Statistics And Research Design - Domain Quiz Flashcards

Question

In a leptokurtic distribution: Select one: a. scores are distributed more evenly throughout the distribution than they are in a normal (mesokurtic) distribution. b. a larger proportion of scores are located in the middle of the distribution than they are in a normal (mesokurtic) distribution. c. most scores are in the negative side of the distribution and the tail is extended on the positive side. d. most scores are in the positive side of the distribution and the tail is extended on the negative side.

Answer 1

For the exam, you want to know the difference between leptokurtic and platykurtic distributions and between positively and negatively skewed distributions. These are described and illustrated in the Statistics and Research Design chapter of the written study materials. Answer B is correct: A leptokurtic distribution is more "peaked" than the normal distribution -- i.e., more scores are located in the middle of the distribution. Answer A is incorrect: A platykurtic distribution is flatter than the normal distribution -- i.e., scores are more evenly distributed throughout the distribution. Answer C is incorrect: In a positively skewed distribution, most scores are in the negative side and the tail is extended on the positive side by the presence of a few high scores. Answer D is incorrect: In a negatively skewed distribution, most scores are in the positive side and the tail is extended on the negative side by the presence of a few low scores. The correct answer is: a larger proportion of scores are located in the middle of the distribution than they are in a normal (mesokurtic) distribution.

Answer 2

For the exam, you want to be familiar with the relationships between the mean, median, and mode in skewed distributions. These relationships are illustrated in figures included in the Statistics and Research Design chapter of the written study materials. Answer A is correct: In a negatively skewed distribution, most scores are "piled up" in the positive tail while a relatively few scores are in the negative tail. Because the mean is more sensitive to the magnitude of all scores, it is affected more by the low scores in the negative tail and is lower in value than the median or the mode. Therefore, the median is greater than the mean in a negatively skewed distribution. Answer B is incorrect: The median is less than the mean in a positively skewed distribution. Answer C is incorrect: In a negatively skewed distribution, the median is greater than the mean and the mode is greater than the median. Answer D is incorrect: The median is greater than the mode in a positively skewed distribution. The correct answer is: the median is greater than the mean.

Answer 3

The four scales of measurement differ with regard to the type of information they provide and the kinds of mathematical operations they permit. Answer C is correct: The ratio scale provides the greatest information. Because it has an absolute zero point, you can conclude that one score is twice as much as another -- for example, that a person who weighs 100 pounds weighs twice as much as a person who weighs 50 pounds. Answer A is incorrect: An ordinal scale divides observations into categories and provides information on the order of those categories. Ranking people in terms of weight is an example of an ordinal scale. When people are ranked in terms of weight, you know that the person with a higher ranking weighs more than a person with a lower ranking but you don't know the actual difference in their weights. Answer B is incorrect: An interval scale provides more information than an ordinal scale -- i.e., when using an interval scale, you know that the difference between scores of 40 and 45 is the same as the difference between scores of 80 and 85. However, because an interval scale does not have an absolute zero point, you can't conclude that 80 is twice as much as 40. Answer D is incorrect: A nominal scale only divides variables into categories and provides no information about the order of those categories (e.g., gender, DSM diagnosis, place of birth). The correct answer is: ratio

Answer 4

For the exam, you want to be familiar with all of the multivariate techniques listed in the responses so that you can answer questions like this one. These techniques and others that you want to be familiar with are described in the written study materials. Answer C is correct: Discriminant function analysis is used to predict or estimate a person's status on a single nominal criterion from two or more predictors. Answer A is incorrect: Multiple regression is used to predict status on a single continuous criterion from scores on two or more predictors. Answer B is incorrect: Path analysis is a causal modeling technique that is an extension of multiple regression and is used to test a theory about the causal relationships among a set of variables. Answer D is incorrect: Canonical correlation is the appropriate technique when two or more predictors will be used to predict status on two or more continuous criteria. The correct answer is: discriminant function analysis

Answer 5

Eta is one of several correlation coefficients that you want to be familiar with for the licensing exam. Answer B is correct: Eta is used to determine the degree of association between two continuous variables when their relationship is nonlinear. The correct answer is: determine the correlation for two variables that have a nonlinear relationship.

Answer 6

Shared variability is one way to interpret the correlation coefficient for two variables. Answer A is correct: A measure of shared variability is obtained by squaring the correlation coefficient: -.40 squared is .16. Therefore, 16% of variability in Test A scores is shared with (or accounted for by) variability in Test B scores. The correct answer is: 16

Answer 7

For the exam, you want to be familiar with the use of each of the correlation coefficients listed in the answers to this question. Answer A is correct: The point biserial coefficient is the appropriate coefficient when one variable represents a true dichotomy and the other is measured on an interval or ratio (continuous) scale. Although gender is not binary, questions on the EPPP may still refer to gender as a dichotomous variable, especially within the statistics questions. Answer B is incorrect: The Pearson r is the appropriate correlation coefficient when both variables are measured on an interval or ratio scale. Answer C is incorrect: The phi coefficient is used when both variables are true dichotomies. Answer D is incorrect: The contingency coefficient is used to determine the degree of association between two nominal variables. The correct answer is: point biserial

Answer 8

Path analysis is a causal modeling technique that is used to test the veracity of a causal theory or model. Answer C is correct: One restriction when using path analysis is that the paths must all involve a one-way causal flow. Answer A is incorrect: See explanation for Answer C. Answer B is incorrect: See explanation for Answer C. Answer D is incorrect: LISREL is more complex than path analysis and not only includes two-way paths but also takes into account both observed variables and the latent traits those variables are believed to measure. The correct answer is: confirming a model involving one-way causal flow between a set of observed variables.

Answer 9

Cohen's d is a measure of effect size. It indicates the difference between the means of two groups in terms of standard deviations. Additional information about methods for calculating an effect size is provided in the Statistics and Research Design chapter of the written study materials. a. CORRECT A Cohen's d of .50 indicates that one group obtained a mean that is one-half standard deviation higher than the mean obtained by the other group. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: there was a difference of one-half standard deviation between the means of the two groups.

Answer 10

Split-plot is another name for "mixed"; and a researcher would use a split-plot ANOVA when the study uses a mixed design - i.e., it includes both between-groups and within-subjects comparisons. a. Incorrect This answer describes a study that includes two between-groups variables. b. Incorrect This answer describes a nested design. c. Incorrect This answer describes a counterbalanced design with a single independent variable. d. CORRECT This is the only answer that describes a mixed design: Variable A is a between-groups variable, while Variable B is a within-subjects variable. Additional information about between-groups and within-subjects designs is provided in the Statistics and Research Design chapter of the written study materials. The correct answer is: a study in which each participant received only one level of Variable A but all levels of Variable B

Answer 11

LISREL, a structural equation (causal) modeling technique, is used to test causal hypotheses about relationships among measured variables and the latent traits those variables are believed to measure. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT In contrast to path analysis models, which predict the causal relationships among measured attributes only, LISREL models incorporate both measured attributes and latent traits. Additional information about LISREL and path analysis is provided in the Statistics and Research Design chapter of the written study materials. d. Incorrect See explanation for response c. The correct answer is: includes both measured (observed) attributes and latent traits.

Answer 12

For the exam, you want to be familiar with the appropriate use of all of the tests listed in the answers to this question. Additional information on these tests is provided in the Statistics and Research Design chapter of the written study materials. a. Incorrect The Mann-Whitney U test is used to analyze rank-ordered data from two independent groups. b. Incorrect The Kruskal-Wallis test is used to analyze rank-ordered data from two or more independent groups. c. CORRECT Knowing that the full name of this test is the Wilcoxon matched-pairs signed-ranks test would have helped you identify it as the correct response. The Wilcoxon is the appropriate test for rank-ordered data that have been obtained from correlated groups (e.g., groups that have been formed by matching participants on the basis of their scores on a pretest or extraneous variable and then randomly assigning members of each matched pair to one of two groups). d. Incorrect The ANCOVA is a type of analysis of variance that allows the researcher to statistically remove the effects of an extraneous variable. Like other forms of the analysis of variance, it is used when data represent an interval or ratio scale of measurement. The correct answer is: Wilcoxon

Answer 13

The types of multiple regression include simple, hierarchical, and stepwise. As its name implies, stepwise regression involves adding (or subtracting) predictors to the multiple regression equation one at a time. a. Incorrect This describes hierarchical multiple regression. b. CORRECT When using stepwise multiple regression, each predictor is retained in the equation as long as it contributes significantly to the total amount of variability in the criterion that is explained by the combined predictors. The proportion of explained variability is measured by R-squared. c. Incorrect This sounds like the basis for determining if predictors have incremental validity, which may be a consideration in multiple regression. However, this is not the best answer to this question. d. Incorrect This refers to factor analysis. The correct answer is: the magnitude of R-squared.

Answer 14

The term "factor" in factorial design refers to independent variables. a. Incorrect See explanation for response b. b. CORRECT A factorial research design is any design that includes two or more "factors" (independent variables). c. Incorrect See explanation for response b. d. Incorrect See explanation for response b. The correct answer is: always includes two or more independent variables.

Answer 15

The level of significance (alpha) determines the location of the boundary between the regions of likely and unlikely values in the sampling distribution. When results are significant at the chosen level of significance, this means that the results are in the region of unlikely values and that the null hypothesis should be rejected. a. CORRECT Significance at the .01 level means that there is a 1% chance that the obtained value (e.g., the mean or the difference between means) could have occurred by chance alone given the value specified in the null hypothesis. In other words, there is a 1% probability that the null hypothesis will be incorrectly rejected (that a Type I error will be made). b. Incorrect The probability of incorrectly retaining the null hypothesis (i.e., of retaining a false null hypothesis) is equal to beta, not alpha. c. Incorrect See explanation for response a. d. Incorrect Because the results are statistically significant, the psychologist will reject (not retain) the null hypothesis. The correct answer is: there is a 1% chance that she will incorrectly reject the null hypothesis.

Answer 16

The study described in this question has one independent variable (gender) and one dependent variable (Holland's six occupational themes), and the dependent variable is measured on a nominal scale. a. Incorrect The single-sample chi-square test is appropriate for descriptive studies that include a single nominal variable. b. CORRECT The multiple-sample chi-square test is appropriate for studies that include two or more variables and the data to be analyzed represent a nominal scale. c. Incorrect The one-way ANOVA is appropriate for studies assessing the impact of a single independent variable on a single dependent variable that is measured on an interval or ratio scale. d. Incorrect The factorial ANOVA is appropriate for studies assessing the impact of two or more independent variables on a single dependent variable that is measured on an interval or ratio scale. The correct answer is: multiple-sample chi-square

Answer 17

For the exam, you want to be familiar with all four techniques listed in the answers to this question, which are described in the Statistics and Research Design chapter of the written study materials. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT Of the techniques listed in the answers, canonical correlation is the only one that is appropriate when two or more continuous (interval or ratio) predictors will be used to predict status on two or more continuous criteria. The correct answer is: canonical correlation

Answer 18

The parametric tests are fairly "robust" with regard to a violation of the assumptions that should be met in order to use them, and their robustness is increased under certain conditions. b. CORRECT When groups are equal in size, a violation of the assumption of homogeneity of variances will probably not substantially affect the results of the t-test. c. Incorrect One of the assumptions of the t-test is that the distribution of scores is normally shaped. d. Incorrect Use of the t-test requires that the data represent an interval or ratio scale. The correct answer is: not likely to be a problem as long as the two groups have the same number of participants.

Answer 19

To identify the correct answer to this question, you have to recognize that the question is describing a nonlinear relationship between anxiety and math achievement and know that an assumption underlying the use of the Pearson r is that the relationship between variables is linear. a. CORRECT In this situation, the relationship between the variables is nonlinear (i.e., the assumption of linearity has been violated). As a result, the Pearson r will underestimate the relationship between test anxiety scores and math achievement scores. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: underestimate the relationship between the variables.

Answer 20

Answer A is correct: Counterbalancing involves administering the treatments (different levels of the IV) in different orders to different groups of participants. It helps control multiple treatment interference (also known as carry-over or order effects) that may result when multiple levels of the independent variable(s) are administered to the same participants. The correct answer is: control multiple treatment interference.

Answer 21

This question is asking about a factor that threatens a study's external - but not internal - validity. Additional information on threats to external and internal validity is provided in the Statistics and Research Design chapter of the written study materials. a. CORRECT Anything that threatens a study's internal validity will also threaten its external validity, but this question is asking about a factor that threatens external validity only. Of the threats listed, only multiple treatment interference is listed by Campbell and Stanley as a direct threat to external validity. b. Incorrect Statistical regression is a threat to internal validity. c. Incorrect Maturation is a threat to internal validity. d. Incorrect Attrition is also a threat to a study's internal validity. The correct answer is: multiple treatment interference

Answer 22

An assumption underlying most correlation coefficients is that the relationship between variables is linear. a. CORRECT Eta is also known as the eta coefficient and the correlation ratio and can be used when the relationship between two continuous variables is nonlinear. b. Incorrect The phi coefficient is used when both variables are true dichotomies. c. Incorrect Spearman rho is used when both variables are ranks. d. Incorrect The Pearson r is used when both variables are continuous and their relationship is linear. The correct answer is: eta coefficient

Answer 23

Even if you're unfamiliar with structural equation modeling (the appropriate technique for the purpose described in the question), you may have been able to identify the correct answer to this question through the process of elimination. a. Incorrect Meta-analysis is used to evaluate an intervention by combining the results of a number of research studies. b. Incorrect The multitrait-multimethod matrix is used to evaluate convergent and divergent validity. c. Incorrect Discriminant function analysis is used to classify people into criterion groups based on their scores or status on two or more predictors. d. CORRECT Structural equation model is used to explore or confirm hypothesized relationships between both measured and latent variables. The correct answer is: structural equation modeling

Answer 24

In the situation described in this question, the researcher is studying a phenomenon (therapist-client interactions) under conditions that only resemble or approximate actual clinical conditions. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT In clinical research, an analogue study is a study in which the conditions are in some way an analogue (approximation) of actual clinical practice. The correct answer is: analogue

Answer 25

The researcher in this experiment is trying to establish a causal relationship between two variables - stress level and processing speed. a. Incorrect Face validity focuses on whether a test looks like it measures what it is intended to measure. b. Incorrect Construct validity is established when a test measures the intended hypothetical trait. c. CORRECT Internal validity focuses on the causal relationship between independent and dependent variables. d. Incorrect External validity focuses on the generalizability of one study to other conditions, individuals, etc. The correct answer is: internal validity

Answer 26

Knowing that an extraneous variable is a source of systematic error would have helped you narrow your choice of a correct response to answers a and c. a. CORRECT An extraneous variable is a variable that is not relevant to the purpose of the study but has a systematic effect on the dependent variable and, as a result, can make it difficult to ascertain the effects of the independent variable on the dependent variable. Additional information on extraneous variables is provided in the Statistics and Research Design chapter of the written study materials. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: has a systematic effect on the study's dependent variable.

Answer 27

The mean square within (MSW) is the denominator of the F-ratio and, as its name implies, is a measure of within-group variability. a. CORRECT Within-group variability is a measure of error; and decreasing within-group variability decreases error and the magnitude of the denominator of the F-ratio. Additional information on the F-ratio is provided in the Statistics and Research Design chapter of the written study materials. b. Incorrect This would increase the size of the denominator of the F-ratio. c. Incorrect MSW is a measure of within-group variability and isn't affected by between-group variability. d. Incorrect See explanation above. The correct answer is: decreasing within-group variability.

Answer 28

The correlation for two variables is zero when scores on one variable are not related to scores on the other variable. a. CORRECT This is the only answer of those given that describes a lack of correlation between X and Y. When the range of Y scores at each value of X equals the total range of Y scores, the scattergram for X and Y is rectangular in shape, indicating a correlation of (or near) zero. b. Incorrect This describes homoscedasticity which is good as long as the range of Y scores at each value of X is not equal to the total range of Y scores. c. Incorrect This condition does not imply a zero correlation between X and Y. d. Incorrect This describes heteroscedasticity. Although heteroscedasticity can produce an inaccurate estimate of the degree of association between variables, it does not necessarily result in a correlation coefficient of zero. The correct answer is: the range of Y scores at each value of X is equal to the total range of Y scores.

Answer 29

Sampling error is due to chance variations that can be expected in samples that have been randomly drawn from the population. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT Sampling error refers to the discrepancies between sample values and corresponding population values (parameters) that are due to chance factors in the selection process. Note that, by definition, sampling error is random, not systematic. d. Incorrect Error due to the unreliability of the measuring instruments is referred to as measurement error, not sampling error. The correct answer is: the tendency of sample statistics to vary from population parameters due to the effects of random error.

Answer 30

For the exam, you want to be familiar with the effects of adding or subtracting a constant to each score in a distribution or multiplying or dividing each score by a constant. a. CORRECT Adding a constant to each score in the distribution affects the distribution's mean (the new mean is the original mean plus the constant). However, adding a constant does not affect the variability of scores - i.e., it does not change the range, standard deviation, or variance. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: affect the mean of the distribution but not its standard deviation or variance.

Answer 31

The multiple-baseline design is a type of single-subject design. a. Incorrect This describes the reversal (withdrawal) design. b. Incorrect This describes the changing criterion design. c. Incorrect This does not describe the multiple-baseline design. d. CORRECT When using the multiple-baseline design a treatment is sequentially applied to different baselines - i.e., to different behaviors, to the same behavior in different settings, or to the same behavior performed by different individuals. Unlike a "reversal" design (which is described by answer a, the multiple-baseline design does not require withdrawal of the treatment during the course of the study. The correct answer is: sequentially apply a treatment to different behaviors or participants.

Answer 32

Be careful not to confuse percentile ranks and percent correct scores. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT Percent correct scores represent a ratio scale of measurement - i.e., they have the properties of order, equal intervals, and an absolute 0 point. Consequently, it is possible to conclude that someone who obtains a score of 50% got twice as many items correct as someone whose score is 25%. d. Incorrect See explanation for response c. The correct answer is: ratio

Answer 33

For the exam, you want to be familiar with the normal curve and distributions that deviate from the normal curve. These are described in the Statistics and Research Design chapter of the written study materials. a. CORRECT "Platy" means flat or broad, and a platykurtic distribution is flatter than the normal distribution. b. Incorrect A leptokurtic distribution is more "peaked" (pointed) than a normal distribution. c. Incorrect Mesokurtic is another term for a normal distribution. d. Incorrect Exokurtic is a "made up" term. The correct answer is: platykurtic.

Answer 34

For the licensing exam, you want to have the least squares criterion associated with regression analysis. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT The least squares criterion is used to locate the regression line in a scatterplot so that the amount of error in prediction is minimized when using the regression line or its equation to predict criterion scores. The correct answer is: identifying the line of best fit in a regression analysis.

Answer 35

To identify the correct answer to this question, you have to be familiar with the areas under the normal curve. a. Incorrect See explanation for response b. b. CORRECT When a distribution is normally shaped, approximately 34% of scores fall between the mean and the score that is one standard deviation above the mean. c. Incorrect See explanation for response b. d. Incorrect In a normal distribution, approximately 68% of scores fall between the scores that are one standard deviation below and above the mean. The correct answer is: 34

Answer 36

Knowing that MSB stands for "mean square between" may have helped you identify the correct answer to this question. (Additional information on the F-ratio is provided in the Statistics and Research Design chapter of the written study materials.) a. Incorrect Since it is not possible to measure treatment effects only, they are estimated by dividing a measure of treatment plus error (MSB) by a measure of error only (MSW). b. CORRECT "Mean square within" is a measure of variability within treatment groups and provides an estimate of the effects of error. c. Incorrect "Mean square between" is the numerator of the F-ratio and provides a measure of variability due to a combination of treatment plus error. d. Incorrect See explanation above. The correct answer is: error only.

Answer 37

Pretest sensitization (also known as an interaction between testing and treatment) is a potential threat to a study's external validity whenever a study includes administering a pretest. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT Pretest sensitization occurs when the administration of a pretest "sensitizes" participants to the purpose of the research study and, as a result, alters their reaction to the treatment. In this case, the results of the study cannot be generalized to situations in which a pretest is not administered prior to delivery of the treatment. The correct answer is: taking the pretest changes how the participants react to the treatment.

Answer 38

Two standards are used to judge the validity of a research study - internal and external validity. a. Incorrect The ability to manipulate the independent variable has a direct impact on a study's internal validity, which refers to the ability to determine if the independent variable has affected the dependent variable. b. Incorrect Random assignment is important for internal validity. c. Incorrect The ability to determine if there is a causal relationship between variables depends on a study's internal (not external) validity. d. CORRECT External validity refers to the ability to generalize the results of a study to groups, settings, and conditions other than those used in the study. Random selection of participants from the population is one factor that helps ensure adequate external validity in terms of generalizability to other individuals. The correct answer is: generalize the results of the study to other individuals, settings, and conditions.

Answer 39

The correlation coefficient is squared to determine the amount of shared variability. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT A correlation coefficient of .60 indicates that 36% of variability is shared variability, while a correlation of .30 indicates that 9% of variability is shared variability. Since 36% is four times larger than 9%, this means that a coefficient of .60 indicates four times as much shared variability as does a coefficient of .30. d. Incorrect See explanation for response c. The correct answer is: four

Answer 40

As its name suggests, "autocorrelation" is a problem whenever observations are not independent. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT Of the research designs listed, only the time-series design involves non-independent observations -- i.e., it involves measuring the same people repeatedly before and after the intervention is applied. The correct answer is: time-series

Answer 41

Experimenter expectancy refers to the bias that results when the experimenter's expectations about the outcomes of the study influence the actual outcomes (e.g., as the result of cues the experimenter inadvertently provides participants). a. Incorrect In a single-blind study, the participants do not know what group they are in. This procedure would not eliminate experimenter expectancies. b. CORRECT When a double-blind procedure is used, research participants and the experimenter are "blind" to the experimental conditions - i.e., they do not know what group participants are in. Because the researcher does not know whether individual participants are in the experimental or control group, this reduces the likelihood that he/she will provide cues to participants that could bias the results of the study. c. Incorrect Random selection of participants to groups would not eliminate experimenter expectancy. d. Incorrect Random assignment of participants to groups would not eliminate experimenter expectancy. The correct answer is: double-blind

Answer 42

In inferential statistics, there are two types of decision errors - rejecting a true null hypothesis and retaining a false null hypothesis. The Statistics and Research Design chapter of the written study materials includes a decision outcome table that you might find useful for clarifying the decisions that are made about the null hypothesis. a. Incorrect Rejecting a true null hypothesis is referred to as a Type I error. b. Incorrect Rejecting a false null hypothesis is a correct decision. c. Incorrect Retaining a true null hypothesis is a correct decision. d. CORRECT Retaining a false null hypothesis is referred to as a Type II error. The correct answer is: retains the null hypothesis when it is false.

Answer 43

Main effects refer to the effects of one independent variable on the dependent variable when considered alone, while interaction effects refer to the effects of one independent variable at different levels of another independent variable. a. Incorrect See explanation for response b. b. CORRECT When the interaction is significant, this means that the effects of one independent variable differ for different levels of another independent variable. Thus, it is not possible to conclude that the independent variable has consistent main effects. For example, a study might find that, overall, Teaching Method #1 is superior to Teaching Method #2 (i.e., there is a main effect of teaching method). However, there might also be an interaction between teaching method and level of self-esteem - for example, Teaching Method #1 might be more effective for students with high and moderate self-esteem, while Teaching Method #2 is more effective for students with low self-esteem. In this situation, the main effect of teaching method would have to be interpreted with caution. c. Incorrect See explanation for response b. d. Incorrect It is possible to have any combination of significant main and/or interaction effects. The correct answer is: interpret the main effects with caution since the interaction is significant.

Answer 44

This study includes one independent variable that has two levels (program and no program); the data to be analyzed represent a ratio scale of measurement (total number of correct items); and matching was used to control an extraneous variable (IQ score). a. Incorrect The t-test for a single sample is used to compare the mean derived from a single sample to a known population mean. b. Incorrect The t-test for independent samples is used compare the means of two unrelated samples. The study described in this question involves comparing means obtained from dependent (related) samples. c. CORRECT The t-test for dependent (related) samples is used when the data are collected from two correlated groups, which is the case when matching has been used to assign participants to groups. d. Incorrect A two-way ANOVA is used when a study involves two independent variables. This study has one independent variable (educational program) with two levels. The correct answer is: t-test for dependent (related) samples

Answer 45

This question describes a type of reactivity in which the behavior of participants is affected by cues in the experimental setting that indicate what behaviors are expected of them. a. Incorrect Treatment diffusion (also known as treatment spillover) occurs when participants in one group inadvertently receive (or learn about) the intervention that has been applied to participants in another group. b. CORRECT Cues in the environment that inform participants about the purpose of the study or indicate what behavior is required of them are referred to as demand characteristics. c. Incorrect The confirmation bias is the tendency to pay attention to information that confirms one's current beliefs or preconceptions. d. Incorrect The placebo effect refers to an improvement in physical or mental health that occurs in response to a "fake" treatment (e.g., a sugar pill), apparently because of the individual's expectation that the treatment will be effective. The correct answer is: demand characteristics

Answer 46

All of the techniques listed in the answers can be described as multivariate techniques, but only one is used to test causal hypotheses or models. a. CORRECT Path analysis and other structural equation modeling techniques are used to determine if sample data fit a predefined model that indicates the direction of the causal links between variables. b. Incorrect Logistic regression is a multivariate technique used to identify the combination of predictors that most accurately categorizes individuals into criterion groups. c. Incorrect Cluster analysis is used to "cluster" or categorize individuals or items. d. Incorrect The multitrait-multimethod matrix is used to evaluate a measure's construct validity and provides information on the test's convergent and divergent validity. The correct answer is: path analysis

Answer 47

A normal distribution of scores is symmetrical and bell-shaped. a. Incorrect See explanation for response b. b. CORRECT By definition, a normal distribution is bell-shaped and, consequently, is also known as the "bell curve." c. Incorrect See explanation for response b. d. Incorrect See explanation for response b. The correct answer is: bell-shaped.

Answer 48

The various single-subject research designs share two characteristics: First, each design includes at least one baseline (no treatment) phase and one treatment phase; and, second, the dependent variable is measured at regular intervals during each phase. a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect See explanation for response d. d. CORRECT This answer is the only one of those given that accurately describes measurement of the dependent variable when using a single-subject research design. The correct answer is: at regular intervals during the baseline and the treatment phases of the study.

Answer 49

Counterbalancing is a within-subjects design that involves changing the order in which each treatment is administered to different groups of participants. The goal of counterbalancing is to use every possible sequence of treatments with an equal number of participants for each sequence. a. Incorrect The changing criterion design is a type of single case design that consists of a series of phases in which a different behavioral criterion is set for each phase. If the behavior reaches the criterion level for each phase, the treatment is considered to be effective. b. CORRECT When the number of participants does not permit a completely counterbalanced design a researcher may use partial counterbalancing. The Latin square design is a type of partial counterbalancing that helps determine exactly what sequences of a treatment will be administered to different groups of participants. c. Incorrect The multiple baseline design is a type of single case design that involves sequentially applying a treatment across participants, settings, or behaviors. d. Incorrect The Solomon four-group design is used to evaluate the effects of pretesting on a study's internal and external validity. The correct answer is: Latin square design

Answer 50

This study is designed to answer the question, "What are the effects of math aptitude and type of instruction on math achievement test scores?" a. Incorrect Type of instruction is only one of two independent variables in this study, so this isn't the best answer. b. CORRECT This study has two independent variables (type of instruction and math aptitude) and one dependent variable (math achievement test scores). c. Incorrect See explanation for response b. d. Incorrect See explanation for response b. The correct answer is: type of instruction and math aptitude.

Answer 51

The MANOVA is used to simultaneously assess the effects of one or more independent variables on two or more dependent variables. a. Incorrect The analysis of covariance (ANCOVA) is used to statistically remove the effects of an extraneous variable on the dependent variable. b. Incorrect This is a distracter and doesn"t make sense in the context of MANOVA. c. CORRECT If three separate ANOVAs were conducted at a level of significance of .05 each, the "experimentwise error rate" (probability of making a Type I error) would be increased to approximately .15. By conducting a single MANOVA at the .05 level, the experimentwise error rate remains at .05. d. Incorrect If the MANOVA yields significant results, the researcher will have to conduct post hoc tests (probably three separate ANOVAs) to determine if the independent variable affected all or only some of the dependent variables. The correct answer is: she wants to reduce the probability of making a Type I error.

Answer 52

Ordinarily, the best measure of central tendency for interval or ratio data is the arithmetic mean. However, in the situation described in this question, the mean would not be the most accurate measure because the distribution of data includes three scores that are estimates (rather than actual measures) of the value of interest a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect Because the mode provides less information about a distribution than the median does (i.e., the mode only indicates the most frequently occurring score in the distribution), the median would be a better measure of central tendency. d. CORRECT The mean is affected by the magnitude of every score in a distribution, but the median is not. Consequently, in this situation, the median would not be affected by the magnitude of the three missing scores and, therefore, would be a more accurate measure of central tendency. The correct answer is: median

Answer 53

Power refers to the ability to detect (reject) a false null hypothesis. Factors that affect the size of power are described in the Statistics and Research Design chapter of the written study materials. a. Incorrect Beta is the probability of retaining a false null hypothesis (i.e., of making a Type II error) - and the larger the size of beta, the greater the probability of retaining (not rejecting) the false null hypothesis. (Also note that researchers do not have direct control over the magnitude of beta.) b. CORRECT The ability to reject the null hypothesis is increased by increasing alpha. When the null hypothesis is actually false, increasing the size of alpha also increases power. c. Incorrect Parametric tests are more powerful than are nonparametric tests. d. Incorrect Decreasing the sample size would decrease power. The correct answer is: increase alpha from .01 to .05.

Answer 54

An ABAB design is a single-subject design that involves collecting baseline data, administering the independent variable, removing the independent variable, and then readministering the independent variable. a. Incorrect Diffusion is a threat to a study's validity when participants in one group (often a no-treatment control group) benefit from the intervention administered to another group - i.e., when participants in the control group inadvertently learn about the treatment or are accidentally exposed to it. b. Incorrect Attrition is a threat to a study's internal validity. It occurs when participants who drop out of one group differ in a systematic way from those who drop out of another group and this difference affects the study's results. c. Incorrect Instrumentation is a threat to internal validity and refers to any change in tests or other measuring devices administered to participants during the course of the study that confounds the study's results. An ABAB design would not help control instrumentation effects. d. CORRECT The use of an ABAB design helps determine whether or not an observed change in the dependent variable is due to the independent variable or to an external event (history). Specifically, if the participant exhibits the same change after the independent variable is administered the second time that he/she exhibited after it was administered the first time, that change can be assumed to have been caused by the independent variable rather than by history. The correct answer is: history

Answer 55

There are several versions of the t-test, but each is used to compare two means. Additional information on the t-test is provided in the Statistics and Research Design chapter of the written study materials. a. Incorrect The t-test for a single sample is used to compare an obtained sample mean to a known population mean. b. CORRECT The t-test for dependent (related) samples is the appropriate test when the two means to be compared are related, which would be the case when they have been obtained from the same group of participants. c. Incorrect The t-test for independent samples is used to compare the means obtained by two unrelated groups. d. Incorrect There is no "t-test for multiple samples." The correct answer is: t-test for dependent samples

Answer 56

Statistical regression refers to the tendency of extreme scores to "regress to the mean" on retesting - i.e., for examinees who obtained extremely high or extremely low scores on a measure to obtain scores closer to the mean when retested. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT When participants are chosen for inclusion in a study because of their extreme scores on the pretest, it is likely that their scores will "regress to the mean" on the post-test, regardless of the effects of the treatment. This can make it difficult to ascertain the effects of the independent variable on the dependent variable. d. Incorrect See explanation for response c. The correct answer is: participants are selected for inclusion in the study because of their extremely high or low scores on the pretest.

Answer 57

The study described in this question involved measuring the dependent variable repeatedly before and after administration of the independent variable. a. CORRECT The time-series design is a type of within-subjects design that involves evaluating the effects of an intervention by comparing multiple quantitative observations of participants before and after they are exposed to the intervention. b. Incorrect The Latin square design is a type of counterbalanced design that is used to control carryover effects. c. Incorrect The static-group comparison design includes two intact groups - one that receives the treatment and one that does not - and involves evaluating the effects of the treatment by comparing the post-test scores of the two groups. d. Incorrect The multiple-baseline design is a single-case design that involves comparing the effects of an intervention across two or more settings, behaviors, or participants. The correct answer is: time-series

Answer 58

Knowing that "homo" means the same or similar may have helped you identify the correct answer to this question even if you aren't familiar with the term "homoscedasticity." a. Incorrect See explanation for response d. b. Incorrect See explanation for response d. c. Incorrect A different range of Y scores at different values of X is referred to as "heteroscedasticity." d. CORRECT Homoscedasticity is one of the assumption for use of the Pearson r and most other correlation coefficients. This assumption is met when the range of Y scores is about the same at every value of X. Homoscedasticity and heteroscedasticity are illustrated in the Statistics and Research Design chapter of the written study materials. The correct answer is: the same range of Y scores at different values of X.

Answer 59

The study described in this question has two independent variables - type of treatment and time. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT Type of treatment is a between-groups variable in this study since each participant is assigned to one of the three interventions. In contrast, time is a within-subjects variable since all participants are being evaluated on the dependent variable at four different times. When a study includes both between-groups and within-subjects variables, it is referred to as a mixed design. d. Incorrect A counterbalanced design is a type of within-subjects design that involves administering the different levels of the independent variable to different groups of participants in a different order. The correct answer is: mixed

Answer 60

To determine the amount of variability in one variable that is explained by variability in another variable (i.e., the amount of shared variability), you must square the correlation coefficient. Conversely, to determine the correlation coefficient when the shared variability is known, you must calculate the square root of the shared variability. a. Incorrect See explanation for response c. b. Incorrect See explanation for response c. c. CORRECT To simplify the task for this question, you could calculate the square root of .16 (rather than .15). The square root of .16 is .40, so the square root of .15 will be slightly less than .40. Therefore, of the numbers listed in the answers, the best estimate of shared variability is .39. d. Incorrect See explanation for response c. The correct answer is: 0.39

Answer 61

Commonly used behavioral observation techniques include event recording, interval recording, time sampling, latency recording, and duration recording. a. Incorrect Event recording (sampling) is useful for behaviors that occur infrequently, have a long duration, and/or leave a permanent record or other product. b. Incorrect Latency recording is used to determine how long it takes for a behavior to begin after a specific event has occurred (e.g., after a verbal demand or request has been made). c. CORRECT Interval recording is useful for behaviors that have no clear beginning or end (e.g., on-task behavior) and involves recording whether or not the behavior occurred during each interval. d. Incorrect Duration recording indicates how long a behavior lasts and is useful for behaviors that have a clear beginning and end. The correct answer is: interval recording

Answer 62

When evaluating a treatment to eliminate self-destructive behaviors, a researcher would not want to use a research design that requires removal or withdrawal of a successful treatment during the course of the study. a. Incorrect Reversal designs (e.g., ABA, ABAB) require withdrawal of the treatment during subsequent baseline (A) phases and, therefore, would not be an appropriate design in this situation. b. Incorrect As its name implies, this design also requires withdrawal of the treatment and, therefore, would not be appropriate. c. CORRECT The multiple baseline design would be an acceptable design in this situation since it does not require withdrawal of the treatment from a baseline once it has been applied to that baseline. d. Incorrect The Latin square design is not a single-subject design but is a group design that is used to control order effects. The correct answer is: multiple baseline

Answer 63

The scale of measurement of the data is the primary consideration when choosing a correlation coefficient. a. Incorrect The Spearman rho is used when both variables are measured in terms of ranks. b. CORRECT The point-biserial correlation coefficient is appropriate when one variable is a true dichotomy (e.g., gender) and the other is measured on an interval or ratio scale (in this case, reaction time is measured on a ratio scale). c. Incorrect The phi coefficient is appropriate when both variables are true dichotomies (e.g., undergraduate degree/no undergraduate degree and employed/unemployed). d. Incorrect The Pearson r is the appropriate correlation coefficient when both variables are continuous (measured on an interval or ratio scale). The correct answer is: point biserial

Answer 64

The ANCOVA is used to statistically remove the effects of an extraneous variable on the dependent variable. a. Incorrect This describes the MANOVA. b. CORRECT When using the ANCOVA, the extraneous variable is the "covariate" and its effects on the dependent variable are statistically removed so that it is easier to detect the effects of the independent variable on the dependent variable. c. Incorrect This does not describe the ANCOVA. d. Incorrect A factorial ANOVA is useful for determining if there is an interaction between independent variables. (There is a factorial ANCOVA that can be used when a study includes two or more independent variables, but this question is asking about the ANCOVA more generally.) The correct answer is: statistically control variability in the dependent variable due to the effects of an extraneous variable.

Answer 65

Squaring a correlation coefficient produces a measure of shared variability. a. Incorrect The coefficient of variation, an infrequently used measure of variability, provides this information. b. Incorrect This describes the standard error of measurement. c. Incorrect This describes first-order correlation. d. CORRECT The multiple correlation coefficient indicates the degree of association between three or more variables. Like any other correlation coefficient, it can be squared in order to obtain a measure of shared variability. The correct answer is: the amount of variability shared by three or more variables.

Answer 66

In research, a control group is a comparison group that does not receive treatment or, alternatively, receives a standard treatment. a. CORRECT In single-subject research, a participant's behavior during the baseline (no treatment) and treatment phases is compared. Thus, when using single-subject designs, participants act as their own no-treatment controls. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: single-subject

Answer 67

Protocol analysis was developed by cognitive psychologists as a method for obtaining information about a person's thoughts while solving problems. a. Incorrect See explanation for response b. b. CORRECT Protocol analysis involves having the individual "think aloud" while solving a problem while an observer records what the individual says. The observer then analyzes the individual's verbalizations in order to identify the cognitive processes involved in solving the problem. c. Incorrect See explanation for response b. d. Incorrect See explanation for response b. The correct answer is: think aloud while solving a complex problem.

Answer 68

There are several methods for combining predictor scores to estimate performance or status on a criterion. When the criterion (Y variable) is group membership, discriminant analysis is an appropriate method. a. CORRECT A discriminant analysis (also known as discriminant function analysis) involves using scores on two or more predictors to predict an individual's membership in a criterion group - i.e., it is used when the criterion is measured on a nominal scale. b. Incorrect Multiple regression is used to combine scores on several predictors to predict or estimate performance or status on a criterion that is measured on an interval or ratio scale. c. Incorrect Factor analysis is used to determine the underlying factors (attributes) that are assessed by set of test items, subtests, or other measures. d. Incorrect Path analysis is used to test a model of the causal links among a set of variables. The correct answer is: discriminant analysis

Answer 69

Methods of behavioral sampling include event sampling, interval recording, and situational sampling. These are described in the Statistics and Research Design chapter of the written study materials. a. CORRECT Event sampling involves observing a behavior each time it occurs and is useful for behaviors that occur infrequently, have a long duration, or leave a permanent record. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: occur infrequently or have a long duration.

Answer 70

Knowing that another name for regression analysis is linear regression would have helped you identify the correct answer to this question. Additional information about regression analysis is provided in the Statistics and Research Design chapter of the written study materials. a. CORRECT Regression analysis is used to predict Y from a single X, and an assumption is that the relationship between X and Y is linear (i.e., the relationship can be described by a straight line). b. Incorrect Regression analysis is NOT based on the assumption that the relationship between X and Y is causal. c. Incorrect Regression analysis is the appropriate technique when there is a single X variable. d. Incorrect Regression analysis is used when there is a single Y variable. The correct answer is: the relationship between X and Y is linear.

Answer 71

The Central Limit Theorem makes three predictions about the sampling distribution of means: (a) if repeated random samples of size N are drawn from the population, as N increases in size, the sampling distribution of means increasingly approaches normal regardless of the shape of the population distribution; (b) the mean of the sampling distribution of means equals the population mean; and (c) the standard deviation of the sampling distribution of means equals the population standard deviation divided by the square root of N. a. CORRECT This response is consistent with the first prediction of the Central Limit Theorem. b. Incorrect The Central Limit Theorem assumes an infinite number of samples. In other words, it is the size of the samples, not the number of samples, that impacts the shape of the sampling distribution. c. Incorrect This is not what is predicted by the Central Limit Theorem. Instead, the Central Limit Theorem predicts that, even when the population distribution is not normal, the sampling distribution of means will increasingly approach normal as the sample size increases. d. Incorrect This is not what is predicted by the Central Limit Theorem. The correct answer is: a normal shape as the sample size increases regardless of the shape of the distribution of scores in the population.

Answer 72

A variety of sampling techniques are used to select individuals from the population. a. Incorrect Stratified random sampling involves first dividing the population into strata (segments) and randomly selecting individuals from those strata. Common strata include gender, SES, and race/ethnicity. b. CORRECT As its name implies, cluster sampling involves selecting clusters (units) rather than individuals. It is a useful technique when it is not feasible to identify or obtain access to individuals in the entire population. For example, rather than randomly selecting individual high school students from the population of all high school students, it may be necessary to select a sample of high schools and then include all students in those schools or a sample of students from those schools. c. Incorrect A convenience sample involves selecting people on the basis of their availability. d. Systematic sampling is similar to simple random sampling and involves selecting every nth (e.g., 10th, 100th) participant from a list of individuals in the population beginning with the random selection of the first participant. The correct answer is: cluster sampling

Answer 73

Trend analysis is a type of analysis of variance that is used to evaluate linear and nonlinear trends. a. CORRECT Trend analysis is used when a study involves a quantitative independent variable, and the researcher wants to determine the shape of the relationship between the independent and dependent variables. The results of the trend analysis indicate whether or not there is a statistically significant linear or nonlinear (quadratic, cubic, quartic) trend. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: a studys independent variable is quantitative.

Answer 74

Trend analysis is a type of analysis of variance that is used to evaluate linear and nonlinear trends. a. CORRECT Trend analysis is used when a study involves a quantitative independent variable, and the researcher wants to determine the shape of the relationship between the independent and dependent variables. The results of the trend analysis indicate whether or not there is a statistically significant linear or nonlinear (quadratic, cubic, quartic) trend. b. Incorrect See explanation for response a. c. Incorrect See explanation for response a. d. Incorrect See explanation for response a. The correct answer is: a study's independent variable is quantitative.

Answer 75

Correct answer: C

Answer 76

Correct answer: A

Answer 77

Correct answer: B Type I = ALPHA (the risk of saying it works when it doesn't) a. multivariate = many variables; univariate = makes it bigger b. Good answer c. Beta = Type II error. If you make beta bigger, alpha gets smaller d. No.

Answer 78

Correct answer: C Error = something is wrong with the sample a. Reliability of a test is established BEFORE we administer it. b. Validity = I'm measuring what I think I'm measuring. c. Random = unexpected, unpredictable, things we couldn't have anticipated d. Non-random = systematic and known

Answer 79

Correct answer: D Power = "Statistics" -- finding a real difference quickly Power = make sample bigger, maximise treatment (instead of 1/week, give it 3x/week)

Answer 80

Correct answer: B Ratio = you have to divide B/W B = Beginning of alphabet W = End of alphabet

Answer 81

Correct answer: D Average # = parametric

Answer 82

Correct answer: C