Deck 16: Testing for Association: Pearsons R and Regression

Question 1

An r-squared value of .40 means that the independent variable has accounted for 16% of the variance in the dependent variable.

Answer 1

True

Question 2

Given the formula Y = a + bX, if the Pearson's r value is .60 the value of b would have to be:

A) half of .60 or .30
B) 60
C) .36
D) none of the above

Answer 2

D

Question 3

It is possible to have both a high correlation coefficient and a low regression coefficient.

Answer 3

True

Question 4

The correlation value for Pearson's r does not reach 1.0 (unity) but will after you square it.

Question 5

If you calculated a Pearson's r-square of .36 between median incomes and city crime rates it means that:

A) 36 percent of the variance in city crime rates is explained
B) 60 percent of the variance in city crime rates is explained
C) 6 percent of the variance in median incomes is explained
D) 13 percent of the variance in either median incomes or city crime rates is explained

Question 6

In order to assess the results of a prediction using Pearson's r, you would need

A) both the correlation coefficient and the regression coefficient
B) both b and Beta
C) only r
D) none of the above

Question 7

An r-squared value of .50 means that the relationship is positive.

Question 8

Pearson's r is known as a parametric statistic.

Question 9

If the correlation value for Pearson's r is positive, an asymmetric version of the test has been used.

Question 10

Pearson's r is reasonably able to withstand the effect of violating one of its assumptions.

Question 11

Pearson's r can easily be extended to a multivariate statistical technique

A) with multiple independent variables
B) with multiple dependent variables
C) with the restriction that only one control variable can be used
D) all of the above

Question 12

Pearson's r will provide both symmetric and asymmetric relationship values, but you have to know which one to use.

Question 13

If Pearson's r = .50 and N = 20, then the proportion of explained variance is:

A) .50
B) 20
C) 10
D) .25

Question 14

Pearson's r is designed for data measured at the

A) nominal level
B) ordinal level
C) interval level
D) metric ordinal level

Question 15

[Instructor note: If you are creating an exam with all measures of association, then use this.] Outliers (extreme scores) will most affect:

A) The Uncertainty Coefficient (U)
B) Gamma
C) Pearson's r
D) Lambda

Question 16

The most important advantage of using Pearson's r is that

A) it can be interpreted as a direct-PRE measure
B) it is the only statistic that has multivariate capabilities
C) the measure itself has very few assumptions about data
D) it is sensitive to the magnitude of change between pairs of scores

Question 17

The linear function, Y = a + bx, is used to estimate the actual amount of effect variable X will have on variable Y

A) under the condition that X = 2Y
B) under the condition that Y = 2X
C) for each unit increase in Y
D) for each unit increase in X

Question 18

If you wanted to know the amount of change in Y as a produce of X, you would be interested in

A) the correlation coefficient or r
B) the coefficient of nondetermination or c
C) the standardized coefficient or s
D) the regression coefficient or b

Question 19

The type of relationship measured by Pearson's r is

A) exponential
B) uniform
C) linear
D) curvilinear

Question 20

Assumptions of Pearson's r include which of the following?

A) normally distributed data
B) homoscedasticity
C) the residual variance in uncorrelated with the independent variable
D) all of the above

Question 21

As a member of the General Linear Model, Pearson's r can be calculated from ANOVA.

Question 22

An alternative to regression, when assumptions are severely violated, is the Contingency Coefficient statistic.