Deck 8: Experimental Research

Question 1

Describe an example of the parts of the classical experimental design and three preexperimental designs.

Answer 1

An example of the parts of the classical experimental design includes the following:

1. Independent variable: This is the variable that is manipulated or changed by the researcher. For example, in a study on the effects of caffeine on memory, the independent variable would be the amount of caffeine consumed.

2. Dependent variable: This is the variable that is measured or observed to see if it is affected by the independent variable. In the caffeine and memory study, the dependent variable would be the participants' performance on a memory test.

3. Control group: This is a group of participants who do not receive the experimental treatment or intervention. In the caffeine and memory study, the control group would be given a placebo instead of caffeine.

4. Experimental group: This is the group of participants who receive the experimental treatment or intervention. In the caffeine and memory study, the experimental group would be given varying amounts of caffeine.

Three preexperimental designs include:

1. One-shot case study: This design involves a single group of participants who are exposed to a treatment or intervention, and then their behavior is observed. For example, a researcher might give a group of students a study skills workshop and then measure their academic performance.

2. One-group pretest-posttest design: In this design, a single group of participants is measured on a dependent variable before and after they receive a treatment or intervention. For instance, a researcher might measure participants' anxiety levels before and after they undergo a stress reduction program.

3. Static-group comparison: This design involves comparing the performance of two different groups on a dependent variable, with one group receiving the treatment or intervention and the other group not receiving it. For example, a researcher might compare the test scores of students who received a tutoring program with those who did not receive the program.

Question 2

Explain a 3 X 2 design and factorial designs. Why do experimenters use them?

Answer 2

A 3 X 2 design refers to an experimental design with three levels of one independent variable and two levels of another independent variable. This means that there are a total of six different conditions in the experiment. For example, in a study on the effects of caffeine and exercise on cognitive performance, the 3 X 2 design could involve three levels of caffeine intake (no caffeine, low caffeine, high caffeine) and two levels of exercise (sedentary, active).

Factorial designs, on the other hand, involve manipulating two or more independent variables to study their individual and combined effects on the dependent variable. For example, a 2 X 2 factorial design involves two independent variables, each with two levels, resulting in four different conditions.

Experimenters use these designs because they allow for the examination of multiple factors and their interactions on the dependent variable. By manipulating and controlling different variables, experimenters can better understand the complex relationships between variables and make more accurate conclusions about cause and effect. Additionally, factorial designs allow for the study of interactions between variables, which can provide valuable insights into how different factors may work together to influence the outcome. Overall, these designs provide a more comprehensive and nuanced understanding of the relationships between variables compared to simpler experimental designs.

Question 3

How do the interrupted and the equivalent time-series designs differ?

Answer 3

Interrupted time-series design and equivalent time-series design are both quasi-experimental research designs used to evaluate the effects of interventions over time. However, they differ in their structure and the way they control for potential confounding variables.

Interrupted Time-Series Design:
An interrupted time-series design involves collecting data at multiple time points before and after an intervention is implemented. The key feature of this design is the "interruption" of the time series by the introduction of the intervention. Researchers look for changes in the trend or level of the outcome variable that can be attributed to the intervention. This design is particularly useful when randomization is not possible, and it can provide strong evidence for causality if the pattern of change is consistent with the hypothesized effect of the intervention.

The strength of the interrupted time-series design lies in its ability to detect changes in trends over time. However, it can be vulnerable to confounding factors that also change around the time of the intervention, such as other policy changes or seasonal effects.

Equivalent Time-Series Design:
An equivalent time-series design, also known as a non-equivalent control group time-series design, adds a comparison group to the interrupted time-series design. This comparison group is not exposed to the intervention. By including a control group, researchers can compare the trends and changes in the outcome variable for both the intervention group and the control group. This helps to control for external factors that might be influencing the outcome variable, as any such factors should affect both groups similarly.

The equivalent time-series design is stronger than the simple interrupted time-series design because it helps to account for alternative explanations for the observed changes, other than the intervention. This design is particularly useful when researchers suspect that other factors might be influencing the outcome variable and when these factors cannot be easily measured or controlled for.

In summary, while both designs are used to assess the impact of interventions over time, the equivalent time-series design provides a more robust approach by including a control group to help account for potential confounding variables. The interrupted time-series design, on the other hand, does not include a control group and focuses solely on the changes in the outcome variable following the intervention.

Question 4

What is control important in experimental research? How is it related to internal validity?

Question 5

Describe the testing effect and its relation to the Solomon four-group design.

Question 6

What is a double-blind experiment? Why would a researcher use it?

Question 7

What is reactivity and why is it less an issue in field experiments than ones conducted in a laboratory?

Question 8

Explain why mortality, maturation, and history effects threaten a researcher's ability to demonstrate a causal relationship. Why are they more common in longer experiments?

Question 9

Why is randomization used instead of matching when selecting subjects for most experiments?

Question 10

Many experiments use confederates, placebos, and various types of deception. Why is deception used in many experiments and how should researchers act ethically when using it?

Question 11

The difference between randomization and random sampling is

A) nothing, they are two words for the same thing.
B) randomization means randomly assigning people into two or more groups, random sampling meaning choosing a sample from a population in a random way.
C) random sampling means randomly assigning people into two or more groups, randomization means choosing a sample from a population in a random way.
D) randomization is a close approximation of true random sampling.
E) randomization uses random numbers while random sampling does not.

Question 12

Dr. Shortstuff conducted an experiment to raise IQ scores by having experimental subjects eat three bananas each day for a month, while control groups had no bananas. Below are his results (scores = average IQ score for group of subjects).
Which of the following can you say about the results?

A) The experiment proves that the Dr. Shortstuff's treatment works.
B) The experiment has selection bias problems.
C) All experimental subjects ended up the same whether or not they had the treatment.
D) A and B.
E) B and C.

Question 13

After the delivery of a baby, new parents at General Hospital were given a poison knowledge test. While at the hospital they attended a three-hour training session on poison. A week later, they took the poison test again. After the delivery of a baby, new parents at University Hospital were given the poison knowledge test just before leaving for home. One week later, they took the test again. What kind of design is being used?

A) one-group pretest-posttest control group
B) time-series
C) nonequivalent control group
D) cross-sectional design
E) multiple time-series

Question 14

Dr. Hooknose said, "I've got to debrief the subjects" and walked into the experimental lab. What is he going to do?

A) Measure the dependent variable.
B) Measure the independent variable.
C) Randomly distribute subjects into groups.
D) Tell the subjects what the experiment is really all about, because he had lied to them earlier.
E) Tell the subjects a story to deceive them, so they won't know the true hypothesis being tested.

Question 15

An article you read refers to the Hawthorne effect. What does this mean?

A) It is a special instrument for measuring emotionally-charged attitudes invented in the 1930s Hawthorne, England.
B) It is a type of experimental design using three control groups developed by Sir Thomas Hawthorne in the 1960s, when investigating the effects of noise on mental concentration.
C) It is what happens when the researcher uses a two-way mirror in research to observe people without them being aware of it.
D) It is the effect that occurs when a researcher measures the dependent variable such that all the subjects get a perfect, or near-perfect score. It means that by random chance alone the scores are likely to be lower the next time the same subjects are measured.
E) none of the above.

Question 16

Several people criticized Dr. Bluenose's experiment on the effects of density of living conditions on aggression for its external validity problems. This means problems with

A) randomization to the experimental or control group.
B) generalizability.
C) ethics.
D) posttest effects.
E) selection of subjects into experimental and control groups.

Question 17

Which of the following is NOT one of the seven parts of an experiment?

A) control group
B) random assignment
C) RDD
D) posttest
E) dependent variable

Question 18

The Vice President for Human Relations at Widget Manufacturing Co. wanted to see whether a flex-time program would improve worker morale. He randomly assigned employees to two groups. He measured each group's job satisfaction; then he gave one group flexible working hours. The other group remained on fixed working hours. After six months, he again measured the degree of job satisfaction for both groups. During the six-month period, the Board of Directors granted all workers a six percent raise to quiet complaints. During his analysis, he found that all of the experimental-group subjects happened to be office employees and all of the control-group people were factory employees. In addition, a high turnover rate meant that only half of the factory workers who began the experiment were still with the firm six months later. He found that the experimental group had higher satisfaction. Which of the following threats to internal validity DID NOT occur during the experiment?

A) history effect
B) selection bias
C) mortality
D) statistical regression
E) all of the above

Question 19

What experimental design did she use?

A) classical experiment
B) one-shot case study
C) time-series
D) static group comparison
E) one-group pretest-posttest

Question 20

During the course of her experiment, Yolanda learned that seven of the 12 clients in her new four-hour session had quit and walked out of the session before it was completed. What threat to internal validity does her experiment have?

A) selection bias
B) history effect
C) experimental mortality
D) maturation
E) testing effect

Question 21

Timothy got a job working for a political candidate who was running for mayor. He decided to apply his knowledge about experiments in his new job and test whether the method of distributing campaign literature caused more people to vote for a candidate. He distributed campaign literature about the candidate to the Riverdale neighborhood in a bright blue envelope that a volunteer hand-placed in each person's front door. He distributed the same campaign literature about the candidate to the Golfview neighborhood by mailing it in a plain white envelope that went to each address. He later learned that Riverdale was a low-income area of the city that was almost entirely Latino and was 95% Catholic. Golfview was a mostly upper-income area in the city and its residents were 100% white Protestants. What problem with internal validity might threaten his ability to trust the experimental results?

A) selection bias
B) history effect
C) experimental mortality
D) maturation
E) testing effect

Question 22

George used the classical experimental design to test ways to raise self-assurance in stressful situations. He gathered 100 students from Introduction to Psychology in a room and randomly assigned each to one of two groups. His independent variable was a toy called Rubik's Cube, in which a person needs to manipulate boxes bound together in a multi-colored cube to align their colors, and he created stress by having a large clock in the room. The dependent variable was speed in getting the Cube's colors to align. He measured the level of self-assurance of people in one group with a questionnaire with 15 Likert scale items; then had them manipulate Cubes that were rigged to be easy. They answered another self-assurance questionnaire after completing the task. He measured self-assurance in the second group with the same questionnaire; then had the people manipulate Cubes that were rigged to be impossible to solve. The subjects answered another self-assurance questionnaire after 50 minutes of trying to complete the impossible task. In what part of the experiment does George measure the level of his dependent variable after he introduced the independent variable?

A) control group
B) experimental group
C) pretest
D) posttest
E) random assignment

Question 23

Refer to the following paragraph to answer the questions below.
Jack Short and Sharon Tall want to increase the percentage of high-school freshmen who pass an algebra exam (dependent variable). They conduct experiments on five sets of classes (named A to E) in Chicago-area schools. They randomly assign students and give a pretest in September followed by a posttest in January. The treatment is the teaching method. The experimental groups are taught with special in-class demonstrations and pictorial games-quizzes on computers, but no homework assignments, textbooks, or tests. The control groups are taught in the traditional way with a textbook, homework assignments, formulas on the blackboard, and tests.


-Looking at all five sets, which set of classes appears to be made up of a nonrandom sample of students from the Chicago-area schools?

A) A
B) B
C) C
D) D
E) E

Question 24

Talk about:
-control group

Question 25

Talk about:
-debrief

Question 26

Talk about:
-deception

Question 27

Talk about:
-demand characteristics

Question 28

Talk about:
-design notation

Question 29

Talk about:
-experimental design

Question 30

Talk about:
-experimental group

Question 31

Talk about:
-experimental mortality

Question 32

Talk about:
-experimental realism

Question 33

Talk about:
-experimental expectancy

Question 34

Talk about:
-external validity

Question 35

Talk about:
-factorial design

Question 36

Talk about:
-field experiment

Question 37

Talk about:
-Hawthorne effect

Question 38

Talk about:
-history effects

Question 39

Talk about:
-interaction effect

Question 40

Talk about:
-internal validity

Question 41

Talk about:
-laboratory experiment

Question 42

Talk about:
-maturation effect

Question 43

Talk about:
-placebo

Question 44

Talk about:
-posttest

Question 45

Talk about:
-preexperimental designs

Question 46

Talk about:
-pretest

Question 47

Talk about:
-quasi-experimental designs

Question 48

Talk about:
-random assignment

Question 49

Talk about:
-reactivity

Question 50

Talk about:
-selection bias

Question 51

Talk about:
-subjects

Question 52

Talk about:
-testing effect

Question 53

Talk about:
-treatment