Deck 13: Within-Subjects Designs

Question 1

A major disadvantage of between subjects designs is that

A) the effects of one treatment may alter the effectiveness of the other
B) subject differences may obscure treatment effects
C) one must use fewer independent variables
D) one can use only one dependent variable

Answer 1

subject differences may obscure treatment effects

Question 2

Which design would require the fewest participants?

A) between subjects design
B) counterbalanced design
C) matched pairs designs
D) within-subjects designs

Answer 2

within-subjects designs

Question 3

Which design would have the least power?

A) between subjects design
B) counterbalanced design
C) matched pairs designs
D) within-subjects designs

Answer 3

between subjects design

Question 4

Which design would be least likely to have internal validity?

A) between subjects design
B) matched pairs design
C) mixed design
D) within-subjects designs

Question 5

Relative to a pure within-subjects design, a between-subjects design will tend to have:

A) less power but more internal validity
B) less power and less internal validity
C) more power and more internal validity
D) more power and less internal validity

Question 6

Participants serve as their own controls in all of the following EXCEPT:

A) counterbalanced designs
B) matched pairs designs
C) mixed designs
D) within-subjects designs

Question 7

The matched pairs design establishes internal validity by

A) matching
B) random assignment
C) both a and b
D) neither a nor b

Question 8

The matched pairs design has more power than the simple experimentbecause of

A) matching
B) random assignment
C) both a and b
D) neither a nor b

Question 9

If matching is based on an irrelevant variable, the net effect might be

A) a very powerful experiment
B) a less powerful experiment
C) a more valid experiment
D) a less valid experiment

Question 10

Matching is desirable only when some uncontrolled variable has a considerable impact on the

A) investigator
B) participant
C) independent variable
D) dependent variable

Question 11

Twenty participants are studied in a simple experiment. Then, 20 participants are studied in a matched pairs experiment. Relative to the simple experiment, the matched pairs experiment has

A) less power, fewer df
B) more power, fewer df
C) less power, more df
D) more power, more df

Question 12

If the average difference between conditions is the same in a simple, between-subjects experiment as it is in a matched pairs experiment, then the t value for the matched pairs design will tend to be ____ the t value for the simple, between-subjects experiment.

A) the same as
B) bigger than
C) smaller than
D) very different from

Question 13

The matched pairs design should be analyzed by

A) the between groups t test
B) the within subjects t test
C) the between subjects ANOVA
D) the appropriate post hoc test

Question 14

Properly executed, the matched pairs design should have which of the following advantages over the simple experiment?

A) more internal validity
B) more power
C) both a and b
D) neither a nor b

Question 15

Matching on pretest scores can

A) improve power
B) improve external validity
C) harm external validity
D) all of the above

Question 16

Matching on pretest scores can do all of the following EXCEPT

A) improve power
B) hurt external validity
C) hurt construct validity
D) hurt internal validity

Question 17

The difference between a within-subjects and a between-subjects simple experiment is

A) there is no problem of group equivalence with simple experiment
B) each subject serves as his or her own control in a within-subjects design
C) a given subject's behavior is only measured once in a within-subject design
D) data from a simple experiment is analyzed using a dependent groups t test

Question 18

Using a within-subjects design,

A) eliminates the effects of random error due to individual differences
B) eliminates the effects of random error
C) both a and b
D) neither a nor b

Question 19

The within-subjects design has more power than between-subjects design because it

A) eliminates the effects of random error due to individual differences
B) collects more observations from each participant
C) both a and b
D) neither a nor b

Question 20

Advantages of within- over between-subjects designs include all of the following except that

A) there is less chance of carry-over
B) fewer subjects are needed
C) there are fewer chance differences between groups
D) each subject serves as his or her own control

Question 21

A major disadvantage of within-subjects designs is that

A) they require more subjects
B) subject differences may obscure treatment effects
C) the effect of one treatment may alter the effectiveness of another
D) one must use fewer independent variables

Question 22

Lengthening the time between trials should reduce

A) power
B) treatment carry-over effects
C) practice effects
D) fatigue effects

Question 23

What problem is associated with within-subjects experimental designs?

A) subject selection bias
B) regression toward the mean
C) order effects
D) none of the above

Question 24

As compared to between-subjects designs, within-subjects designs may

A) have more power
B) make subjects more aware of the experimental hypothesis
C) mistake practice and/or fatigue effects for treatment effects
D) all of the above

Question 25

Using placebo treatments would be most effective in reducing

A) fatigue effects
B) practice effects
C) sensitization
D) treatment carry-over effects

Question 26

In a within-subjects design, the order of treatments should be

A) the same for all participant
B) randomly determined for each participant
C) arbitrarily determined
D) left up to the participant

Question 27

When subjects perform double duty in both conditions of an experiment, the most efficient way to deal with order effects is to _______ the order in which conditions are experienced.

A) counterbalance
B) control
C) randomize
D) hold constant

Question 28

If the treatment is a two-level within-subjects factor and counterbalancing is the between-subjects factor, a significant interaction between treatment and counterbalancing tells you

A) very little
B) that a between-subjects design should have been used instead
C) that subjects getting one sequence of treatments (A-B) had a different total score than subjects getting another sequence (B-A)
D) that subjects tended to score differently on their first trial than they did on their second trial

Question 29

If the treatment is a two-level within-subjects factor and counterbalancing is the between-subjects factor, a significant counterbalancing main effect tells you

A) very little
B) that a between subjects design should have been used instead
C) that subjects getting one sequence of treatments (A-B) had a different total score than subjects getting another sequence (B-A)
D) that subjects tended to score differently on their first trial than they did on their second trial

Question 30

If you have a counterbalanced design and one group did better than the other, then you have a(n)

A) treatment main effect
B) treatment X counterbalancing interaction
C) counterbalancing main effect
D) order effect

Question 31

If we want to know whether participants score differently on the first trial than on the second, you should look at the:

A) treatment main effect
B) treatment X counterbalancing interaction
C) counterbalancing main effect
D) sequence effect

Question 32

A mixed design is one in which

A) there is one independent and one dependent variable
B) one independent variable is tested within subjects and another is tested between subjects
C) more than two kinds of dependent variables are used
D) carry-over effects are stronger in one direction than in another

Question 33

In a factorial experiment, when one independent variable is manipulated between subjects and a second independent variable is manipulated within subjects,the experimental design is referred to as a _______ design.

A) matched pairs
B) non-counterbalanced
C) mixed
D) completely randomized

Question 34

If you are interested in order effects, you should use

A) a pure within-subjects design
B) a pure between subjects design
C) a counterbalanced design
D) a matched pairs design

Question 35

If you cannot study many participants and order effects are a problem, you should use a

A) simple experiment
B) matched pairs design
C) mixed design
D) within subjects design

Question 36

The matched-pairs design uses matching to reduce the effects of random differences between participants.

Question 37

The matched pairs design has internal validity because of its use of matching.

Question 38

Because matched-pairs designs also use random assignment, you do not have to be concerned with the possibility that participants may figure out your hypothesis.

Question 39

You should never use a t test when analyzing data from a matched pairs design.

Question 40

One of the best ways to combat order effects is to randomly determine the sequence in which each participant will get the treatments.

Question 41

In the counterbalanced design, participants are randomly assigned to systematically varying sequences of conditions.

Question 42

Order effects are the same thing as sequence effects.

Question 43

Order is a within subjects factor.

Question 44

Counterbalanced designs permit you to assess both the effects of sequence and of order.

Question 45

Counterbalanced designs require more participants than pure within-subjects designs.

Question 46

A counterbalanced design is one form of a mixed design.

Question 47

What steps can be taken to reduce sensitization, practice, fatigue, and carry-over effects?

Question 48

You wish to evaluate the effects of a program designed to improve IQ scores. What design would you use? Why?