Deck 11: Correlational Designs

Question 1

Describe a study that uses a correlational design and explain why a true experimental design could not be used for that study.

Answer 1

No Answer

Question 2

List two reasons why it would be necessary to use a correlational design rather than a true experiment.

Answer 2

No Answer

Question 3

Which of John Stuart Mill's three rules for causality can be assessed using a correlational research design? Which of the three rules cannot be assessed with this type of design?

Answer 3

No Answer

Question 4

What is a zero-order correlation?

Question 5

What does the correlation coefficient (r) indicate? What does the square of the correlation coefficient (r²) indicate?

Question 6

What is the purpose of a bivariate regression equation? What are its components?

Question 7

Why might a researcher compare the strength of a correlation of two variables in different samples? What statistic would be used for this comparison?

Question 8

A researcher finds that the slope of a bivariate regression analysis differs between two subgroups. What does this indicate about the accuracy of predicting Y from X for these two subgroups? Explain your reasoning.

Question 9

Explain the assumptions of linearity and additivity as they apply to correlational research.

Question 10

Explain the difference between a linear relationship between two variables and a curvilinear relationship.

Question 11

Why is it important to plot the scores on the variables you wish to correlate prior to conducting your statistical analysis?

Question 12

Distinguish between a true score correlation and an observed correlation.

Question 13

Define attenuation and explain how it affects the observed correlation relative to the true score correlation.

Question 14

Explain how the reliability of a measure affects the outcome of a correlational analysis.

Question 15

Define restriction in range and explain how it might affect an observed correlation.

Question 16

What are outliers? What effect might they have on a correlational analysis?

Question 17

Imagine that you have measured self-esteem and anxiety in both a Japanese and a British sample of respondents. You plan to examine the correlation between these two measures. What steps should you take to evaluate whether there are subgroup differences on these measures? Why would this be important?

Question 18

Distinguish between multifaceted and multidimensional constructs.

Question 19

You have identified multifaceted constructs in your research. What issues should you consider in deciding whether these constructs should be combined?

Question 20

List four guidelines that should be followed in conducting quality correlational research.

Question 21

What is a partial correlational analysis and when would it be used? How is a partial correlation interpreted?

Question 22

What is multiple regression analysis? List two purposes of this type of analysis.

Question 23

Distinguish between simultaneous and sequential (hierarchical) multiple regression analysis.

Question 24

What is simultaneous multiple regression analysis and when would it be used?

Question 25

What is sequential (hierarchical) multiple regression analysis and when would it be used?

Question 26

What is a multiple correlational coefficient and what information does it provide to researchers?

Question 27

Distinguish between a standardized and an unstandardized regression coefficient. When would each be used and why?

Question 28

What does the square of the multiple regression coefficient (R²) represent in a multiple regression analysis?

Question 29

Under what circumstances would a researcher test for a change in R²? If the change is statistically significant, what would information would that give the researcher?

Question 30

Describe the problem of multicollinearity and explain how a researcher can test for this possibility. Why would it be important to do so?

Question 31

List four possible causes of multicollinearity.

Question 32

You are conducting a multiple regression analysis. You correlate the variables you plan to include in that analysis and discover that two of your variables are strongly correlated (r = .83). Is this a problem? Why or why not? If you believe it is a problem, what can you do to address it?

Question 33

Under what circumstances is conducting a multiple regression analysis a better option than conducting an analysis of variance? Why?

Question 34

What assumptions of an analysis of variance do not apply to a multiple regression analysis?

Question 35

What is meant by dichotomization of a continuous variable? Explain why dichotomizing a continuous variable is problematic.

Question 36

What is a logistic regression analysis? For what type of data would this analysis be appropriate?

Question 37

What is an odds ratio and how is it interpreted?

Question 38

What is a multiway frequency analysis? For what type of data would this analysis be appropriate?

Question 39

What is log linear analysis?

Question 40

What is a logit analysis? For what type of data would this analysis be appropriate?

Question 41

You have conducted a study using a categorical independent variable and a categorical dependent variable. What three types of statistical analysis could you conduct? How would you decide which is most appropriate?

Question 42

Which of the following statements about correlational research is FALSE?

A) Correlational research can be used to study variables that cannot be manipulated.
B) Correlational research can determine whether one variable is the cause of another variable.
C) Correlational research can be used to study variables that it would be unethical to manipulate.
D) Correlational research has no advantages over experimental research.

Question 43

Professor Li finds that high school students who spend more time on social media have higher levels of depression. This difference is statistically significant. Assuming that the measures had adequate validity and that proper research procedures were followed, it would be most appropriate to conclude that, for this population,

A) spending time on social media increases depression.
B) being more depressed causes students to spend more time on social media.
C) some unmeasured third variable causes both higher social media use and increased depression.
D) any of the above could be true.

Question 44

You calculate the correlation between height and weight in a group of research participants. The resulting correlation coefficient represents the ________ correlation between height and weight in your sample.

A) zero-order
B) partial
C) multiple
D) multi-linear

Question 45

A correlation coefficient (r) represents

A) the amount of change that occurs in one variable associated with each unit change in the other variable.
B) the size and direction (positive or negative) of the relationship between the two variables.
C) the proportion of variance in one of the variables that can be accounted for by the other variable.
D) the difference in standard deviation units of the means of the two variables.

Question 46

The square of a correlation coefficient (r²) represents

A) the amount of change that occurs in one variable associated with each unit change in the other variable.
B) the joint reliability of two variables.
C) the proportion of variance in one variable that can be accounted for by the other variable.
D) the difference in standard deviation units of the means of two variables.

Question 47

A bivariate regression equation is used to

A) predict scores on one variable from scores on a correlated variable.
B) compute the joint reliability of two correlated variables.
C) compute the proportion of variance in one variable that can be accounted for by the other variable.
D) compute the difference in standard deviation units of the means of two correlated variables.

Question 48

The slope component of a bivariate regression equation represents

A) the amount of change that occurs in the dependent variable associated with each unit change in the independent variable.
B) the joint reliability of two correlated variables.
C) the proportion of variance in a variable that can be accounted for by another variable.
D) the predicted score on the dependent variable when the score on the independent variable is zero.

Question 49

The intercept component of a bivariate regression equation represents

A) the amount of change that occurs in the dependent variable associated with each unit change in the independent variable.
B) the predicted score on the dependent variable when the score on the independent variable is zero.
C) the joint reliability of two correlated variables.
D) the proportion of variance in one variable that can be accounted for by another variable.

Question 50

In the bivariate regression equation Y = a + bX, the value a is called the

A) slope
B) intercept
C) variance percentage
D) correlation coefficient for the relationship between X and Y

Question 51

In the bivariate regression equation Y = a + bX, the value b is called the

A) slope
B) intercept
C) variance percentage
D) correlation coefficient for the relationship between X and Y

Question 52

You compute the following bivariate regression equation: Y = 3.11 + 1.25X. The 1.25 represents the

A) joint reliability of X and Y.
B) proportion of variance in Y that can be accounted for by X.
C) amount of change that occurs in Y variable associated for each unit change in X.
D) predicted score on Y when the score on X is zero.

Question 53

You compute the following bivariate regression equation: Y = 3.11 + 1.25X. The 3.11 represents the

A) amount of change that occurs in Y associated with each unit change in X.
B) joint reliability of X and Y.
C) proportion of variance in Y that can be accounted for by X.
D) predicted score on Y when the score on X is zero.

Question 54

Dr. Thompson wants to know if the correlation between height (variable X) and weight (variable Y) is the same or different for women and men. Which of the following is TRUE about calculating this difference?

A) It is not possible to calculate the difference between correlations from different groups.
B) If the difference between correlations is not statistically significant, then there is no difference in the relationship between height and weight for women and men.
C) A difference in the slopes calculated for the correlations indicates an interaction between gender and height in predicting weight.
D) If the correlation is the same in both groups, then the intercepts and slopes for the bivariate regression equation are the same for women and men.

Question 55

You compute the correlation between two variables twice, once for the female participants in your study and once for the male subjects. You find that the difference between these two correlations is statistically significant. This finding

A) represents a case of reverse causation.
B) suggests that the measures used were unreliable.
C) means that gender of participant moderates the relationship between the two variables.
D) is mathematically impossible.

Question 56

The use of simple correlational techniques implies the assumption(s) that

A) there is a linear relationship between the independent and dependent variables.
B) if there is more than one independent variable, there are no interactions between independent variables.
C) if there is more than one independent variable, the independent variables are multicollinear.
D) a and b but not c.

Question 57

You calculate the correlation between amount of job experience and work productivity in a group of employees in an organization you are studying. You find that as experience increases productivity also increases. Your finding represents a ________ relationship between job experience and productivity.

A) positive
B) curvilinear
C) negative
D) polylinear

Question 58

You calculate the correlation between work load and job satisfaction in a group of employees in an organization you are studying. You find that as work load increases satisfaction decreases. Your finding represents a ________ relationship between job experience and productivity.

A) positive
B) curvilinear
C) negative
D) polylinear

Question 59

You examine the relationship between stress and job performance in a group of employees in an organization you are studying. You find that, up to a point, as stress increases performance also increases, but that after that point, as stress increases performances decreases. Your finding represents a ________ relationship between stress and performance.

A) positive linear
B) curvilinear
C) negative linear
D) polylinear

Question 60

In correlational analyses that involve more than one independent variable, the assumption of additivity means that

A) all the independent variables have linear relationships with the dependent variable.
B) the sum of the correlations between the independent variables and dependent variables equals 1.
C) all the independent variables have the same type of relationship (positive or negative) with the dependent variable.
D) there are no interactions between any of the independent variables.

Question 61

We say that a correlation is attenuated when

A) the observed correlation between two measures greatly underestimates the true correlation between the constructs they measure.
B) its scores include a large number of outliers.
C) the observed correlation violates the assumption of additivity.
D) there are subgroup differences in the correlation between two measures.

Question 62

An attenuated correlation can result when

A) one or both of the measures involved in the correlation has low reliability.
B) the range of scores on one or both measures found in the sample is less than that found in the population.
C) the relationship between two variables is curvilinear rather than linear.
D) any of the above situations occur.

Question 63

The reliability of the measures used in a correlational analysis is important because measures low in reliability

A) produce observed correlations that greatly underestimate the true correlation between the constructs they measure.
B) are likely to be high in multicollinearity.
C) cannot meet the assumption of linearity.
D) produce an undesirable restriction in range of the observed correlation.

Question 64

Restriction in range means that

A) a measure exhibits an undesirable range of reliability.
B) scores of a measure observed in a sample have a range of values that is less than the range of scores in the popu?lation.
C) no score in a sample is greater than three standard deviations from the mean.
D) there are no subgroup differences in the correlation between two measures.

Question 65

In a correlational study it is important to have a range of scores in the research sample that is similar to the range of scores in the population. This is because a restricted range of scores

A) is likely to result in high multicollinearity.
B) produces observed correlations that greatly underestimate the true correlation between the constructs they measure.
C) creates outliers that threaten the statistical conclusion validity of the data analysis.
D) is likely to lead to a violation of the assumption of additivity.

Question 66

An outlier is a(n)

A) score that is three or more standard deviations above or below the sample mean.
B) measure that is low in reliability.
C) score that is very close to the mean and so causes a restriction in the range of scores.
D) observed correlation that underestimates the true correlation between two measures.

Question 67

Which of the following is a DESIRABLE characteristic of a set of scores to be used in a correlational analysis?

A) restriction in range
B) the presence of several outliers
C) multicollinearity
D) none of the above

Question 68

When identifiable subgroups (such as different ethnic groups) exist in a sample, it can be useful to compute the correlations found in each group because

A) subgroups represent outliers that must be taken into account when interpreting the results of the research.
B) if there are subgroup differences in correlations then the overall correlation will not accurately represent the results of the study.
C) subgroup differences in correlations is one indicator of multicollinearity in the data.
D) a large difference in subgroup correlations is an indicator of restriction in range.

Question 69

One issue that arises when conducting research with multifaceted constructs as independent variables is whether to treat the facets as separate variables or to combine them into a single variable. It is best to treat the facets as SEPARATE variables when

A) the facets are theoretically or empirically related to differ?ent dependent variables or to different facets of a dependent variable.
B) the theory of the construct predicts an interaction among the facets.
C) it is convenient to do so.
D) a and b, but not c.

Question 70

One issue that arises when conducting research with multifaceted constructs as independent variables is whether to treat the facets as separate variables or to combine them into a single variable. It is appropriate to COMBINE the facets when

A) the research focuses on the latent variable represented by the combination of facets rather than in the particular aspects of the variable repre?sented by the facets.
B) from a theoretical perspective, the latent variable is more important, more interesting, or represents a more appropriate level of abstraction than do the facets.
C) the facets are highly correlated and so represent a single construct rather than different facets of a construct.
D) any of the above situations occur.

Question 71

You have a situation in which three variables-X, Y, and Z-are correlated with one another. The correlations of X with Y, X with Z, and Y with Z are referred to as ________ correlations.

A) partial
B) multiple
C) differential
D) zero-order

Question 72

You have a situation in which three variables-X, Y, and Z-are correlated with one another. The partial correlation of X with Y controlling for Z represents

A) what the relationship between X and Y would be if all the research participants had the same score on Z.
B) the difference between the correlation of X with Y and X with Z.
C) the average of the correlations of X with Y, X with Z, and Y with Z.
D) the amount of variance in Z accounted for by the combination of X and Y.

Question 73

You have a situation in which three variables-X, Y, and Z-are correlated with one another. You find that the partial correlation of X with Y controlling for Z is zero. This finding means that

A) Z is a cause of both X and Y.
B) the relationship between X and Y can be completely accounted for by their common relationships with Z.
C) scores on Z cannot be used to predict scores on either X or Y.
D) both a and b

Question 74

A ________ variable in multiple regression analysis is analogous to a(n) ________ variable in experimental research.

A) criterion; independent
B) predictor; dependent
C) predictor; independent
D) criterion; extraneous

Question 75

A ________ variable in multiple regression analysis is analogous to a(n) ________ variable in experimental research.

A) criterion; independent
B) predictor; dependent
C) predictor; extraneous
D) criterion; dependent

Question 76

You want to know the extent to which political orientation can predict prejudice with age, level of education, and amount of contact with minority group members controlled. The most appropriate form of multiple regression analysis to use with your data would be

A) sequential.
B) statistical.
C) simultaneous.
D) synchronous.

Question 77

You want to derive an equation to predict first-year college grade point average from Scholastic Assessment Test scores, high school grade point average, and a measure of motivation to success in college. The most appropriate form of multiple regression analysis to use with your data would be

A) sequential.
B) statistical.
C) simultaneous.
D) synchronous.

Question 78

In multiple regression analysis, the multiple correlation coefficient (R)

A) is the value by which the score on an independent variable is multiplied to predict the score on the dependent variable.
B) represents the degree of association between the independent variables as a set and the dependent variable.
C) represents the increase in the proportion of variance in the dependent variable that is accounted for by adding another independent variable to the regression equation.
D) represents the condition that arises when two or more predictor variables are highly correlated with each other.

Question 79

In multiple regression analysis, the square of the multiple correlation coefficient (R²) represents

A) the amount of change that occurs in the dependent variable associated with each unit change in the independent variables.
B) the joint reliability of the all the variables involved in the analysis.
C) the proportion of variance in the dependent variables that can be accounted for by the independent variables considered as a set.
D) the mean of the dependent variable.

Question 80

In multiple regression analysis, the regression coefficient

A) is the value by which the score on an independent variable is multiplied to predict the score on the dependent variable.
B) represents the degree of association between the independent variables as a set and the dependent variable.
C) represents the increase in the proportion of variance in the dependent variable that is accounted for by adding another independent variable to the regression equation.
D) represents the condition that arises when two or more predictor variables are highly correlated with each other.