Blog

Statistical Analysis: Controlling Confounding

QUESTION 1: The best way to deal with confounding variables is to statistically control for their effects after your experiment has been conducted.adjust your independent variables during your experiment when you discover the confounding.carefully plan how your independent variables are to be executed.

QUESTION 2: Confounding variables should be included in your experiments, because they clarify causal relationships among variables.can be eliminated by using factorial designs.can be substituted for independent variables.pose a serious threat to the internal validity of experiments.

QUESTION 3: Within-subjects designs cope with the problem of error variance bymatching subjects prior to random assignment to conditions.treating each individual as a separate unit in the statistical analysis of the data.using the same subjects in all treatment conditions.using different subjects in each treatment condition.

QUESTION 4: According to the text, a disadvantage of a within-subjects design is thatit is less powerful than a between-subjects design.more subjects are required than in a between-subjects design.each subject must spend more time in your experiment and may become fatigued.matching of subjects becomes cumbersome even in simple within-subjects designs.

QUESTION 5: Within-subjects designs are most useful whensubject differences contribute heavily to variation in the dependent variable.carryover effects are large.carryover effects vary unpredictably.subject differences do not significantly affect the value of the dependent variable.

QUESTION 6: In the ________ design, you chart changes in behavior as a function of some naturally occurring event.interrupted time seriestime seriestime samplesSolomon four-group

QUESTION 7: In the ________ design, treatments are administered repeatedly and alternated with periods of observation.nonequivalent control groupequivalent time samplesinterrupted time seriesSolomon four-group.

QUESTION 8: A drawback to quasi-experimental research studying the impact of naturally occurring events on behavior is thatsuch events rarely occur.it is difficult to predict when such events will occur.when natural events occur, they are usually too subtle to have a great impact on behavior.

QUESTION 9: Quasi-experimental research is used whenyou can manipulate independent variables only in the field.participants can be randomly assigned to treatment groups.random assignment of subjects is not possible.

QUESTION 10: A ________ design includes a single dependent variable.univariatemultivariatesingle-factormultifactor

QUESTION 11: The presence of error variancemakes it difficult to determine if your independent variable was effective.makes it easier to determine if your independent variable was effective.increases the internal validity of an experiment.increases the external validity of an experiment.

QUESTION 12: Participants in one group of an experiment on the effects of a drug on behavior receive a sugar pill rather than either an active drug or no drug. Those participants constitute acompound control group.placebo control group.blind control group.

QUESTION 13: In the nonequivalent control group design,a second group of subjects is included that is not exposed to your treatment.observations are made only after introduction of your treatment.observations are made only before introduction of your treatment.periods of observation are alternated with repeated introductions of your treatment.

QUESTION 14: As a way of handling the sensitizing effects of a pretest, Campbell and Stanley (1963) suggestusing a cohort-sequential design.eliminating the pretest entirely and running your study as a true experiment.allowing several weeks to elapse between the pretest and posttest.

QUESTION 15: One way to combine experimental and correlational designs is to includea between-subjects and a within-subjects factor in an experimental design.a quasi-independent variable in an experiment.multiple dependent variables in an experiment.

QUESTION 16: A design in which every possible order of treatments is represented once is theLatin square design.partially counterbalanced design.Solomon four-group design.completely counterbalanced design.

QUESTION 17: The most serious form of carryover effects occurs when your experimental treatment producesfatigue.habituation.learning.irreversible changes.

QUESTION 18: ________ involves assigning the various treatments of the experiment in a different order for different subjects.CounterbalancingReverse orderingEqualizing=

QUESTION 19: The matched groups design is most useful whena randomized two-group design would be too time-consuming.you suspect that some subject characteristic is correlated with your independent variable.you suspect that no relationship exists between subject characteristics and your dependent variable.you cannot manipulate independent variables.

QUESTION 20: To ensure internal validity of a pretest−posttest study, you mustinclude a large sample of subjects.include a control group that is not exposed to your treatment.conduct your research in your subjects natural environment.

QUESTION 21: A drawback to the randomized two-group design is that itprovides a limited amount of information about the effect of the independent variable.leaves too many sources of variance unaccounted for.is time-consuming to run.produces data that are difficult to analyze.

QUESTION 22: Dr. Jones conducts a randomized two-group experiment on the effect of a drug on learning. She finds that subjects who receive the drug do more poorly on a memory test than those who do not receive the drug. Which of the following would Dr. Jones be justified in concluding?There is a linear relationship between the independent variable and the dependent variable.There is a curvilinear relationship between the independent variable and the dependent variable.There is no relationship between the independent variable and the dependent variable.It would be dangerous to speculate on the nature of the function relating the independent and dependent variables.

QUESTION 23: If you vary the quantity of a drug administered to subjects by using 10 mg, 20 mg, and 40 mg, you are using a ________ design.parametricnonparametricmultifactor.

QUESTION 24: In a memory experiment, you expose one group of participants to a list of concrete words, another group to a list of abstract words, and a third group to a list of concrete and abstract words. This design is a ________ design.single-factor parametricsingle-factor nonparametrictwo-factor parametrictwo-factor nonparametric.

QUESTION 25: ________ occur(s) when a previous treatment alters the observed behavior in subsequent treatments.Error varianceCarryover effectsExperimenter biasSensitization effects.

QUESTION 26: In a factorial design, the main effects arethe effects of the most important independent variables on your dependent variable.the separate effects of each independent variable on your dependent variable.the changes in the effect of one independent variable over levels of a second.never interpreted because they yield no interesting information.

QUESTION 27: A factorial design always has more than onelevel of a single independent variable.dependent variable.independent variable.