Deck 13: Experimental Research: Designs, Part 2

Question 1

The statistical procedure that typically is used to analyze data from an experiment that employs a nonequivalent control-group design is

A) analysis of variance.
B) analysis of covariance.
C) a t test for independent means.
D) a t test for correlated means.

Answer 1

analysis of covariance.

Question 2

$\begin{array}{lll}0 & \mathrm{X} & 0 \\ 0 & & 0\end{array}$
The graph above describes the

A) nonequivalent control-group design.
B) time series design.
C) counterbalanced design.
D) static-group comparison design.

Answer 2

nonequivalent control-group design.

Question 3

A static-group comparison experiment

A) involves only a single group of research participants, all of whom receive the same treatment.
B) requires that the research participants remain in the same location during the entire experiment, so as to avoid threats to the experiment's internal validity.
C) involves a treatment and control group, but only the treatment group is given the posttest.
D) involves a treatment and control group, but of which are given a posttest, but no pretest.

Answer 3

involves a treatment and control group, but of which are given a posttest, but no pretest.

Question 4

The main threat to the internal validity of a nonequivalent control-group experiment is

A) the control group will feel resentful, because control groups generally perform at a lower level on the pretest than do treatment groups.
B) the possibility that observed differences between the experimental and control groups might reflect preexisting differences between the groups rather than treatment effects.
C) differential attrition in the two groups across the duration of the experiment.
D) all of the above.

Question 5

Match each example with the type of treatment variable it represents.

-Traditional vs. innovative method of teaching reading

A) Potentially manipulable aspects of the environment
B) Relatively fixed aspects of the environment
C) Manipulated variables
D) Response characteristics of research participants
E) Organismic characteristics of research participants

Question 6

Match each example with the type of treatment variable it represents.

-Male vs. female students

A) Potentially manipulable aspects of the environment
B) Relatively fixed aspects of the environment
C) Manipulated variables
D) Response characteristics of research participants
E) Organismic characteristics of research participants

Question 7

Match each example with the type of treatment variable it represents.

-High vs. low scorers on an achievement test

A) Potentially manipulable aspects of the environment
B) Relatively fixed aspects of the environment
C) Manipulated variables
D) Response characteristics of research participants
E) Organismic characteristics of research participants

Question 8

Match each example with the type of treatment variable it represents.

-High vs. low socioeconomic level of a school's student population

A) Potentially manipulable aspects of the environment
B) Relatively fixed aspects of the environment
C) Manipulated variables
D) Response characteristics of research participants
E) Organismic characteristics of research participants

Question 9

Match each example with the type of treatment variable it represents.

-Vocational vs. academic curriculum

A) Potentially manipulable aspects of the environment
B) Relatively fixed aspects of the environment
C) Manipulated variables
D) Response characteristics of research participants
E) Organismic characteristics of research participants

Question 10

Students' scores on a test are an example of

A) a manipulated variable.
B) an organismic characteristic of research participants.
C) a response characteristic of research participants.
D) a relatively fixed aspect of the environment.

Question 11

In a factorial experiment, the researcher

A) analyzes the treatment variable into two or more factors.
B) analyzes the dependent variable into two or more factors.
C) determines the effect of one treatment variable on a dependent variable.
D) determines the effect of two or more treatment variables on a dependent variable.

Question 12

Aptitude-treatment-interaction research is primarily concerned with

A) the exploration of matches between learner characteristics and different instructional methods.
B) the use of the experimental method to develop aptitude tests.
C) the combined effects of response and organismic characteristics on research participants' performance.
D) whether the intensity and duration of treatments can increase students' aptitude for mastering different academic subjects.

Question 13

A 3 X 2 factorial design means that there are

A) six interaction effects.
B) three levels of one factor and two levels of another factor.
C) six factors.
D) three factors, two of which are manipulated.

Question 14

The main purpose of the Solomon four-group design is to determine the presence of

A) experimenter bias.
B) an aptitude-treatment interaction.
C) an attribute-treatment interaction.
D) pretest sensitization.

Question 15

Experiments can vary in the

A) number of independent variables.
B) number of research participants in each treatment or control group.
C) treatment and control conditions that each research participant is assigned.
D) all of the above.

Question 16

The single-case experiment is particularly well suited to

A) survey research methodology.
B) research on behavior modification.
C) use in cross-cultural studies.
D) case studies.

Question 17

Careful training of observers, operational definition of behaviors, and control of observer bias are all used to ensurein single-subject experiments.

A) precise description of the experimental conditions
B) baseline and treatment stability
C) reliable observation
D) all of the above

Question 18

Because of the need for repeated measurements in single-case designs, it is important to

A) obtain measurements of behavioral products rather than behavior.
B) standardize the measurement procedure.
C) discontinue treatment as soon as improvement occurs.
D) keep baseline observations to a ratio of 1 for every 5 intervention observations.

Question 19

The A-B-A-B design

A) avoids the problem of ending an experiment with withdrawal or reversal of the treatment.
B) eliminates the possibility that an observed treatment effect will be dependent on the particular baseline conditions included in the experiment.
C) has lower internal validity than the A-B-A design.
D) all of the above.

Question 20

Conditions other than the naturally occurring target behavior are used as controls in

A) an A-B design.
B) an A-B-A design.
C) an A-B-A-B design.
D) a multiple-baseline design.

Question 21

In the analysis of single-case data, it has been found that

A) graphical data plots are best, because they cannot be used to distort treatment effects.
B) the observations made at different intervals of the experiment are susceptible to serial dependency.
C) the percentage of treatment data points that exceed the highest baseline data point yields highly misleading results.
D) analysis of variance for repeated measures is superior to any other statistical technique.

Question 22

The A-B-A-B design

A) is the best single-case design for studying multiple treatments.
B) ends with treatment reversal.
C) yields findings about treatment effects that are dependent on the particular baseline conditions included in the experiment.
D) is low in internal validity.

Question 23

The major threat to the external validity of single-case experiments is

A) the use of one or a few cases in most single-case experiments as a basis for generalizing to a population.
B) the small number of observations during the course of a typical single-case experiment.
C) the focus on observing behavior rather than the research participant's cognitions during the course of most single-case experiments.
D) the use of a withdrawal procedure rather than a reversal procedure in the second baseline of most single-case experiments.

Question 24

If the range of difficulty of test items is limited, the test may be subject to

A) statistical regression.
B) instrumentation effects.
C) a lack of replicability.
D) a ceiling effect.

Question 25

An appropriate technique for determining the statistical significance of pretest-posttest change in experiments is

A) analysis of variance for repeated measures.
B) the t test for independent means.
C) the t test for correlated means.
D) a Mann-Whitney U test.

Question 26

The validity of gain scores based on repeated administration of the same test to research participants

A) does not depend on whether the test has equal intervals across all of its measurement points.
B) depends on whether the test has equal intervals across all of its measurement points.
C) does not depend on whether a given score on a test measures more than one ability.
D) depends on the number of factors manipulated in the experiment.

Question 27

In an experiment that includes a pretest and posttest, describe a statistical technique that can be used to reduce initial group differences resulting from nonrandom assignment of research participants to the experimental and control treatments. Indicate the function of the pretest and posttest in thestatistical technique.

Question 28

a. What is the main difference between a static-group comparison design and a nonequivalent controlgroup design?
b. Which design is stronger in internal validity, and why?

Question 29

a. What threat to external validity is controlled by the Solomon four-group design?
b. How is the threat to external validity controlled by this design?

Question 30

Five types of independent variables can be employed in an experiment. Name and describe three of these types of independent variables.

Question 31

a. What is the purpose of aptitude-treatment interaction (ATI) research?
b. What type of experimental design is employed in ATI research?

Question 32

Using a factorial design, a researcher conducted an experiment to determine how prior vocabulary knowledge and a new reading program influence students' reading comprehension. The new reading program is the experimental treatment, and the existing reading program is the control treatment.
a. Identify what would be a main effect in this experiment, and define what is meant by a main effect.
b. Identify what would be an interaction effect in this experiment, and define what is meant by an interaction effect.

Question 33

Why is the A-B-A-B design considered more acceptable from an ethical standpoint than the A-B-A design?

Question 34

What is the main difference between an A-B-A design and a multiple-baseline design in a single-case experiment?

Question 35

Which of the two single-case designs, A-B or A-B-A, has superior internal validity? Why?

Question 36

Describe two potential problems that can occur when raw gain scores (posttest scores minus pretest scores) are used to measure the amount of learning that individuals achieve as the result of an experimental intervention.

Question 37

Describe two alternatives to the use of gain scores to analyze research data from a correlational or experimental study that included administration of a pretest and a posttest.

Question 38

A research team conducted an experiment in which they administered a mathematics test before and after students had participated in a new individualized curriculum. Parallel forms of the test were used as the pretest and posttest, and the range of possible scores was 0-100 on each. In analyzing data from the experiment, the researcher found that students in the lowest quartile on the pretest (M = 15.5) made a good gain (posttest M = 37.6), whereas students in the highest quartile on the pretest (M = 83.7) declined slightly (posttest M = 80.1). From these differences in gain scores, they concluded that the new curriculum is more effective for students of low ability than for students of high ability. What are two other interpretations that can be made of the differences in gain scores?

Question 39

A research team wishes to determine which of two types of staff development is more effective. The administrators of a school district agree to the experiment, and all the teachers in two high schools agree to participate and to take the pretest and posttest. However, the administrators insist that the teachers cannot be randomly assigned to the staff development programs. All teachers in one high school must participate in program A, and all teachers in the other high school must participate in program B.
a. If the researchers agrees to this condition, what type of experimental design must she use and why?
b. What is the main threat to internal validity in this type of experimental design, and how can it be handled?

Question 40

A researcher plans to test the effectiveness of self-paced instruction vs. conventional instruction for male and female students of varying levels of ability. There are a total of 150 students: 60 boys and 90 girls. Within each of these groups (boys and girls), there are an equal number of high-, middle-, and low-ability students. Make a chart that illustrates the experimental design for this research project, and show the number of students in each experimental group. Also, list each interaction effect that can be analyzed.

Question 41

A new program has been developed to increase students' speed and accuracy in simple mathematical computations. You have been asked by the school district's curriculum specialist to evaluate its effectiveness. One of the first decisions you must make is whether to use a group or a single-case experimental design. What questions could you ask the curriculum specialist about the program and the purpose of the experiment to determine whether a group or single-case design is more appropriate, and what is the purpose of each question?