Research Methods in Psychology 10th Edition by John Shaughnessy – Test Bank

$20.00

Description

INSTANT DOWNLOAD COMPLETE TEST BANK WITH ANSWERS

 

Research Methods in Psychology 10th Edition by John Shaughnessy – Test Bank

 

Sample  Questions

 

Chapter 06

Independent Groups Designs

 

 

Short Answer Questions

  1. (p. 181-182)Identify and briefly define the three primary types of control used in experiments.

The three primary types of experimental control are: manipulation, holding conditions constant, and balancing. Manipulation refers to controlling an independent variable, that is, by manipulating the levels of an independent variable. Holding conditions constant involves holding a potentially important factor (or factors) constant in an experiment to keep that factor from confounding the intended independent variable. Balancing is the type of control used to control individual differences variables; in the random groups design balancing is usually accomplished by randomly assigning participants to conditions of the experiment.

 

Level: Factual

  1. (p. 193-194)Clearly and concisely describe how a researcher could minimize the likelihood of selective subject loss in an experiment—be sure to include in your answer any risks that the researcher would be taking in trying to prevent the selective loss.

The researcher could use a pretest to screen out participants who would be likely to fail to complete the experiment successfully. The pretest could also be used to identify subjects in the group in which there was no subject loss who are comparable to the subjects who were lost. These comparable subjects in the no loss group could then be dropped to restore the comparability of the groups. The risk with both these procedures is a potential decrease in the generality of the findings because of testing subjects who are more homogeneous than the general population.

 

Level: Conceptual

 

  1. (p. 202-203)Briefly outline the steps involved in order to make an inference about the effect of an independent variable using null hypothesis significance testing; be sure to indicate the conditions in which an independent variable is judged to be nonsignificant and when it is statistically significant.

The first step is to assume that the null hypothesis is true; that is, there is no difference in the population means for the conditions of an experiment. Next, sample means for groups in an experiment are used to estimate the population means. The difference between these means is computed and an inferential statistic (e.g., t-test) is used to determine the probability of obtaining the difference under the null hypothesis. This probability value associated with the test statistic is compared to the level of significance, alpha (typically, p = .05). If the observed probability value is less than .05 we reject the null hypothesis of no difference and state that the difference between means is statistically significant. We infer that the independent variable produced a statistically significant effect on the dependent variable (assuming the experiment is internally valid). If the observed probability value is greater than .05 we do not reject the null hypothesis (i.e., no difference between the population means). We infer there is not sufficient evidence in the experiment to conclude that the independent variable produced a reliable effect.

 

Level: Factual

  1. (p. 203-204)Describe briefly how confidence intervals can be used to determine if the means for the two conditions of an experiment differ.

A confidence interval tells us the range of values that contains the true population mean with a specified level of confidence, usually .95. Confidence intervals can be used to determine if the means for the two conditions of an experiment differ. A confidence interval is computed for the mean in each condition. If the two confidence intervals do not overlap, then we can conclude that the population means for the two conditions differ and that the independent variable produced a reliable effect. If the intervals overlap slightly, however, we are uncertain about the difference between population means and postpone judgment about the independent variable. If the confidence intervals overlap such that the sample mean of one group falls within the confidence interval of another group, we conclude it’s likely the population means do not differ.

 

Level: Factual

 

  1. (p. 208)Describe how partial replications can be used to establish both experimental reliability and the external validity of research findings.

A partial replication involves changing some but not all of the original experiment. For instance, if an independent variable was shown to have an effect in an experiment testing only men, then a useful partial replication of this experiment would include the dimension of gender. In the partial replication both men and women would be tested while replicating all other conditions of the initial experiment. Reliability can be demonstrated in partial replications when the same effect of the independent variable is obtained across the two experiments (e.g., the independent variable has the same effect for men in both studies). Partial replications help to establish external validity by showing that a similar result occurs when slightly different experimental procedures are used, for example, with different participants and in a different setting. When similar results are obtained, we can say that the findings generalize across the participants (e.g., men and women) and setting.

 

Level: Conceptual

  1. (p. 209-210, 212)Explain how the procedures for conducting a matched groups design and a natural groups design differ from the random groups design.

A matched groups design often is used when the number of participants available for an experiment is too few for random assignment to balance effectively individual differences across conditions of the experiment (as in the random groups design). Instead, the researcher chooses a matching variable preferably the dependent variable measure or a similar variable and matches participants who have the same or similar score on the measure. Once matched, these pairs (triplets, etc.) of participants are randomly assigned to the conditions of the experiment. In a natural groups design, researchers select an individual differences variable for comparing groups of participants (e.g., male, female). Thus, rather than attempting to balance these individual differences (as in the random groups design), the researcher examines participants’ responses on the dependent variable as a function of the individual differences independent variable.

 

Level: Conceptual

 

  1. (p. 211-212)A researcher was interested in whether divorce and remarriage influence the extent to which children are sociable. At a nearby school, the researcher classified 5th-graders into one of three groups: intact parents (no divorce), divorced (single-parent families), and remarried parents. The researcher interviewed the 5th-graders and rated their sociability, and also asked them to complete a questionnaire that assessed their level of comfort in different social situations.What type of experimental design did this researcher use?

A natural groups design was used.

 

Level: Applied

  1. (p. 211-212)A researcher was interested in whether divorce and remarriage influence the extent to which children are sociable. At a nearby school, the researcher classified 5th-graders into one of three groups: intact parents (no divorce), divorced (single-parent families), and remarried parents. The researcher interviewed the 5th-graders and rated their sociability, and also asked them to complete a questionnaire that assessed their level of comfort in different social situations.Describe the independent variable in this study.

The independent variable was the family environment with three groups: intact parents, divorce/single-parent families, and remarried parents.

 

Level: Applied

 

  1. (p. 181)A researcher was interested in whether divorce and remarriage influence the extent to which children are sociable. At a nearby school, the researcher classified 5th-graders into one of three groups: intact parents (no divorce), divorced (single-parent families), and remarried parents. The researcher interviewed the 5th-graders and rated their sociability, and also asked them to complete a questionnaire that assessed their level of comfort in different social situations.Describe the dependent variable(s) in this study.

The dependent variables were the researcher’s ratings of 5th-graders’ sociability and these students’ questionnaire responses regarding their level of comfort in different social situations.

 

Level: Applied

  1. (p. 212)A researcher was interested in whether divorce and remarriage influence the extent to which children are sociable. At a nearby school, the researcher classified 5th-graders into one of three groups: intact parents (no divorce), divorced (single-parent families), and remarried parents. The researcher interviewed the 5th-graders and rated their sociability, and also asked them to complete a questionnaire that assessed their level of comfort in different social situations.Suppose that the researcher finds that the 5th graders in the divorce (single-parent) group are less sociable than are 5th graders in the intact and remarried groups. The researcher concludes that divorce and living in a single-parent home causes children to be less sociable, and that remarriage causes improvements in sociability. Do you accept this researcher’s conclusions? Why or why not?

This researcher incorrectly made a causal inference about the effects of divorce and remarriage based on a natural groups design. Natural groups designs provide correlational data regarding the relationship between an individual differences variable and a dependent variable; therefore, causal inferences cannot be made. These three groups may differ on a number of other variables (e.g., income, household stability, time since divorce) that may be related to children’s sociability. These variables are potential alternative explanations for the observed group differences, rather than the three groups the researcher formed.

 

Level: Applied

 

  1. (p. 182)A researcher tested whether exposure to images of very thin fashion models causes young women to be dissatisfied with their own body, compared to exposure to athletic body images or neutral (non-body) images. She randomly assigned 120 women from an introductory psychology course to one of three exposure conditions: very thin female images, athletic female images, or neutral images (e.g., household objects). Each condition had 10 images, projected individually on a large screen. The young women participated in small groups. Each image was displayed for 1 minute, for a total of 10 minutes of exposure. After viewing each image, participants wrote for 30 seconds a description of the image (the participants were led to believe their memory was being tested). After viewing the images, the women completed a questionnaire about satisfaction with their body. Negative scores indicate body dissatisfaction and positive scores indicate satisfaction with their body. The mean scores for each condition were as follows:What type of experimental design did this researcher use?

A random groups design was used.

 

Level: Applied

 

  1. (p. 182)A researcher tested whether exposure to images of very thin fashion models causes young women to be dissatisfied with their own body, compared to exposure to athletic body images or neutral (non-body) images. She randomly assigned 120 women from an introductory psychology course to one of three exposure conditions: very thin female images, athletic female images, or neutral images (e.g., household objects). Each condition had 10 images, projected individually on a large screen. The young women participated in small groups. Each image was displayed for 1 minute, for a total of 10 minutes of exposure. After viewing each image, participants wrote for 30 seconds a description of the image (the participants were led to believe their memory was being tested). After viewing the images, the women completed a questionnaire about satisfaction with their body. Negative scores indicate body dissatisfaction and positive scores indicate satisfaction with their body. The mean scores for each condition were as follows:Describe the independent variable in this study.

The independent variable was the type of image displayed with three exposure conditions: very thin female image, athletic female image, and neutral image.

 

Level: Applied

 

  1. (p. 181)A researcher tested whether exposure to images of very thin fashion models causes young women to be dissatisfied with their own body, compared to exposure to athletic body images or neutral (non-body) images. She randomly assigned 120 women from an introductory psychology course to one of three exposure conditions: very thin female images, athletic female images, or neutral images (e.g., household objects). Each condition had 10 images, projected individually on a large screen. The young women participated in small groups. Each image was displayed for 1 minute, for a total of 10 minutes of exposure. After viewing each image, participants wrote for 30 seconds a description of the image (the participants were led to believe their memory was being tested). After viewing the images, the women completed a questionnaire about satisfaction with their body. Negative scores indicate body dissatisfaction and positive scores indicate satisfaction with their body. The mean scores for each condition were as follows:Describe the dependent variable(s) in this study.

The dependent variable measured women’s satisfaction with their body using a questionnaire. Negative scores indicated dissatisfaction with their body and positive scores indicated satisfaction.

 

Level: Applied

 

  1. (p. 185-186)A researcher tested whether exposure to images of very thin fashion models causes young women to be dissatisfied with their own body, compared to exposure to athletic body images or neutral (non-body) images. She randomly assigned 120 women from an introductory psychology course to one of three exposure conditions: very thin female images, athletic female images, or neutral images (e.g., household objects). Each condition had 10 images, projected individually on a large screen. The young women participated in small groups. Each image was displayed for 1 minute, for a total of 10 minutes of exposure. After viewing each image, participants wrote for 30 seconds a description of the image (the participants were led to believe their memory was being tested). After viewing the images, the women completed a questionnaire about satisfaction with their body. Negative scores indicate body dissatisfaction and positive scores indicate satisfaction with their body. The mean scores for each condition were as follows:Someone unfamiliar with research methods criticizes the findings by arguing that women who viewed the very thin images probably were more dissatisfied with their bodies even before participating in the experiment. State whether you agree or disagree with this argument and explain your reasoning.

Disagree. Because the researcher randomly assigned sufficient numbers of participants to each condition of the experiment, we can be confident that individual differences were balanced across the conditions to create equivalent groups prior to the independent variable manipulation. Thus, prior to viewing the images, the three groups of women were likely equivalent, on average, in their level of body satisfaction.

 

Level: Applied

 

  1. (p. 203-204)A researcher tested whether exposure to images of very thin fashion models causes young women to be dissatisfied with their own body, compared to exposure to athletic body images or neutral (non-body) images. She randomly assigned 120 women from an introductory psychology course to one of three exposure conditions: very thin female images, athletic female images, or neutral images (e.g., household objects). Each condition had 10 images, projected individually on a large screen. The young women participated in small groups. Each image was displayed for 1 minute, for a total of 10 minutes of exposure. After viewing each image, participants wrote for 30 seconds a description of the image (the participants were led to believe their memory was being tested). After viewing the images, the women completed a questionnaire about satisfaction with their body. Negative scores indicate body dissatisfaction and positive scores indicate satisfaction with their body. The mean scores for each condition were as follows:Suppose the .95 confidence interval (CI) for the very thin condition is -1.75 to -2.50; the CI for the Athletic condition is -.75 to -1.25; and the CI for the neutral condition is 0 to .50. What claim would you make based on the estimates of the population means for the three groups in the experiment based on a comparison of these confidence intervals?

The confidence intervals for the three conditions do not overlap; based on this we can conclude that the population means for the three groups differ. Body dissatisfaction was greatest in the very thin image condition, followed by the athletic image condition. The least amount of body dissatisfaction was in the neutral image condition.

 

Level: Applied

 

Multiple Choice Questions

  1. (p. 179)When researchers use the multimethod approach they can reach comparable conclusions about a research question after using different methods to study it. Our confidence in these conclusions increases and the conclusions are said to have
    A. convergent validity.
    B. concurrent validity.
    C. multiple validity.
    D. correlational validity.

 

Level: Factual

  1. (p. 179)One major purpose of conducting experiments is to decide whether a treatment or program effectively changes behavior. A second major purpose for doing experiments is to provide
    A. definitive answers to theoretical questions.
    B. simple and relatively quick ways of testing and revising hypotheses.
    C. methods to obtain results that confirm our hypothesis.
    D. an empirical test of hypothesis derived from theories.

 

Level: Factual

  1. (p. 181)The factors that researchers control or manipulate in order to determine their effect on behavior are called the
    A. intervention variables.
    B. dependent variables.
    C. independent variables.
    D. confounding variables.

 

Level: Factual

 

  1. (p. 181)In a study that investigates the effects of two different doses of a drug on memory performance, memory performance represent the ____ variable and doses of the drug represent the ____ variable.
    A. correlational; confounding
    B. experimental; control
    C. dependent; independent
    D. independent; dependent

 

Level: Applied

  1. (p. 181)When the three requirements for causal inference are met, an experiment is said to be
    A. balanced.
    B. internally valid.
    C. an independent groups design.
    D. held constant.

 

Level: Factual

  1. (p. 181)Two control techniques that allow researchers to rule out alternative explanations for an outcome are balancing and
    A. holding conditions constant.
    B. establishing a time-order relationship.
    C. validation.
    D. establishing a covariation.

 

Level: Factual

 

  1. (p. 181)A researcher has manipulated only one independent variable at two levels, has held constant as many other variables as possible and has balanced individual differences by using random assignment. The researcher is likely to be able to claim that the independent variable caused the observed changes in the dependent variable because the experiment is
    A. externally valid.
    B. internally valid.
    C. a natural groups design.
    D. statistical.

 

Level: Factual

  1. (p. 182)The goal of a random groups design experiment is to establish the independent variable as the cause of any differences in the dependent variable. The logic of accomplishing this involves
    A. beginning with comparable groups, treating them differently, and ending with differences among groups.
    B. beginning with comparable groups, treating them the same, and ending with no differences among groups.
    C. beginning with noncomparable groups, treating them differently, and ending with differences among groups.
    D. beginning with noncomparable groups, treating them the same, and ending with no differences among groups.

 

Level: Conceptual

  1. (p. 182)The most common solution to the problem of forming comparable groups in the random groups design is
    A. random-digit dialing.
    B. random assignment.
    C. random selection.
    D. matching participants on the dependent variable.

 

Level: Factual

 

  1. (p. 185)Which of the following arises when the independent variable of interest and an unintended independent variable are allowed to covary?
    A. contamination effect
    B. decrease in external validity
    C. illusory correlation
    D. confounding

 

Level: Factual

  1. (p. 185)A researcher examined participants’ memory following emotionally shocking events by manipulating whether participants viewed a violent or a nonviolent version of a video. The video in the two conditions was identical except for two seconds in which the violence variable was manipulated. That the video was identical except for the manipulation represents the control technique of
    A. balancing the individual differences across the groups of the experiment.
    B. holding conditions constant in the experiment.
    C. a time-order relationship between the independent and dependent variables.
    D. confounding the independent variable.

 

Level: Applied

  1. (p. 185)Individual differences among participants in a random groups design are controlled by
    A. eliminating the individual differences from the experiment.
    B. ignoring the individual differences because they cannot confound the experiment.
    C. balancing the individual differences across the conditions of the experiment.
    D. holding the individual differences constant in the experiment.

 

Level: Factual

 

  1. (p. 185-186)Random assignment to conditions works to balance participants’ individual differences across conditions of the experiment by
    A. matching individuals on key variables.
    B. asking individuals to participate in each condition of the experiment.
    C. forming groups based on an individual differences variable the researcher selects.
    D. generating groups of participants that are equivalent, on average.

 

Level: Factual

  1. (p. 185-186)Researchers manipulated exposure to body images by randomly assigning young girls to one of three conditions. They observed that young girls’ body dissatisfaction was greatest after viewing very thin images of a Barbie doll compared to normal-weight images (“Emme” doll) or neutral images. The alternative explanation that girls in the Barbie condition might have weighed more prior to exposure to body images can be ruled out because the researchers
    A. held constant the amount of time girls looked at images.
    B. manipulated the type of exposure with three conditions.
    C. randomly assigned girls to the different exposure conditions.
    D. all of these

 

Level: Conceptual

  1. (p. 185-186)A researcher interested in young girls’ exposure to thin images of Barbie dolls compared to normal-weight images (“Emme” doll) wanted to make sure girls in the two conditions (Barbie, Emme) did not differ in the number of Barbie dolls they own at home. The best way to do this would be to
    A. hold conditions constant by making sure each girl in the experiment only had three Barbie dolls.
    B. randomly assign girls to the two conditions of the experiment.
    C. give each girl in the experiment a Barbie doll and an Emme doll.
    D. none of these

 

Level: Applied

 

  1. (p. 188)The procedure for using block randomization in an independent groups experiment with three conditions (A, B, C) and 60 subjects is to
    A. generate twenty random orders of the conditions (e.g., ACB, BAC) and assign subjects one block at a time (i.e., 3 subjects in the first block, 3 in the second block, and so on).
    B. generate one random order of conditions (e.g., ACB) and test the first 20 subjects in condition A, the second 20 subjects in condition C, and the last 20 subjects in condition B.
    C. select intact groups of 20 subjects each and randomly assign the groups to condition A, B, or C.
    D. generate sixty random orders of the conditions, one for each subject.

 

Level: Factual

  1. (p. 188-189)In addition to creating groups of equal size, an advantage of block randomization is that it
    A. decreases the number of participants needed in each condition.
    B. averages the effect of the dependent variable across the conditions of the experiment.
    C. holds constant the time in which an experiment is run.
    D. balances potential confoundings that occur during the time in which an experiment is run.

 

Level: Factual

  1. (p. 189-190)A researcher randomly assigns one classroom to a new teaching method and a second classroom to the control condition (the regular teaching method). This researcher faces the potential problem of confounding due to
    A. experimenter effects.
    B. selective subject loss.
    C. intact groups.
    D. extraneous variables.

 

Level: Applied

 

  1. (p. 189-190)An instructor randomly assigns two sections of his course to an experimental condition (new teaching method) or control condition (regular teaching method). Students in the 9:30 section receive the experimental treatment and students in the 1:30 section participate in the control condition. At the end of the semester the instructor’s tests reveal that students in the 9:30 section had significantly higher test scores than students in the 1:30 section. Based on this summary, we can state that
    A. the results are uninterpretable because of intact groups.
    B. the findings have external validity across two times of day.
    C. the experiment has internal validity.
    D. all of these

 

Level: Applied

  1. (p. 190)In order to conduct an experiment more efficiently, a researcher tests groups of participants in several small groups. The number of participants in each group differs because of participants’ availability at different times. The different number of participants in each group represents a potential
    A. intact group problem.
    B. selective subject loss problem.
    C. experimenter effect.
    D. extraneous variable problem.

 

Level: Applied

  1. (p. 190-191)Four students plan to conduct an experiment with four conditions (A, B, C, and D) during one semester. Each student will serve as an experimenter. The best way to control for the extraneous variable of having four experimenters test participants in the study is to
    A. make sure each experimenter tests only one of the conditions.
    B. have the first experimenter test a random order of conditions during the first two weeks of the semester, the second experimenter test a second random order during the next two weeks, and so on.
    C. have each experimenter test randomized blocks of conditions.
    D. have each experimenter test condition A during the first two weeks of the semester, condition B during the next two weeks, and so on.

 

Level: Applied

 

  1. (p. 191)When participants begin an experiment but fail to complete it, the internal validity of the experiment can be threatened. Which of the following types of subject loss poses the most serious threat to internal validity?
    A. the loss occurs because of an error by the experimenter.
    B. the loss leads to different numbers of participants in the groups in the experiment.
    C. the loss occurs because of equipment failure.
    D. the loss occurs because of some characteristic of the participant that is related to the outcome of the study.

 

Level: Factual

  1. (p. 191)Subject loss (attrition) poses a problem for a random groups design because
    A. extraneous variables are more likely to affect subjects who drop out.
    B. group equivalence established at the beginning of the experiment may be lost.
    C. participants may respond to demand characteristics.
    D. participants may change their natural group designation.

 

Level: Factual

  1. (p. 193)One preventive step researchers can take when they suspect selective subject loss might occur in their experiment involves using a pretest to screen out participants who may drop out of the study. The disadvantage of this method is decreased
    A. external validity.
    B. internal validity.
    C. statistical significance.
    D. all of these

 

Level: Factual

 

  1. (p. 194)If participants know they have been given alcohol in an experiment, they may expect certain effects such as giddiness or relaxation. The general term for such cues that guide participants’ behavior in a study is
    A. double-blind effects.
    B. self-inflicted characteristics.
    C. expectation effects.
    D. demand characteristics.

 

Level: Factual

  1. (p. 194)When observers have biases regarding what they expect participants’ behavior to be like in an experiment, the observations may be influenced by
    A. placebos.
    B. demand characteristics.
    C. experimenter effects.
    D. selective subject loss.

 

Level: Factual

  1. (p. 194-195)Placebo control groups and double-blind procedures are typically used to control for
    A. individual differences variables and matching variables.
    B. demand characteristics and experimenter effects.
    C. selective subject loss and mechanical subject loss.
    D. manipulated variables and extraneous variables.

 

Level: Factual

  1. (p. 196)Which of the following is the best way to determine whether the differences in means obtained in an experiment are reliable?
    A. Replicate the experiment.
    B. Use confidence intervals to compare the means.
    C. Perform an inferential statistics test (null hypothesis significance testing).
    D. Find the difference between two sample means.

 

Level: Factual

 

  1. (p. 196)The three steps of data analysis are
    A. find any outliers, compute confidence intervals, and do null hypothesis significance testing.
    B. check the data, summarize the data, and confirm what the data reveal.
    C. find Type I and Type II errors, compute means, and compute inferential statistics.
    D. calculate the standard deviation, find the effect size, and do null hypothesis significance testing.

 

Level: Factual

  1. (p. 198)After checking the data for errors and outliers, the next step in analyzing the data from a research study is to use
    A. confidence intervals.
    B. meta-analysis.
    C. descriptive statistics.
    D. inferential statistics and null hypothesis significance testing.

 

Level: Factual

  1. (p. 198-199)Which of the following is a measure of the strength of the relationship between the independent and dependent variables that is independent of sample size?
    A. t-test
    B. standard deviation
    C. mean difference score
    D. effect size

 

Level: Factual

 

  1. (p. 199)A researcher conducted an experiment in which participants played either a violent or a nonviolent video game. After the game, the researcher measured hostile cognitions. The effect size, Cohen’s d, for the difference in mean hostile cognition between the violent and nonviolent conditions was .83. Based on this, it is possible to state that the video game independent variable had a _______ effect on hostile cognition in this experiment.
    A. nonsignificant
    B. small
    C. medium
    D. large

 

Level: Applied

  1. (p. 198-199)Which of the following is not an advantage of using measures of effect size?
    A. Measures of effect size provide information about the strength of the relationship between an independent variable and a dependent variable.
    B. Measures of effect size provide the best information about whether the effect of the independent variable is statistically significant.
    C. Measures of effect size can provide an estimate of the overall effect size for an independent variable by averaging effect sizes across a series of experiments.
    D. Measures of effect size can be used to make quantitative comparisons of the different outcomes found in a series of experiments involving the same independent variable.

 

Level: Factual

  1. (p. 199)The statistical technique that is used to analyze the results of several independent experiments is called
    A. F-test.
    B. central tendency.
    C. meta-analysis.
    D. Cohen’s d analysis.

 

Level: Factual

 

  1. (p. 201)Two statistical methods that researchers use to determine whether an independent variable has a reliable effect on a dependent variable are
    A. null hypothesis significance testing and confidence intervals.
    B. statistical significance and odds-likelihood analysis.
    C. analysis of error variation and effect size.
    D. Cohen’s d and alpha.

 

Level: Factual

  1. (p. 201)The nonsystematic (random) variation due to differences among subjects within each group is called
    A. spurious variation.
    B. nonsignificant variation.
    C. inferential variation.
    D. error variation.

 

Level: Factual

  1. (p. 202)A statistically significant outcome is an outcome that
    A. leads us to reject the null hypothesis.
    B. assures the internal validity of the experiment.
    C. assures that practical applications of the outcome will be successful.
    D. all of these

 

Level: Factual

  1. (p. 202)A statistically significant outcome is an outcome that
    A. has a large likelihood of occurring if the null hypothesis is true.
    B. has a small likelihood of occurring if the null hypothesis is true.
    C. always has a large effect size.
    D. has neither Type I nor Type II errors.

 

Level: Factual

 

  1. (p. 202-203)A researcher computes an inferential statistic to test the difference between mean scores for an experimental group and a control group. The probability of the obtained statistical value for the t-test is .025, which is less than the alpha level of significance (p < .05). The researcher should
    A. reduce the alpha level of significance to .025 to form a definite conclusion.
    B. accept the null hypothesis of no difference between the experimental and control groups because the probability is so small; thus, the independent variable had no effect.
    C. reject the null hypothesis of no difference between the experimental and control groups because the probability is so small; thus, the independent variable had a reliable effect on the dependent variable.
    D. neither accept nor reject the null hypothesis of no difference because the probability value of .025 is significantly different than .05.

 

Level: Applied

  1. (p. 203)We can be confident that the population means differ for two conditions of an experiment when the confidence intervals for the two sample means
    A. are of different size.
    B. have different standard deviations.
    C. overlap.
    D. do not overlap.

 

Level: Factual

  1. (p. 203-204)A researcher computes a .95 confidence interval for an experimental group to be 3.0 B 7.0, and computes a .95 confidence interval for a control group to be 0.0 B 4.0. Based on these confidence intervals, the researcher can state that
    A. the results for the effect of the independent variable on the dependent variable are inconclusive.
    B. the independent variable definitely influenced participants’ scores on the dependent variable because the confidence interval for the experimental group is higher than that that of the control group.
    C. the independent variable definitely did not affect participants’ scores because the confidence interval for the control group includes zero.
    D. the population value for the experimental group is .95.

 

Level: Applied

 

  1. (p. 204)Our ability to draw appropriate conclusions based on the results of an experiment depends most of all on the
    A. external validity of the experiment.
    B. internal validity of the findings.
    C. amount of variation in the experiment.
    D. statistical significance of the findings.

 

Level: Factual

  1. (p. 204-205)The problem of Type I and Type II errors occurs because
    A. decision making based on inferential statistics depends on probabilities.
    B. researchers rarely conduct internally valid experiments.
    C. the null hypothesis is difficult to define.
    D. all of these

 

Level: Conceptual

  1. (p. 204)A Type I error occurs when the null hypothesis is really true and we claim
    A. the independent variable did not have an effect on behavior.
    B. the independent variable did have an effect on behavior.
    C. the inferential test was not statistically significant.
    D. that the results are not statistically significant.

 

Level: Factual

  1. (p. 205)When we conclude that an experiment does not provide sufficient evidence to reject the null hypothesis when, in fact, an independent variable does produce an effect, the inferential statistics problem is called a
    A. Type I error.
    B. Type II error.
    C. no-confidence interval.
    D. probability error.

 

Level: Factual

 

  1. (p. 181, 205)________ refers to the researcher’s ability to make causal inferences regarding an experimental outcome and _________ refers to the researcher’s ability to generalize the findings beyond the scope of the specific experiment.
    A. Reliability; validity
    B. Replication; confounding
    C. Internal validity; external validity
    D. External validity; internal validity

 

Level: Factual

  1. (p. 205)When the findings of experiments testing the effectiveness and safety of drugs on animals are replicated in clinical trials with people, the original findings of the animal research are shown to have
    A. external validity.
    B. internal validity.
    C. statistical significance.
    D. sensitivity.

 

Level: Conceptual

  1. (p. 205)In what situation is it argued that external validity is irrelevant?
    A. when psychologists seek to describe real-world settings based on laboratory research.
    B. when research is conducted with college student samples.
    C. when the goal of an experiment is to observe the most typical case found in nature.
    D. when the purpose of the experiment is to test a specific hypothesis derived from a theory.

 

Level: Factual

 

  1. (p. 208)A researcher read a research report indicating that a certain medication was found to be effective when tested on men. The researcher plans to do an experiment testing the effectiveness of the same medication, but in his experiment both men and women will be tested. The researcher is planning to do a
    A. partial replication of the experiment to test the reliability and external validity of the finding from the original experiment.
    B. replication of the experiment to test the internal validity of the original experiment.
    C. replication of the experiment to test the sensitivity of the original experiment.
    D. partial replication of the experiment to test the statistical power of the original experiment.

 

Level: Applied

  1. (p. 208)Researchers found a relationship between insults and aggressive behavior for 5-year olds. If they want to test the external validity of the conceptual relationship between insults and aggressive behavior for a sample of 35-year olds, they should
    A. test the findings in the real world rather than in a laboratory setting.
    B. use the same measure of aggression as was used with the 5-year olds.
    C. use age-appropriate insults and measures of aggression for the 35-year olds.
    D. use the same insults as was used with the 5-year olds.

 

Level: Conceptual

  1. (p. 209)A cognitive psychologist wants to do an experiment to test a variable that may influence the memory of elderly people. She has the following constraints: only 12 people are available as possible participants, and the independent variable she plans to manipulate requires the use of a separate group for each of the two conditions. The psychologist has a reliable and valid pretest she can use to evaluate participants’ memory prior to the manipulation. Which design is this psychologist likely to use?
    A. random groups design
    B. natural groups design
    C. matched groups design
    D. placebo control design

 

Level: Applied

 

  1. (p. 210)When conducting an experiment using the matched groups design, the preferred matching variable is
    A. completely different from the dependent variable.
    B. an inexpensive test that participants could complete quickly.
    C. a measure with limited reliability and validity.
    D. the same measure that will be used as the dependent variable.

 

Level: Factual

  1. (p. 209-210)Which of the following conditions would lead you to recommend against the use of a matched groups design and in favor of a random groups design?
    A. a small number of participants from a heterogeneous population is available
    B. a large number of participants from a homogeneous population is available
    C. a separate group is required for each level of the independent variable of interest
    D. a reliable and valid matching variable is available

 

Level: Factual

  1. (p. 211)A researcher plans to use a matched groups design to assess a treatment for increasing body weight of premature infants. The independent variable has two conditions, treatment and control. Eight pairs of premature infants are matched according to their body weight prior to manipulating the independent variable. In order to control for other potentially relevant characteristics of the participants, the researcher should
    A. identify a matching variable uncorrelated with body weight.
    B. assign the lower weight infant in each pair to the treatment condition.
    C. randomly assign infants in each pair to the treatment or control group.
    D. none of these

 

Level: Applied

 

  1. (p. 211-212)Which of the following is an example of an individual differences (subject) variable?
    A. individuals randomly assigned to one of three different treatments in a diabetes study
    B. an independent variable involving three levels of nicotine in a smoking cessation program
    C. a law enforcement training program compared over a 4-week or an 8-week period
    D. individuals classified as introverted or extraverted in a study of alcohol use

 

Level: Applied

  1. (p. 212)To differentiate experiments involving individual differences (subject) variables and those involving manipulated independent variables, those experiments involving groups selected based on individual differences (subject) variables are called
    A. selected subjects designs.
    B. matched groups designs.
    C. random groups designs.
    D. natural groups designs.

 

Level: Factual

  1. (p. 212)The natural groups design represents an illustration of the general research approach that is called
    A. correlational research.
    B. descriptive research.
    C. experimental research.
    D. observational research.

 

Level: Factual

  1. (p. 212)Which of the following is the most critical problem in drawing causal inferences based on the natural groups design?
    A. establishing covariation
    B. correlating participants’ characteristics and their performance
    C. eliminating plausible alternative causes for the obtained relationship
    D. using the natural groups variable to make predictions for the dependent variable

 

Level: Factual

Chapter 07

Repeated Measures Designs

 

 

Short Answer Questions

  1. (p. 220-221)Cite two reasons why a researcher might choose to use a repeated measures design.

The four reasons cited in the text are that repeated measures designs are ideal when only a small number of subjects is available; that repeated measures designs are more convenient and efficient; that repeated measures designs are generally more sensitive; and that repeated measures designs can be used to study changes in participants’ behavior over time or across stimuli.

 

Level: Factual

  1. (p. 224)Provide brief answers to the following question regarding the repeated measures designs.What distinguishes the complete repeated measures design from the incomplete repeated measures design?

In the complete repeated measures design each subject experiences each condition at least twice such that the results are interpretable for each subject (i.e., practice effects are balanced within subjects). In the incomplete repeated measures design each subject experiences each condition exactly once so the results are interpretable only by averaging across subjects (i.e., conditions of the independent variable are confounded with practice effects within subjects).

 

Level: Factual

 

  1. (p. 224-227, 230-232)Provide brief answers to the following question regarding the repeated measures designs.What techniques are used in the complete and incomplete repeated measures designs to balance practice effects?

Block randomization and ABBA counterbalancing are used to balance practice effects in the complete repeated measures design, and some researchers use simple randomization in which new random orders are created for each subject. All possible orders or selected specific orders (Latin Square and random starting order with rotation) are used to balance practice effects in the incomplete repeated measures design.

 

Level: Factual

  1. (p. 227-228)Explain why ABBA counterbalancing would be ineffective in balancing practice effects in the complete repeated measures design when the practice effects involve an abrupt initial change followed by little change thereafter.

ABBA counterbalancing is most effective when practice effects are linear, meaning that each trial adds the same amount of practice and each condition accumulates the same amount of practice across the trials. When practice effects are nonlinear (as in this case), practice effects will contribute less to performance in the conditions administered following the initial trial. Practice effects will not be balanced across the conditions.

 

Level: Conceptual

  1. (p. 230-231)What are one advantage and one disadvantage of using all possible orders to balance practice effects in the incomplete repeated measures design?

The advantage of using all possible orders, the preferred method for balancing practice effects in the incomplete repeated measures design, is that every condition occurs in every ordinal position and precedes and follows every condition in every ordinal position. The disadvantage of using all possible orders is that it is generally limited to experiments with four or fewer conditions. [With four conditions, there are 24 possible orders, requiring at least one subject assigned to each.]

 

Level: Factual

 

  1. (p. 237-238)Define differential transfer and describe one method a researcher could use to detect whether differential transfer is a problem in a repeated measures experiment.

Differential transfer occurs when performance in one condition differs depending on the condition that precedes it. That is, the effects of one condition persist and influence performance differently in subsequent conditions. The best way to determine whether differential transfer is a problem is to test the independent variable in two separate experiments, a repeated measures experiment and a random groups experiment. If the results of the two experiments differ, differential transfer may be a problem. A second method is to examine the results for the first ordinal position in an incomplete design, which represents an random groups design (if participants are randomly assigned to different orders). If the results differ when compared to the repeated measures design as a whole, then differential transfer may be a problem.

 

Level: Factual

  1. (p. 220, 222-224)Identify the differences between the random groups design and the repeated measures design in terms of how the independent variable is manipulated and what potential confounding variable must be balanced in each design. What method(s) are available for balancing in each design?

In a random groups design the researcher manipulates the independent variable such that different groups of participants experience each level of the independent variable. Individual differences among participants across the conditions of the experiment must be balanced; random assignment to conditions is the preferred method for balancing individual differences variables. In the repeated measures design the independent variable is manipulated such that the same group of participants experiences every condition of the experiment. Practice effects due to repeated measurement must be counterbalanced in repeated measures designs. In the complete repeated measures design, in which each participants experiences every condition more than once, ABBA counterbalancing or block randomization may be used to balance practice effects across conditions. In the incomplete design, in which each participant experiences every condition exactly once, all possible orders or selected orders may be used to balance practice effects.

 

Level: Factual

 

  1. (p. 220)A psychologist wishes to have participants rate two different sets of inkblots (labeled A and B) for their “emotionality.” The inkblots differ in their use of color. Ten inkblots in set A are created in shades of gray; ten inkblots in set B use shades of red, blue, and yellow. The psychologist hypothesizes that color will influence participants’ ratings of emotionality. Each participant will rate all twenty inkblots.What is the independent variable in this experiment?

The independent variable is the color of the inkblot with two levels, gray (A) and color (B).

 

Level: Applied

  1. (p. 220)A psychologist wishes to have participants rate two different sets of inkblots (labeled A and B) for their “emotionality.” The inkblots differ in their use of color. Ten inkblots in set A are created in shades of gray; ten inkblots in set B use shades of red, blue, and yellow. The psychologist hypothesizes that color will influence participants’ ratings of emotionality. Each participant will rate all twenty inkblots.What is the dependent variable in this experiment?

The dependent variable is participants’ ratings of emotionality of the inkblots.

 

Level: Applied

  1. (p. 220)A psychologist wishes to have participants rate two different sets of inkblots (labeled A and B) for their “emotionality.” The inkblots differ in their use of color. Ten inkblots in set A are created in shades of gray; ten inkblots in set B use shades of red, blue, and yellow. The psychologist hypothesizes that color will influence participants’ ratings of emotionality. Each participant will rate all twenty inkblots.If the psychologist would like to have 50 ratings for each inkblot, how many participants will the psychologist need to recruit for the experiment?

Because this is a repeated measures design, the psychologist needs to recruit 50 participants.

 

Level: Applied

 

  1. (p. 224-227)A psychologist wishes to have participants rate two different sets of inkblots (labeled A and B) for their “emotionality.” The inkblots differ in their use of color. Ten inkblots in set A are created in shades of gray; ten inkblots in set B use shades of red, blue, and yellow. The psychologist hypothesizes that color will influence participants’ ratings of emotionality. Each participant will rate all twenty inkblots.What potential confounding variable must be balanced in this research? Explain what balancing procedure would be best for this experiment.

Practice effects need to be balanced in this repeated measures experiment. Participants should not rate the set of A inkblots followed by the set of B inkblots (or vice versa). Because this experiment uses a complete repeated measures design, ABBA counterbalancing or block randomization is appropriate. Block randomization is preferable when participants might anticipate the presentation of particular conditions; these anticipation effects could influence participants’ responses of the emotionality of the inkblot. In block randomization each block would contain one inkblot from set A and one inkblot from set B. Ten random sequences of A and B inkblots would counterbalance practice effects.

 

Level: Applied

 

Multiple Choice Questions

  1. (p. 220)In an independent groups design, a separate group of people serves as a control group. In the repeated measures design,
    A. there is no control.
    B. participants serve as their own controls.
    C. all participants participate in one condition of the experiment.
    D. test-retest reliability is the main goal of the research.

 

Level: Factual

 

  1. (p. 220)The need to balance practice effects in the repeated measures design is analogous to the need to balance ________ in the independent groups design.
    A. order effects
    B. reactivity
    C. individual differences
    D. subject loss

 

Level: Conceptual

  1. (p. 220-221)Which of the following is not one of the reasons researchers choose to use the repeated measures design?
    A. Repeated measures designs are generally less sensitive than are independent groups designs.
    B. Repeated measures designs are more convenient and efficient.
    C. Repeated measures designs require fewer subjects.
    D. Repeated measures designs are needed when the experimental procedures require participants to compare two or more stimuli.

 

Level: Factual

  1. (p. 221)A sensitive experiment refers to the ability to detect the effect of the independent variable even if the effect is _________.
    A. uneven
    B. small
    C. large
    D. very large

 

Level: Factual

 

  1. (p. 221)In general, participants in a repeated measures design will vary within themselves less over conditions of the experiment than participants in a random groups design will vary from other participants across conditions. This means there is likely to be ________ in repeated measures designs compared to random groups designs.
    A. less error variation
    B. less sensitivity
    C. more bias
    D. more individual differences

 

Level: Conceptual

  1. (p. 222)The difference between repeated measurements to establish test-retest reliability and experiments using repeated measures designs is that in experiments with repeated measures designs
    A. the reliability of measurement is not important.
    B. a correlation coefficient would be used to determine consistency across measurements.
    C. measurements occur only twice.
    D. measurements are compared for different conditions of an independent variable.

 

Level: Conceptual

  1. (p. 223)There can be no confounding by individual differences variables in the repeated measures designs because
    A. intact groups of participants are assigned to the different conditions of the experiment.
    B. individual differences variables are held constant.
    C. the same participants are tested in all conditions.
    D. participants are tested in only one condition of the experiment.

 

Level: Factual

 

  1. (p. 223)When using a repeated measures design, researchers must control the potential threat to internal validity referred to as
    A. individual differences variables.
    B. sensitivity.
    C. incomplete orders.
    D. practice effects.

 

Level: Factual

  1. (p. 223, 228-229)A researcher compares students’ performance using a new learning strategy to their performance using the old strategy. Students’ performance is first tested with the old strategy, followed by the new strategy. The results indicate that students perform better with the new strategy. These results
    A. show that the order of the two learning strategies does not matter.
    B. are uninterpretable due to the confounding with practice effects.
    C. indicate that teachers should use the new strategy.
    D. all of these

 

Level: Applied

  1. (p. 224)Counterbalancing is an important research technique because it is used to
    A. average practice effects across conditions of a repeated measures design.
    B. average practice effects across conditions of an independent groups design.
    C. average individual differences variables across conditions of a repeated measures design.
    D. all of these

 

Level: Conceptual

  1. (p. 224)Balancing the order of conditions in a repeated measures design
    A. equates practice effects with those found in random groups designs.
    B. eliminates practice effects in the conditions of the experiment.
    C. averages practice effects across the conditions of the experiment.
    D. balances individual differences variables across the conditions of the experiment.

 

Level: Factual

 

  1. (p. 224)A marketing researcher wants to test the effectiveness of three different types of advertisement. Participants will respond to 50 examples of each type of advertisement so that the researcher can determine each person’s preference. What type of design is this researcher using?
    A. conditional repeated measures design
    B. partial repeated measures design
    C. complete repeated measures design
    D. incomplete repeated measures design

 

Level: Applied

  1. (p. 224)A researcher plans to conduct a complete repeated measures design in which the independent variable is manipulated using four conditions that vary the level of violence depicted in a photograph (none, low, medium, high). There are 10 photographs in each of the four conditions. Each participant will rate each photo once. If the researcher wishes to have 80 ratings for each photo, how many participants should the researcher recruit?
    A. 2
    B. 10
    C. 40
    D. 80

 

Level: Applied

  1. (p. 224)A researcher designs an experiment with six conditions but each condition takes only a very short amount of time to administer. The researcher has the opportunity, therefore, to administer many trials of each condition in the experiment. Which of the following techniques is likely to be most effective for balancing practice effects?
    A. ABBA counterbalancing
    B. block randomization
    C. stratified randomization
    D. all possible orders of conditions

 

Level: Applied

 

  1. (p. 226)An advantage of block randomization in the complete repeated measures design experiment is that
    A. each condition of the experiment is represented an equal number of times at all stages of practice.
    B. all trials of one condition are presented first, followed by all trials of a second condition, etc., making statistical computations easier.
    C. one condition is assigned to one “block” of participants, a second condition is assigned to a second block of participants, etc.
    D. stages of practice are randomly assigned to the different blocks of conditions.

 

Level: Factual

  1. (p. 227-228)A student is considering doing a complete repeated measures design experiment involving motor skills. The student’s advisor has told him that people show a large initial improvement on the task followed by slow steady improvement after this initial change. The student must choose a technique for balancing practice effects. Which technique should the student not use?
    A. ABBA counterbalancing
    B. block randomization
    C. Latin Square
    D. all possible orders

 

Level: Applied

  1. (p. 228)When a participant in an experiment that involves the complete repeated measures design develops expectations about which condition should occur next in the sequence, the methodological problem that occurs is called
    A. reactivity effects.
    B. sequential effects.
    C. differential transfer.
    D. anticipation effects.

 

Level: Factual

 

  1. (p. 228-229)In the incomplete repeated measures design, the levels of the independent variable for each participant are perfectly confounded with the
    A. characteristics of that particular participant.
    B. order in which the levels are presented.
    C. individual differences variables in the experiment.
    D. characteristics of the experimental task.

 

Level: Factual

  1. (p. 229)Practice effects in the incomplete repeated measures design are balanced by
    A. ensuring first that the results for each participant are balanced across conditions.
    B. subtracting the results for the different groups of participants in the experiment.
    C. averaging across the results for all participants.
    D. first obtaining a summary score within each condition for each participant.

 

Level: Factual

  1. (p. 230)Each of the three techniques that are used to balance practice effects in the incomplete repeated measures design conforms to a general rule that can be stated as
    A. each condition must appear in only one ordinal position.
    B. each condition must appear in the first ordinal position equally often.
    C. each condition must appear in each ordinal position at least twice.
    D. each condition must appear in each ordinal position equally often.

 

Level: Factual

 

  1. (p. 230-231)A health psychologist conducts an experiment to test the effectiveness of four techniques for helping a person to relax. The psychologist has a limited number of people available to participate in the experiment and each relaxation technique takes some time to complete. The psychologist has decided, therefore, to use the incomplete repeated measures design with all possible orders to balance practice effects. What is the minimum number of participants the psychologist will need for this experiment?
    A. 4
    B. 12
    C. 24
    D. 48

 

Level: Applied

  1. (p. 232)When selected orders of conditions (Latin Square or random starting order with rotation) are used to balance practice effects in the incomplete repeated measures design,
    A. the number of selected orders must be exactly equal to the number of conditions in the experiment.
    B. the number of selected orders will always be equal to some multiple of the number of conditions in the experiment.
    C. the number of selected orders will always be equal to one less than the number of conditions in the experiment.
    D. there is no restriction on the number of possible orders needed to balance practice effects in the incomplete repeated measures design.

 

Level: Conceptual

  1. (p. 232)Compared to random starting order with rotation, an advantage of the Latin Square technique for selecting orders in the incomplete repeated measures design is that in the Latin Square
    A. each condition is presented to each subject several times.
    B. practice effects are balanced within each individual.
    C. a random order of conditions is generated for each participant.
    D. each condition precedes and follows each other condition exactly once.

 

Level: Factual

 

  1. (p. 234)The additional step needed when analyzing the results in a complete repeated measures design is to
    A. compute the mean (or median) score for each participant for each condition of the experiment.
    B. multiply participants’ score in each condition by the order in which the condition appeared.
    C. add the participants’ scores across all conditions and divide by the number of conditions.
    D. subtract the mean of all conditions from each participant’s score.

 

Level: Factual

  1. (p. 235-236)Repeated measures designs are more sensitive than random groups designs because the systematic variation due to individual differences is _________ the statistical analyses.
    A. eliminated from
    B. added to
    C. averaged into
    D. a critical part of

 

Level: Factual

  1. (p. 237)The problem of _____ in a repeated measures design occurs when the effects of a condition persist or carryover to affect performance in the subsequent conditions.
    A. sensitivity
    B. anticipation effects
    C. differential transfer
    D. counterbalancing

 

Level: Factual

 

  1. (p. 237)An incomplete repeated measures design experiment was done to test people’s ability to solve problems that varied in difficulty at three levels (easy, moderate, and hard). The researcher used all possible orders to balance practice effects. Results indicated that performance on the hard problems was better when they were preceded by a moderate difficulty problem than when they were preceded by an easy problem. Which of the following is a possible explanation of this finding?
    A. anticipation effects
    B. nonlinear practice effects
    C. progressive error
    D. differential transfer

 

Level: Applied

  1. (p. 238)The best solution when differential transfer is likely in an experiment is to
    A. use selected orders in an incomplete repeated measures design.
    B. conduct an independent groups design.
    C. choose a different independent variable to investigate.
    D. use block randomization in a complete repeated measures design.

 

Level: Factual