Deck 10: Using Between-Subjects and Within-Subjects Experimental Designs

Question 1

In the context of types of experimental designs, each treatment is administered to a different group of subjects in a _____.

A) higher-order factorial design
B) between-subjects design
C) within-subjects design
D) fractional factorial design

Answer 1

between-subjects design

Question 2

In an experiment on the effects of alcohol on memory, participants are first tested after drinking a nonalcoholic beverage and then tested again after drinking an alcoholic beverage. This is an example of a _____ design.

A) within-subjects
B) between-subjects
C) single-subject
D) matched-pairs

Answer 2

within-subjects

Question 3

In a single-subject design, _____.

A) each subject is exposed to all levels of an independent variable at different times
B) data are not averaged across subjects
C) reliability is assessed with replication
D) All of the answers are correct.

Answer 3

All of the answers are correct.

Question 4

The statistical variability among scores caused by the influence of variables other than your independent variable is called _____ variance.

A) systematic
B) extraneous
C) error
D) random

Question 5

The presence of error variance:

A) makes it difficult to determine if your independent variable was effective.
B) makes it easy to determine if your independent variable was effective.
C) increases the internal validity of an experiment.
D) increases the external validity of an experiment.

Question 6

A way of reducing error variance is to:

A) take steps to control extraneous variables.
B) randomize error variance across groups.
C) increase the effectiveness of your independent variable.
D) All of the answers are correct.

Question 7

Analysis of your data with inferential statistics:

A) eliminates all error variance from your results.
B) helps you decide whether your results are an effect of your independent variable or simply the result of error variance.
C) randomizes error variance across your treatment groups.
D) None of the answers is correct.

Question 8

In a _____ design, you assign subjects to one of two groups on a purely random basis.

A) matched-pairs
B) demonstration
C) randomized two-group
D) All of the answers are correct.

Question 9

An advantage of the randomized two-group design is that it:

A) controls error variance by taking into account the characteristics of your subjects.
B) is simple to carryout, and data are easy to analyze.
C) eliminates random error through selective assignment of subjects to groups.
D) All of the answers are correct.

Question 10

A drawback of the randomized two-group design is that it:

A) provides a limited amount of information about the effect of an independent variable.
B) leaves too many sources of variance unaccounted for.
C) is time-consuming to run.
D) produces data that are difficult to analyze.

Question 11

Dr. Jones conducts a randomized two-group experiment on the effect of a drug on learning. She finds that subjects who receive the drug score more poorly on a memory test than those who do not receive the drug. Which of the following would Dr. Jones be justified in concluding?

A) There is a linear relationship between the independent variable and the dependent variable.
B) There is a curvilinear relationship between the independent variable and the dependent variable.
C) There is no relationship between the independent variable and the dependent variable.
D) It would be dangerous to speculate on the nature of the function relating the independent and dependent variables.

Question 12

An experimental design with more than two levels of a single independent variable is called a _____ design.

A) matched-groups
B) single-factor multigroup
C) factorial
D) higher-order

Question 13

When you manipulate your independent variable quantitatively, you are using a _____.

A) nonparametric design
B) qualitative design
C) parametric design
D) quantitative design

Question 14

If you vary the quantity of a drug administered to subjects by using 10 mg, 20 mg, and 40 mg, you are using a _____ design.

A) parametric
B) nonparametric
C) multifactor
D) parametric and multifactor

Question 15

Manipulating your independent variable qualitatively results in a _____.

A) parametric design
B) quantitative design
C) multigroup design
D) nonparametric design

Question 16

Often in studies that look at the effects of drugs on behavior, it is found that a participant's behavior can be altered if he or she believes a drug was taken. To test for this possibility, you would use a:

A) single-subject design.
B) matched-pairs design.
C) multiple control group design.
D) design with several dependent variables.

Question 17

In an experiment designed to study the effects of a drug on behavior, participants in the control group received a sugar pill rather than either the active drug or no drug. These participants constitute a:

A) compound control group.
B) placebo control group.
C) blind control group.
D) None of the answers is correct.

Question 18

A design including a placebo control group is used to detect:

A) excessive error variance.
B) the effect of a participant's belief that he or she has received a drug.
C) volunteer bias.
D) All of the answers are correct.

Question 19

The matched-groups design is most useful when:

A) a randomized two-group design would be too time-consuming.
B) you suspect that some subject characteristics are strongly correlated with your independent variable.
C) you suspect that no relationship exists between subject characteristics and your dependent variable.
D) you cannot manipulate independent variables.

Question 20

In a matched-groups design, after selecting a sample of subjects, you:

A) randomly assign subjects to your groups.
B) select only the most intelligent subjects for inclusion in your research.
C) match pairs of subjects on some measured characteristic (e.g., intelligence), and randomly assign one member of each pair across experimental groups.
D) None of the answers is correct.

Question 21

An advantage of the matched-groups design over the random assignment is that the matched-groups design:

A) allows you to control subject variables that might otherwise obscure the effect of your independent variable.
B) requires fewer subjects.
C) uses more subjects, which increases the sensitivity of the experiment.
D) eliminates error variance completely.

Question 22

In a matched-groups design, if the matched characteristic does not have much effect on the dependent variable under the conditions of the study, then it results in:

A) an increase in the sensitivity of your experiment.
B) no effect on the sensitivity of your experiment.
C) a loss of statistical power that will reduce the ability to detect the effect of the independent variable.
D) None of the answers is correct.

Question 23

The matched equivalent of a randomized two-group design is the:

A) matched subjects design.
B) matched-pairs design.
C) matched-groups design.
D) None of the answers is correct.

Question 24

Although you can use matching with a single-factor multigroup design, it may not be advisable because:

A) matching on more than one variable is usually impossible.
B) subjects cannot complete enough measures to make matching worthwhile.
C) it becomes difficult to find three or more subjects that can be matched.
D) All of the answers are correct.

Question 25

Within-subjects designs cope with the problem of error variance by:

A) matching subjects prior to random assignment to conditions.
B) treating each individual as a separate unit in the statistical analysis of data.
C) using the same subjects in all treatment conditions.
D) using different subjects in each treatment condition.

Question 26

All other things being equal, compared to the between-subjects design, the within-subjects design is:

A) more powerful.
B) less powerful.
C) as powerful.
D) less time-consuming.

Question 27

An advantage of a within-subjects design over a matched-pairs between-subjects design is that:

A) in the within-subjects design, it is easier to measure and take into account relevant subject characteristics.
B) measuring subject characteristics is unnecessary in a within-subjects design.
C) error variance is totally eliminated in the within-subjects design.
D) it is easier to measure and take into account relevant subject characteristics, and error variance is totally eliminated.

Question 28

Which of the following is true of the within-subjects design?

A) It is less powerful than an equivalent between-subjects design.
B) Fewer subjects are required than in an equivalent between-subjects design.
C) Error variance is less adequately handled in a within-subjects design than in an equivalent between-subjects design.
D) All of the answers are correct.

Question 29

According to the text, a disadvantage of a within-subjects design is that:

A) it is less powerful than a between-subjects design.
B) more subjects are required than in a between-subjects design.
C) each subject must spend more time in your experiment and thus may become fatigued.
D) matching of subjects becomes cumbersome even in simple within-subjects designs.

Question 30

A serious problem with the within-subjects design is that:

A) it is unable to deal with error variance.
B) independent variables are more difficult to manipulate in a within-subjects design than in a between-subjects design.
C) there is a loss of power when a within-subjects design is adopted.
D) there is a potential for carryover effects.

Question 31

_____ occur(s) when a previous treatment alters the observed behavior in subsequent treatments.

A) Error variance
B) Carryover effects
C) Experimenter bias
D) Sensitization effects

Question 32

Which of the following is the most likely disadvantage of a within-subjects design experiment?

A) It is difficult to detect the effects of an independent variable.
B) It requires a greater number of subjects than a between-subjects experiment.
C) It suffers from carryover effects, which is a serious problem.
D) It is hard to expose every subject to all levels of an experimental treatment.

Question 33

Which of the following is an advantage of a within-subjects experimental design?

A) The likelihood of detecting the effects of an independent variable is greater than in a between-subjects design.
B) The effects of a prior treatment are not observed in subsequent treatments.
C) Subjects are not exposed to every level of the treatment.
D) The problem of subject attrition is avoided if a mistake occurs while administering the treatment.

Question 34

Identify a true statement about matched-groups experimental designs.

A) Matching with a large number of groups becomes unwieldy because it is difficult to find subjects with equivalent scores on the variables to be matched.
B) Control of subject variables is difficult because it obscures the effect of the independent variable under investigation.
C) This design works best when a matched characteristic has little or no effect on a dependent variable.
D) Lack of requirement for pretesting makes this design less demanding and time-consuming than a randomized design.

Question 35

Carryover effects are:

A) easy to eliminate.
B) potentially problematic in both between-subjects and within-subjects designs.
C) potentially problematic only in within-subjects designs.
D) not a serious problem in either between-subjects or within-subjects designs.

Question 36

Which of the following was not listed in the text as a source of carryover effects in within-subjects designs?

A) Contrast
B) Learning
C) Sensitization
D) Demand characteristics

Question 37

Subjects in a two-level within-subjects experiment on reaction time respond to 2,000 stimuli in the first condition and then immediately respond to 2,000 more stimuli in the second condition. You notice that the subjects perform more poorly in the second condition than in the first. The confounding variable here is:

A) learning.
B) fatigue.
C) adaptation.
D) habituation.

Question 38

In a within-subjects experiment, subjects are required to perform a complex motor task for the dependent variable. You find that they perform better in the second treatment than in the first. You conclude that your independent variable had a significant effect on motor behavior. However, an alternative explanation is that _____ produced a carryover effect.

A) habituation
B) sensitization
C) learning
D) fatigue

Question 39

One assigns the various treatments of an experiment in a different order for different subjects in _____.

A) counterbalancing
B) reverse ordering
C) a factorial design
D) confounding

Question 40

_____ provides every possible ordering of treatments and assigns at least one subject to each ordering.

A) The Latin square design
B) Partial counterbalancing
C) The Solomon four-group design
D) Complete counterbalancing

Question 41

With complete counterbalancing, order effects are dealt with by:

A) randomly selecting two or three treatment orders for inclusion in an experiment.
B) having treatment conditions presented in every possible order.
C) summing them across treatment conditions.
D) None of the answers is correct.

Question 42

A complete counterbalanced design is most practical in experiments:

A) where carryover effects are a minimal problem to start with.
B) where carryover effects can be totally eliminated with the proper counterbalancing procedures.
C) with a small number of treatments.
D) with a large number of treatments.

Question 43

In a(n) _____ design, only some of the possible treatment orders are included in a within-subjects experiment.

A) partially counterbalanced
B) incompletely counterbalanced
C) truncated counterbalanced
D) semi-counterbalanced

Question 44

A key assumption behind partial counterbalancing in a within-subjects experiment is that:

A) subject factors will have little effect on the dependent variable.
B) randomly chosen treatment orders will randomly distribute carryover effects among the treatments.
C) all treatment orders will produce equivalent carryover effects.
D) statistical control of carryover effects is impossible.

Question 45

In the _____ design, the number of treatment orders is equal to the number of treatments.

A) Solomon four-group
B) Rubic's cube
C) Latin square
D) mixed

Question 46

Counterbalanced designs are effective in reducing carryover effects in within-subjects designs if:

A) the magnitude of the carryover effect is small to begin with.
B) the carryover from Group A to B is different from the carryover from Group B to A.
C) no appropriate statistical controls exist.
D) the order effects induced by different orders are of the same approximate magnitude.

Question 47

No amount of counterbalancing will eliminate _____ carryover effects.

A) differential
B) subject-induced
C) within-group
D) nonequivalent

Question 48

The most serious form of carryover effects occurs when your experimental treatment produces:

A) fatigue.
B) habituation.
C) learning.
D) irreversible changes.

Question 49

One way to reduce the impact of carryover caused by practice effects is to:

A) habituate your subjects to your experimental situation prior to beginning it.
B) give subjects equivalent practice sessions to bring them all up to the same level of performance.
C) make treatment order an independent variable and counterbalance it.
D) None of the answers is correct.

Question 50

Carryover effects caused by adaptation and habituation can be reduced by:

A) allowing a rest period between treatment conditions.
B) giving subjects practice sessions to bring them all up to the same level of performance.
C) allowing time for all subjects to adapt or habituate to the experimental conditions.
D) scheduling treatment conditions out over several days.

Question 51

A viable alternative to counterbalancing is to:

A) assume that there will be no carryover effects.
B) make treatment order an independent variable in your experiment.
C) run only a single treatment order.
D) None of the answers is correct.

Question 52

Making treatment order an independent variable allows you to:

A) completely eliminate carryover effects.
B) partially eliminate carryover effects.
C) measure the amount of carryover and take it into account in future experiments.
D) not worry about carryover effects anymore.

Question 53

A drawback of making treatment order an independent variable is that:

A) subjects must be tested under every treatment order, resulting in a complex and demanding experiment.
B) sometimes, treatment order cannot be made into an independent variable.
C) treatment order is a correlational variable.
D) None of the answers is correct.

Question 54

Within-subjects designs are most useful when:

A) subject differences contribute heavily to variation in the dependent variable.
B) carryover effects are large.
C) carryover effects vary unpredictably.
D) subject differences do not significantly affect the value of the dependent variable.

Question 55

You are interested in studying how airline pilots react in emergency situations. To preserve external validity, you want to use real airline captains. In this case, a within-subjects design would be most desirable because:

A) you do not have to worry about carryover effects with a highly trained subject sample.
B) subject variables probably do not contribute much to the value of the dependent variable.
C) highly trained airline pilots probably don't have the time to participate in a between-subjects design.
D) you will probably be faced with a limited supply of participants.

Question 56

In cases where you want to assess changes in performance as a function of increasing exposure to treatment conditions, the most appropriate design is a _____.

A) between-subjects design
B) correlational design
C) within-subjects design
D) combined design

Question 57

The simplest form of the within-subjects design is the _____ design.

A) single-factor two-level
B) within-subjects factorial
C) Latin square
D) partially counterbalanced

Question 58

A drawback to the completely within-subjects factorial design is that:

A) too many subjects are required as the design gets larger and larger.
B) carryover effects cannot be dealt with.
C) as the number of factors increases, so does the number of treatments that the subjects must go through, making the design cumbersome and complex for subjects.
D) None of the answers is correct.

Question 59

A factorial design always has more than one:

A) level of a single independent variable.
B) dependent variable.
C) independent variable.
D) dependent and independent variable.

Question 60

An advantage of a factorial design is that you can identify:

A) both the main effects of independent variables and any interactions between independent variables.
B) potential sources of confounding.
C) variables on which subjects can be matched.
D) All of the answers are correct.

Question 61

In a factorial design, the main effects are:

A) the effects of the most important independent variables on your dependent variable.
B) the separate effects of each independent variable on your dependent variable.
C) the changes in the effect of one independent variable over levels of a second.
D) never interpreted because they yield no interesting information.

Question 62

If the effect of one independent variable changes over the levels of a second, a(n) _____ is present.

A) multiple main effect
B) interaction
C) confounding variable
D) matching variable

Question 63

In a two-factor factorial experiment, the effect of Factor A at a given level of Factor B is called _____.

A) the main effect of Factor A
B) the main effect of Factor B
C) the interaction of Factor A with Factor B
D) the simple effect of Factor A at the given level of Factor B

Question 64

Dr. Harris conducts a two-factor experiment where she plots her results on a graph and notices that the lines on the graph are not parallel (which is confirmed when she runs a statistical analysis). This pattern suggests the presence of:

A) only one main effect.
B) two main effects.
C) an interaction between two independent variables.
D) None of the answers is correct.

Question 65

In a _____ factorial design, each subject is exposed to every combination of levels of all the factors.

A) between-subjects
B) within-subjects
C) mixed between-within
D) higher-order

Question 66

A factorial design with more than two independent variables is called a _____ design.

A) multiple
B) multivariate
C) lower-order factorial
D) higher-order factorial

Question 67

In a 2* 2 1* 2 factorial design, there are:

A) two main effects and two interactions.
B) three main effects and two interactions.
C) three main effects and three interactions.
D) three main effects and four interactions.

Question 68

When more than three factors are included in a factorial design, _____.

A) the interpretation of main effects becomes difficult
B) the interpretation of higher-order interactions becomes increasingly difficult as the number of factors increases
C) fewer subjects can be used
D) None of the answers is correct.

Question 69

A _____ design includes a single dependent variable.

A) univariate
B) multivariate
C) single-factor
D) multifactor

Question 70

An experimental design with two or more dependent variables is termed _____ design.

A) multifactor
B) factorial
C) multivariate
D) multigroup

Question 71

Which of the following statements is true of confounding variables in an experiment?

A) They arise as a result of sensitization and habituation.
B) They can be eliminated by using factorial designs.
C) They alter the observed behavior in subsequent treatments.
D) They damage the internal validity of an experiment.

Question 72

Dr. Franke is conducting an experiment on group decision making. The experiment will run for an entire academic semester. Dr. Franke runs his experimental group at the beginning of the semester and his control group at the end of the semester. The confounding variable here is:

A) that different groups were run at different times during the semester.
B) not including more than one control group.
C) choosing to run his experiment at an academic institution.
D) None of the answers is correct.

Question 73

Which of the following is the most likely source of confounding?

A) Two-way interactions
B) Experimenter bias
C) Carryover effects
D) Fractional factorial designs

Question 74

According to the text, the confounding in the original "Pepsi Challenge" could have been avoided by:

A) randomly assigning participants to conditions.
B) counterbalancing.
C) using blind experimenters.
D) designing effective dependent variables.

Question 75

The best way to avoid confounding variables in an experiment is to:

A) statistically control for their effects after your experiment has been conducted.
B) adjust your independent variables during your experiment when you discover the confounding.
C) plan carefully how your independent variables are to be executed.
D) All of the answers are correct.

Question 76

Which of the following is effective in dealing with carryover effects in an experiment?

A) Sensitizing and habituating participants to a stimulus
B) Introducing confounding variables
C) Showing main interactions between dependent variables
D) Making treatment order an independent variable

Question 77

Which of the following is a way of dealing with the problem of error variance?

A) Increasing the effectiveness of the independent variable
B) Introducing confounding variables in an experiment
C) Avoiding the assignment of subjects to groups on a random basis
D) Choosing variables that specifically produce carryover effects

Question 78

The different values an independent variable takes on in an experiment are called its levels.

Question 79

Within-subjects designs can include only two levels of your independent variable.

Question 80

The performance of subjects in a control group indicates how they perform in the absence of an experimental treatment.