Deck 14: Elaborating Bivariate Tables

Question 1

Which of the following describes a spurious relationship between X, Y, and Z?

A) X causes Z, which causes Y.
B) Z causes X, which causes Y.
C) Z has no effect on the relationship between X and Y.
D) Z causes both X and Y.

Answer 1

D

Question 2

Another term for interaction is________.

A) intervention.
B) spurions.
C) direct.
D) specification.

Answer 2

D

Question 3

With spurious relationships, the Z (spurious) variable comes _______ the independent and dependent variables in order of time.

A) before
B) during
C) after
D) at any time

Answer 3

A

Question 4

Which of the following describes an intervening relationship between X, Y, and Z?

A) X causes Z, which causes Y.
B) X causes Z, but neither has a relationship with Y.
C) Z causes both X and Y.
D) Z causes neither X nor Y.

Question 5

Multivariate analysis provides techniques by which the researcher can________.

A) gather more information about a bivariate relationship.
B) probe the possible causal relationships among variables.
C) test and revise theory.
D) All of the answer choices

Question 6

When there is evidence of a(n) __________ relationship, partial gamma (G p) should not be calculated.

A) interactive
B) direct
C) intervening
D) spurious

Question 7

Spurious or intervening relationships among variables are indicated by a pattern of outcomes in which the bivariate relationship is the same across all partial tables and________.

A) the same as in the original bivariate table.
B) much weaker than in the bivariate table.
C) much stronger than in the bivariate table.
D) the variables are not associated in the bivariate table.

Question 8

In an investigation of the relationship between political ideology (liberal - conservative) and support for civil liberties such as free speech, a researcher calculated a bivariate gamma of 0.65. After controlling for gender, the partial gamma was 0.60. Which of the following correctly describes this outcome?

A) There seems to be a direct relationship between ideology and support for civil liberties.
B) This relationship appears to be spurious.
C) There is an interactive relationship between ideology and support for civil liberties.
D) Women are much more supportive of civil liberties than men.

Question 9

Gammas computed on partial tables are both .49 and the bivariate gamma is .52. This pattern of outcomes would be taken as evidence of________.

A) a direct relationship between the independent and dependent variables.
B) a spurious relationship.
C) a direct relationship between the control variable and the independent.
D) no association between any of the three variables.

Question 10

A measure of association for X and Y shows that the two variables are associated at -.63. After controlling for Z, the measures for the partial tables are -.33 and -.78. This pattern of outcomes would be consistent with________.

A) a direct relationship between X and Y.
B) a spurious relationship between X and Y and a direct relationship between X and Z.
C) interaction between X, Y, and Z.
D) an intervening relationship between X, Y, and Z.

Question 11

If a researcher finds interaction between variable X and Y after controlling for Z, what would be the next step in the statistical analysis?

A) Disregard Z and select a new control variable.
B) Disregard X and select a new independent variable.
C) Disregard Y and select a new dependent variable.
D) Incorporate Z into the analysis.

Question 12

In multivariate analysis we investigate________.

A) the mutual effects of two variables on each other.
B) what happens to the third variable when the first two are controlled.
C) what happens when many variables are free to vary.
D) how a specific bivariate relationship is affected by a control variable.

Question 13

The gamma for a bivariate table is 0.42 and the partial gamma (G p) is 0.40. This is consistent with________.

A) interaction between the variables.
B) an intervening relationship between X, Y, and Z.
C) a direct relationship between X and Y.
D) a spurious relationship between X and Z.

Question 14

In an interactive relationship, the conditional distributions of Y in the partial tables would be ___________ and the conditional distributions of Y in the bivariate table would be ____________.

A) different from each other, different from the partial tables
B) the same, different from the partial tables
C) different from each other, the same as the partial tables
D) the same, the same as the partial tables

Question 15

Interaction is said to exist when________.

A) the partial tables differ from each other but are the same as the bivariate table.
B) the partial tables are the same but are different from the bivariate table.
C) the partial tables display a weaker relationship than the bivariate table.
D) the partial tables differ from each other and from the bivariate table.

Question 16

In an investigation of the relationship between political ideology (liberal - conservative) and support for civil liberties such as free speech, a researcher calculated a bivariate gamma of 0.65. After controlling for gender, the partial gamma was 0.07. Which of the following correctly describes this outcome?

A) There seems to be a direct relationship between ideology and support for civil liberties.
B) This relationship appears to be spurious.
C) There is an interactive relationship between ideology and support for civil liberties.
D) Women are much more supportive of civil liberties than men.

Question 17

The gamma for a bivariate table is 0.42 and the partial gamma (G p) is 0.05. This is consistent with________.

A) interaction between X and Z.
B) interaction between X and Y.
C) direct relationship between X and Y.
D) a spurious relationship between X and Y.

Question 18

Which of the following describes an interactive relationship between X, Y, and Z?

A) X causes Z, which causes Y.
B) X and Y are associated for some categories of Z but not for others.
C) X and Y are associated for all categories of Z.
D) X and Y are not associated for any categories of Z.

Question 19

Another term used to describe intervening relations is________.

A) specification.
B) interpretation.
C) direct.
D) unusual.

Question 20

Partial gamma (G p)________.

A) measures the association between X and Y after removing the effects of Z.
B) is the average of all gammas computed on partial tables.
C) can never be lower in value than the bivariate gamma.
D) is computed for each partial table separately.

Question 21

Under what conditions should partial gamma (G p) NOT be calculated?

A) When sample size is very large.
B) When there is no dependent variable.
C) When the variables are ordinal in level of measurement.
D) When there is evidence of interaction.

Question 22

The technique of elaboration is appropriate for ___________ variables which have been organized into _______________.

A) interval-ratio level, bivariate tables
B) interval-ratio level, scatterplots
C) nominal or ordinal level, scatterplots
D) nominal or ordinal level, bivariate tables

Question 23

Which of the following describes a direct relationship between independent and dependent variables after controlling for a third variable?

A) The bivariate relationship changes across the partial tables.
B) One partial table is the same as the bivariate table, the other(s) are very different.
C) The bivariate relationship is the same across the partial tables.
D) The partial tables have many empty cells.

Question 24

An important limitation of the elaboration technique is sample size because________.

A) too many cases destabilize the value of partial gamma.
B) statistical significance is reduced with large samples.
C) cases must be divided over a series of partial tables.
D) with many cases, not enough cells in the partial tables will be empty.

Question 25

If we investigate a bivariate relationship by controlling for gender, which of the following would be consistent with an intervening relationship?

A) The relationship between X and Y is the same for men and women.
B) The bivariate relationship between X and Y is zero for both men and women.
C) X and Y are related for men but not for women.
D) There are many fewer men than women.

Question 26

With the elaboration technique, it is important to have large samples to________.

A) avoid having to control for a third variable.
B) fill up the cells in the partial tables.
C) eliminate the possibility of spurious relationships.
D) increase the value of the measure of association.

Question 27

How many partial tables will be produced if a researcher controlled for gender?

A) One
B) Two
C) Three
D) Four

Question 28

Examining the bivariate tables in Section 14.1, which group was happiest as a function of their level of education?

A) males with a graduate degree.
B) males and females with at least a Bachelor's degree.
C) females with a graduate degree.
D) males and females with less than a high school diploma.

Question 29

If the contents of partial tables are _________, the outcome will be the original frequency table.

A) added
B) subtracted
C) divided
D) multiplied

Question 30

Which of the following describes a direct relationship between X and Y?

A) X causes Z, which causes Y
B) Y causes Z, which causes X
C) Both Z and Y cause Z
D) X causes Y

Question 31

With the elaboration technique, one solution for the problem of empty cells is to________.

A) increase the number of control variables.
B) collapse categories of all variables as much as possible.
C) decrease sample size.
D) disregard the control variable.

Question 32

The chief psychologist at a state prison has developed a rehabilitation program for inmates convicted of violent crimes against females. All of the prisoners have been rated on the frequency of their attendance at sessions (the program is voluntary) and have completed a battery of tests which measures their level of hostile and aggressive feelings towards females. Are less hostile attitudes associated with frequent attendance at sessions? Does the program work?

Calculate percentages and a measure of association for the table. Choose a measure of association appropriate for these variables. Characterize the relationship in terms of the existence, strength, and pattern or direction of the relationship in a paragraph. To further test the nature of the association between attendance at the program and hostility towards women, the investigator controlled for the age of the inmates. For each partial table below, compute percentages and a measure of association and interpret the pattern of relationships. Write a paragraph analyzing the effect of controlling for age on the bivariate relationship.Hostility Towards Women by Frequency of Attendance, Controlling for Ag
A. Younger Inmates (less than 35):
B. Older Inmates (35 and older):

Question 33

Regarding the study in your textbook, which of the following statements is correct?

A) Conservatives were most likely to think greenhouse gases were dangerous.
B) Liberals were least likely to think greenhouse gases were dangerous.
C) Moderates were most likely to think greenhouse gases were dangerous.
D) Liberals were most likely to think greenhouse gases were dangerous.

Question 34

What variables should we control for? We are most likely to find the answer in________.

A) a good statistics text.
B) our research design.
C) our theory.
D) common sense.

Question 35

A possible solution for dealing with the problem of empty cells in the partial tables when using the elaboration technique is to________.

A) collapse the categories of all variables to the minimum possible.
B) avoid using a control variable.
C) use more than one control variable.
D) eliminate the dependent variable.

Question 36

The elaboration technique controls for a third variable by________.

A) fixing its values so they are no longer free to vary.
B) eliminating the variable from the analysis.
C) replacing its values with those of the independent.
D) adding a column to the bivariate table to include the control variable.

Question 37

The three variables involved in the elaboration of bivariate tables are________.

A) independent, control, continuous.
B) nominal, ordinal, interval-ratio.
C) first dependent, second dependent, independent.
D) independent, dependent, control.