is established when you make sure that the change in behavior occurs before the treatment.

Question 47

Random assignment works by turning nontreatment variables into random variables.

Question 48

Both randomized experimenter and single-n experimenters compare conditions to determine covariation.

Question 49

The A-B-A design is useful in single-n research for ruling out consistent maturation and testing effects.

Question 50

Like single-n research, quasi-experiments can keep relevant nontreatment factors from varying.

Question 51

If you can keep participants from dropping out of your study, you can rule out the effects of mortality.

Question 52

In order to rule out the effects of regression in a quasi-experiment, simply select only those participants who have extreme scores.

Question 53

The time-series design uses fewer participants than the A-B single-n design.

Question 54

The quasi-experimental design is more versatile than an experimental design.

Question 55

Compare and contrast the single-n design with a within-subjects design.

Question 56

How does a quasi-experimenter argue that her results reach the equivalence of statistical significance?

Question 57

Distinguish between the single-n design and the quasi-experiment.

Question 58

What strategies would you use to eliminate or minimize each of the eight threats to validity?

Question 59

Identify the potential threats to validity for the following studies:
a. Dr. Grueloff gives her patients a depression inventory shortly after they begin therapy. Then, she gives them caffeine therapy for three months and gives them the inventory again. They score less depressed than when they entered therapy so she concludes that caffeine therapy works.
b. Dr. Hardloff matches two groups of patients on initial level of depression and income level. Then, he gives one group caffeine therapy and the other group regular therapy. Three months later, he retests each group and finds that the caffeine group is less depressed. Therefore, he concludes that caffeine therapy is better than regular therapy.