Deck 7: The Internal Validity of Research

Question 1

Define internal validity in the context of research design and explain its role in determining causality.

Answer 1

No Answer

Question 2

Distinguish between a confound and an artifact. Give an example of each.

Answer 2

No Answer

Question 3

Describe an experiment, based on your own interests, that has strong internal validity. Explain how the experiment achieved that level of internal validity.

Answer 3

No Answer

Question 4

Name and define the three types of confounds described in your text. Give an example of each.

Question 5

Define treatment confounds. Under what conditions are they likely to occur?

Question 6

Define history confound. Give an example.

Question 7

Define maturation confound. Describe a study in which this type of confound might be operating.

Question 8

Define testing confound. Describe a study in which this type of confound might emerge.

Question 9

Distinguish between a testing confound and a confound due to instrumentation change.

Question 10

What is the Solomon Four Group Experimental Design? What confounds can be ruled out when it is used?

Question 11

Define observer drift. How might it affect the results of an experiment?

Question 12

Define statistical regression. For what types of scores is it most likely to be a problem?

Question 13

Explain how a control group can be used to rule out time-related threats to validity.

Question 14

Under what conditions is it useful to pretest on the dependent variable? Under what conditions is it unnecessary?

Question 15

What is theoretical validity? How does it affect whether a study provides a valid test of a theory?

Question 16

Define selection bias. List two forms it can take.

Question 17

Choose one of the selection biases described in the text and explain how that bias can be controlled.

Question 18

What is a volunteer bias and how might it create a threat to internal validity?

Question 19

Explain the ways in which a mortality confound could limit the internal validity of longitudinal research.

Question 20

Distinguish between time threats and selection threats to internal validity. Give an example of each.

Question 21

Define the concept of reactivity. Explain how reactivity can threaten the internal validity of a research study.

Question 22

List two possible sources of reactivity in experimental research.

Question 23

Define evaluation apprehension. Explain how it might affect participants' responses on a dependent variable.

Question 24

Describe the ways in which evaluation apprehension differs cross-culturally.

Question 25

Define novelty effects. Explain how they might affect participants' responses in a study using a physiological measure of the dependent variable. What might you do to limit those effects?

Question 26

List four ways of reducing evaluation apprehension.

Question 27

List two ways of reducing novelty effects.

Question 28

Explain why it is important to keep a record of any instances of reactivity observed in research.

Question 29

Define demand characteristics. Explain how they threaten the internal validity of a research study.

Question 30

Define psychological reactance. What effects might it have on participants' responses on a dependent variable?

Question 31

Describe how aspects of the experimental procedure can increase the possibility that demand characteristics are operating.

Question 32

Name and define the three roles that participants can take in response to experimental demand.

Question 33

List four strategies experimenters can use to reduce the possibility that demand characteristics are operating.

Question 34

Define role-play control groups. Explain how they can be used to address the possibility that one's experimental procedures create demand characteristics.

Question 35

Define experimenter expectancy effects. Give an example.

Question 36

Describe two ways that experimenters can unconsciously create experimenter expectancy effects.

Question 37

Describe three strategies for reducing the possibility that experimenter expectancy effects are operating.

Question 38

What is condition masking? Give an example.

Question 39

Describe four ways an experimenter can reduce threats to internal validity in a study.

Question 40

A study exhibits internal validity when

A) its results generalize across a wide variety of settings and participant populations.
B) one can have confidence that the results of a study are due only to the effects of the independent variable.
C) it is carried out in a laboratory setting.
D) random sampling is used to select the research participants.

Question 41

To which principle of causality is the concept of internal validity most closely related?

A) covariation between proposed cause and effect
B) time precedence of the cause
C) absence of plausible alternative explanations for the effect of the independent variable
D) demonstration of the effect of the independent variable in a variety of settings

Question 42

If an extraneous variable is found to be present, the researcher

A) can control for it by using appropriate statistical analyses.
B) cannot establish time precedence of the cause.
C) cannot conclude that the independent variable caused the dependent variable.
D) all of the above.

Question 43

Plausible alternative explanations for the effect of an independent variable can be ruled out

A) through logic, by showing that it is unlikely that the alternative explanatory variable could have had an effect.
B) by instituting control measures as part of the research design that limit the effects of any alternative explanatory variables.
C) by conducting research that shows that the alternative explanatory variables do not, in fact, affect the dependent variable.
D) by any of the above means.

Question 44

A confound exists when

A) two variables are combined so that the effect of one cannot be separated from the effect of the other.
B) some aspect of the research situation other than the independent variable affects the dependent variable.
C) the measures used in a study exhibit high reliability but low validity.
D) two studies testing the same hypothesis get different results.

Question 45

An artifact exists when

A) two variables are combined so that the effect of one cannot be separated from the effect of the other.
B) some aspect of the research situation other than the independent variable affects the dependent variable.
C) the measures used in a study exhibit high reliability but low validity.
D) some aspects of participants' personal histories influence how they respond on the dependent variable.

Question 46

A natural confound exists when

A) the procedures used to manipulate an independent variable affect the dependent variable.
B) two variables are associated with one another in the "real world."
C) the scores on a measure reflect the research participants' standing on a variable in addition to the one the measure is intended to assess.
D) the research participants are people who naturally inhabit the setting where the research is conducted.

Question 47

A treatment confound exists when

A) the procedures used to manipulate an independent variable unintentionally affect the dependent variable.
B) two variables are associated with one another in the "real world."
C) the scores on a measure reflect the research participants' standing on a variable in addition to the one the measure is intended to assess.
D) the research participants respond differently to levels of the independent variable.

Question 48

A measurement confound exists when

A) the procedures used to manipulate an independent variable affect the dependent variable.
B) the research participants respond differently to levels of the independent variable.
C) scores on a measure reflect the research participants' standing on a variable in addition to the one the measure is intended to assess.
D) a measure exhibits low validity but high reliability.

Question 49

Cohort effects are examples of a(n)

A) measurement confound.
B) treatment confound.
C) artifact.
D) natural confound.

Question 50

Professor Yu designs a study testing whether color can account for gender differences in spatial ability. He uses pastel-colored stimuli for the female participants and primary-colored stimuli for the male participants. In this study, stimulus color is a(n)

A) treatment confound
B) measurement confound
C) natural confound
D) artifact

Question 51

A researcher conducted an online study at the beginning of the semester. When the data were downloaded at the end of the term, she realized she had not included one of the experimental conditions, so she reopened the study and collected data from new participants. This procedure resulted in

A) a natural confound.
B) a treatment confound.
C) an artifact.
D) a measurement confound.

Question 52

A measurement confound results in a problem with the ________ of the measure.

A) reliability
B) convergent validity
C) differential validity
D) discriminant validity

Question 53

A history confound exists when, in addition to any effects of the independent variable,

A) participants' past experiences affect their response to the dependent variable.
B) taking a pretest affects participants' scores on the posttest.
C) an event occurs outside the research situation that affects participants' responses on the dependent variable.
D) children's scores on a measure change over time as a result of their cognitive development.

Question 54

Professor Grune is conducting an experiment on perceptions of the credibility of eyewitness testimony. While the study was in progress, a case of mistaken identity received wide publicity: A person who was on death row was exonerated because the key witness had mistakenly identified him as the murderer. Professor Grune found that his participants were less willing to believe an eyewitness after the case was publicized. His research was hampered by a ________ confound.

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

Question 55

Which of the following could create a history confound?

A) one of the participants experienced a major life event that is related to the topic of the study.
B) an event external to the study, but related to the topic, received widespread news coverage.
C) several participants recently completed another study that tested a similar hypothesis.
D) all of the above

Question 56

A maturation confound exists when, in addition to any effects of the independent variable,

A) participants' past experiences affect their response to the dependent variable.
B) taking a pretest affects participants' scores on the posttest.
C) an event occurs outside the research situation that affects participants' responses on the dependent variable.
D) participants' scores on a measure change over time as a result of natural aging processes.

Question 57

Professor Gelbe tested a new method of teaching reading to third graders. Over the course of the school year, the reading scores of Professor Gelbe's participants showed clear increases. A potential problem with the design of this study is

A) maturation.
B) statistical regression.
C) selection.
D) mortality.

Question 58

A testing confound exists when, in addition to any effects of the independent variable,

A) participants' past experiences affect their response to the dependent variable.
B) taking a pretest affects participants' scores on the posttest.
C) an event occurs outside the research situation that affects participants' responses on the dependent variable.
D) participants' scores on a measure change over time as a result of natural aging processes.

Question 59

A statistics professor wants to test whether completing her course increases students' critical thinking skills. She gives students a pretest measuring these skills on the first day of class. The test is repeated on the last day of class; she finds students' scores are higher at the end of the term. Which of the following is a potential confound?

A) testing
B) maturation
C) history
D) a and b, but not c

Question 60

Which of the following statements about the Solomon Four-Group Experimental Design is/are TRUE?

A) It is very simple to carry out.
B) It can be used to assess the extent to which taking a pretest might affect the results of the posttest,
C) It can be used to assess the extent to which statistical regression might be affecting the results of a study.
D) All of the above statements are true

Question 61

In the Solomon Four-Group Experimental Design

A) all participants complete the pretest and its effects are removed using statistical analyses.
B) the researcher can rule out statistical regression as a potential confound.
C) the need for a pretest is eliminated.
D) the researcher can rule out pretest sensitization as a potential confound.

Question 62

Instrumentation change occurs when

A) taking a pretest affects scores on the posttest independently of the effect of the independent variable.
B) the measure used to assess the dependent variable changes over time.
C) participants' scores on a measure change over time as a result of natural aging processes.
D) research participants who obtained extreme scores on a pretest have posttest scores that are closer to the mean.

Question 63

At XYZ Corporation, Professor Martinelli assessed workers' job satisfaction at pretest and posttest using a valid measure; she found that posttest scores were significantly higher than pretest scores. However, during interviews conducted after the post-test, Professor Martinelli found that the intervention had changed how participants interpreted the meaning of the questions on the job satisfaction measure. Professor Marinelli's research was hampered by

A) statistical regression.
B) maturation.
C) a history confound.
D) instrumentation change.

Question 64

Statistical regression occurs when

A) taking a pretest affects scores on the posttest independently of the effect of the independent variable.
B) the measure used to assess the dependent variable changes over time.
C) research participants who obtained extreme scores on a pretest have posttest scores that are closer to the mean.
D) participants' scores on a measure change over time as a result of natural aging processes.

Question 65

Third graders who scored in the bottom 20% of a standardized reading test were selected for a study on a new method for teaching reading. At the end of a 9-month test period, the children's scores had improved significantly more than could be expected by chance or by the passage of time. A potential problem with the design of this study is

A) maturation.
B) statistical regression.
C) selection.
D) mortality.

Question 66

Which of the following is true about statistical regression?

A) Extreme high scores are more likely to regress to the mean than are extreme low scores.
B) Researchers should not conduct research on people with extreme scores on a measure.
C) Adding a pretest is the best way to control for the problem of statistical regression.
D) Adding a control group allows the researcher to test whether statistical regression has occurred.

Question 67

A market researcher asks participants to attend a campus fashion show. At the show, Owen models a pair of Brand A jeans, Noah models a pair of Brand B jeans, and Nate models a pair of Brand C jeans. After each model leaves the stage, the participants rate his attractiveness on a 10-point scale. Which of the following factors could influence the results of the research?

A) the brand of jeans worn by the models
B) the physical attractiveness of the models
C) the order in which the models appeared
D) all of the above

Question 68

Which of the following statements about control groups is TRUE?

A) Control groups help eliminate time-related confounds as alternative explanations for research results.
B) Control groups are not needed when the research uses a pretest-posttest design.
C) Selection as a threat to internal validity is not an issue when the study includes a control group.
D) The use of a pretest requires that an additional control group be added.

Question 69

A researcher investigated the relationship between fraternity membership and drinking. He surveyed the members of 10 fraternities (250 men) and found that 80% agreed with the statement "I drink more than the average college student." The researcher concluded that fraternity membership contributes to excessive drinking. Based on the information given here, a problem with this study is that

A) the sample size was too small.
B) the measure of drinking was unreliable.
C) there is no control or comparison group.
D) fraternity membership and self-reported drinking are confounded.

Question 70

A selection bias occurs when

A) researchers use a different version of the dependent measure at the posttest than at the pretest.
B) research participants in the control condition of an experiment differ in some way from the participants in the experimental condition.
C) the dependent variable in a study is assessed by having research participants choose between performing two or more behaviors.
D) researchers choose one operational definition of a dependent variable over another rather than using multiple operational definitions.

Question 71

Selection biases are LEAST likely to be a problem when

A) researchers randomly assign participants to conditions of the independent variable.
B) researchers randomly assign one pre-existing group (such as children in classroom A) to be their experimental group and those in another pre-existing group (such as children in classroom B) to be their control group.
C) research participants are allowed to choose whether they want to be in the experimental group or the control group.
D) volunteer participants serve as members of the experimental group while non-volunteers from the same population serve as members of the control group.

Question 72

The term mortality refers to

A) a situation in which the instruments used to measure the dependent variable stop functioning part way through the research.
B) research participants failing to show up for their scheduled data-collection session.
C) research participants dropping out at the informed consent stage.
D) research participants dropping out of a study while it is being conducted.

Question 73

Professor Cloud conducts an experiment to test a well-established theory and finds non-significant results. Low theoretical validity is a threat to the validity of his findings if

A) he conducted an exact replication of a previous study that tested the theory.
B) the study excludes one of the experimental conditions proposed to be important by the theory.
C) he failed to include a pretest.
D) any of the above occurred.

Question 74

If the experimental conditions used in a study do not match those proposed by the theory, which of the following threats to validity could be operating?

A) selection bias
B) low theoretical validity
C) instrumentation
D) low statistical conclusion validity

Question 75

Low theoretical validity would NOT be a plausible alternative explanation when

A) an exact replication is being conducted.
B) a control group is included.
C) an appropriate pretest is administered.
D) any of the above occur.

Question 76

Which of the following does NOT have the potential for creating a selection bias?

A) use of volunteer research participants
B) mortality
C) history
D) use of pre-existing groups as experimental and control groups

Question 77

Dr. Anton is clinical psychologist who wants to test a new treatment for severe depression. He posts an announcement for the study online and includes people who score high on a depression inventory, are willing to attend several treatment sessions, and are willing to be randomly assigned to a treatment or control group. Which form of selection bias might be operating?

A) volunteer bias
B) mortality
C) cohort effects
D) instrumentation

Question 78

Professor Prieto teaches two sections of a diversity class. In the 10:00 class, he focuses on discussion; in the 3:00 class, he mainly lectures. At the end of the semester, he finds the students in the 3:00 class score higher on a measure of multicultural competence. Which of the following threats to validity could be operating in this study?

A) a history effect
B) a selection bias
C) mortality
D) a volunteer bias

Question 79

Mortality causes the most problems for a study when

A) there are equal drop-out rates in all conditions of a study.
B) the study is being conducted in a laboratory setting.
C) people in one condition of the study are more likely to withdraw from the research than are people in other conditions.
D) the study's research participants are older adults.

Question 80

A cancer researcher finds that more people drop out of his study if they are assigned to a condition with severe side effects. Data collection is now complete. What can he do to address the extent to which attrition affected his results?

A) Determine whether the people who completed the study differ from those who did not on characteristics relevant to the study.
B) Follow up with the people who dropped out to encourage them to participate in a separate study.
C) Analyze only the data collected after the attrition occurred.
D) do any of the above