ch16, Chapter 16, Bivariate Correlation and Regression

Multiple Choice

1. A scatter diagram of the plot between two variables reveals that the data points
are aligned in a linear fashion, down-sloping to the right. The type of relationship
just described would be?
a. strong positive linear relationship
b. perfect parabolic relationship
c. strong negative linear relationship
d. no relationship
e. none of the above

Ans: C
Response: See page 519

2. If you are studying the impact of training on salesperson performance as indicated

on an exam taken by salespersons at the end of training, salesperson sales would
be the ____________ variable, and salesperson exam score would be the
_____________ variable.
a. criterion, predictor
b. covariate, independent
c. independent, predictor
d. dependent, independent
e. none of the above

Ans: D
Response: See page 518

3. Which of the following bivariate techniques is not appropriate for ratio or interval
data?
a. regression analysis
b. ANOVA
c. z-test
d. Pearson’s correlation
e. all are appropriate

Ans: E

219
Response: See page 518

4. The percentage of total variation in the dependent variable that is described by the
independent variable is expressed by_________.
a. coefficient of determination
b. correlation coefficient
c. coefficient of covariation
d. regression coefficient
e. none of the above

Ans: A
Response: See page 526

5. The change in the dependent variable given a one-unit change in the independent
variable is expressed by the ________________.
a. coefficient of determination
b. correlation coefficient
c. coefficient of covariation
d. regression coefficient
e. none of the above

Ans: D
Response: See pages 518-526

6. ____________ is the mathematical technique for fitting a line that best describes
the relationship between two variables.
a. Pearson’s Product-Moment Correlation
b. least squares procedure
c. chi-square test
d. significance test
e. none of the above

Ans: B
Response: See page 521

220
7. Y

* *
* *
X

Given this plot, which of the following provides the best description?
a. perfect linear association
b. strong exponential association
c. strong parabolic association
d. no association
e. strong curvilinear association

Ans: D
Response: See page 519

8. In the equation Y = a + b(X) + e, b refers to the _____________.

a. slope of the regression line
b. change in Y for a 1 unit change in X
c. amount of error in the equation
d. both (a) and (b)
e. none of the above

Ans: D
Response: See page 521

9. A regression equation that estimates the number of hours studying impact on test
score would most likely produce what type of relationship?
a. positive
b. negative
c. curvilinear
d. none

Ans: A
Response: See page 519

221
10. Research on fear appeals in advertising suggests that a moderate amount of fear is
most effective at achieving persuasion. That is, using not enough fear or too
much fear in advertising will not be effective. This is an example of what type of
relationship that regression might find?
a. positive linear
b. negative linear
c. curvilinear
d. no relationship
e. none of the above

Ans: C
Response: See page 519

11. Which of the following is the appropriate technique for correlation analysis,
which involves metric (interval or ratio) data?
a. Spearman’s rank-order correlation
b. Jensen’s analysis of covariance
c. Kendall’s coefficient of concordance
d. Pearson’s Product-Moment Correlation
e. none of the above

Ans: D
Response: See page 530

12. Computed values in correlation analysis range from?

a. 0 to 1
b. -1 to +1
c. -1 to 0
d. -.99 to + .99
e. none of the above

Ans: B
Response: See page 530

13. When two variables are not correlated at all, the correlation coefficient would be:
a. -1
b. 0
c. 1
d. -2
e. none of the above

222
Ans: B
Response: See page 530

14. Which of the following would not yield the coefficient of determination (R2)?
a. 1 – [(SSE/SST)]
b. the square root of the coefficient of correlation
c. the square of the coefficient of correlation
d. 1 – [(SST – SSR)/SST]
e. none of the above

Ans: B
Response: See pages 525-530

15. IF SST = 90,000 and SSE = 15,000, what will be the value of the calculated
coefficient of determination (R2)?
a. .833
b. .800
c. .200
d. .167
e. none of the above

Ans: A
Response: See pages 525-530

16. Which of the following statements regarding total variation (SST) in regression
analysis is true?
a. SST is a measure of variation of the observed dependent values around
their mean, taking into consideration the values of the independent
variable
b. SST is computed only when there is more than 1 independent variable
c. SST in the dependent variable is the sum of explained and unexplained
variation
d. the larger the total variation, the higher the coefficient of correlation
e. both (c) and (d)

Ans: C
Response: See page 527

223
17. If a marketer wants to correlate an ordinal ranking of TV sets with the nominal
scale variable gender, the marketer would have to use_____________.
a. ANOVA
b. Pearson’s Product Moment Correlation
c. z-test
d. chi-square test
e. none of the above

Ans: E
Response: See page 530

18. Which test statistic is used to test the significance of the results of a regression
analysis?
a. F
b. t
c. Z
d. chi-square
e. none of the above

Ans: A
Response: See page 529

19. Which of the following is true regarding sum of squares due to regression (SSR)?
a. it represents the variation not explained by the regression
b. a larger SSR relative to total variation is an indication of a better fitting
regression line, and hence higher coefficient of determination
c. if the total variation (SST) is very large, then it is highly likely that the
SSR for a regression function will be small
d. both (b) and (c)
e. none of the above

Ans: B
Response: See pages 527-528

20. Bivariate regression can NOT demonstrate:

a. when the two variables are linear
b. when the two variables are strongly inversely related
c. when the two variables are strongly positively related
d. when the two variables are causally related

224
Ans: D
Response: See page 530

21. A correlation analysis between sales and sales training scores results in R = +.98.
Which of the following best interprets the relationship between sales and sales
training?
a. 98% of the salespeople taking the test have higher sales
b. the correlation between sales and sales training is very weak and
insignificant
c. the correlation between sales and sales training is strong and positive,
indicating that higher sales training scores are closely associated with
higher sales and vice-versa
d. 98% of the variation in sales is explained by variations in sales training
scores
e. none of the above

Ans: C
Response: See page 530

22.
Y
*
*
* *
*

*
*
X

Given this plot, which of the following provides the best description?
a. strong inverse linear association
b. strong exponential association
c. strong parabolic association
d. no association
e. strong curvilinear association

Ans: A
Response: See page 519

225
23. A department store asked a market researcher to examine the relationship between
the percent change in the number of salespeople in its various departments and the
corresponding percent increase in volume of sales. The marketer decided to use
bivariate regression to evaluate the relationship. The results were as follows:
R2 = -.12 with a regression coefficient (b) of .004. Which of the following would
be the best explanation for the market researcher to provide the department store
manager?
a. the department store should continue to add more salespeople
b. the R2 indicates that there is little or no association between increasing the
number of salespeople and increasing sales; the regression coefficient
indicates little change in sales associated with change is the number of
salespeople
c. the regression coefficient (b) indicates a strong positive and significant
relationship between adding salespeople and higher sales
d. both (a) and (c)
e. none of the above

Ans: B
Response: See pages 522-528

24. A market researcher observes the following graph depicting the relationship
between sales promotion and sales:

Y (sales)

*
*
X ($ spent on sales promotion)
The computed F-statistic of 125.26 is significant at .01 or 99% confidence.
Which of the following explanations would you provide for your client?

a. there is a strong positive linear relationship between sales and dollars

spent on sales promotion
b. the store should quit spending money on sales promotion
c. the information is not conclusive
d. there is an insignificant linear association between sales and sales
promotion
e. none of the above

226
Ans: A
Response: See page 519

25. A marketer wants to produce a model, which will depict the relationship between
changes in advertising expenditures and sales, and hence, whether more should be
spend on advertising. Which of statistical techniques will produce a descriptive
model of advertising and sales?
a. Pearson’s Product-Moment Correlation Analysis
b. Bivariate regression with sales as the independent variable and advertising
expenditures as the dependent variable
c. Bivariate regression with sales as the dependent variable and advertising
expenditures as the independent variable
d. T-test
e. None of the above

Ans: C
Response: See page 518

26. In a regression output in SPSS, R2 represents:

a. the independent variable
b. the dependent variable
c. the estimated slope of the regression line
d. the coefficient of determination

Ans: D
Response: See page 526

27. An R2 of .05 would indicate:

a. A significant relationship between X and Y
b. A negative relationship between X and Y
c. A very small relationship between X and Y
d. none of the above

Ans: C
Response: See page 526

28. In a regression equation, which symbol represents the independent variable?

a. Y

227
 b. X
c. e
d. r
e. none of the above

Ans: B
Response: See page 521

29. Which of the following is NOT included in the equation for the coefficient of
determination?
a. explained variation
b. total variation
c. estimated variation
d. unexplained variation

Ans: C
Response: See page 527

True – False

30. Bivariate statistical techniques analyze the relationship between two or more
variables.

Ans: False
Response: See page 518

31. Dependent variables are also known as predictor variables.

Ans: False
Response: See page 518

32. Bivariate statistical techniques are often used to establish cause and effect
relationships between two or more variables.

Ans: False
Response: See page 518

228
33. If a marketer was trying to determine the productivity of an advertising campaign,
the marketer might invoke a bivariate regression analysis, in which sales would be
the dependent variable and advertising expenditures the independent variable.

Ans: True
Response: See page 518

34. The coefficient of determination is used in correlation analysis.

Ans: False
Response: See page 530

35. In correlation analysis, the closer the value of R is to –1 or +1, the stronger the
correlation between the two variables in question.

Ans: True
Response: See page 530

36. A perfect correlation indicates that two variables are causally related.

Ans: False
Response: See page 530

37. If the relationships we are investigating involve more than one independent
variable, then we would want to use correlation analysis in lieu of regression
analysis.

Ans: False
Response: See page 530

38. Bivariate correlation examines the relationship between two variables.

Ans: True

229
Response: See page 530

39. A R = -.95 is always weaker than +.95, because positive relationships are more
significant than negative relationships.

Ans: False
Response: See page 530

40. In bivariate regression, both the independent and dependent variables are
measured in the form of nominal and/or ordinal data.

Ans: False
Response: See page 530

41. A positive correlation between X and Y suggests that higher values of X are
consistently associated with higher values of Y, and vice-versa.

Ans: True
Response: See page 530

42. If the coefficient of determination resulting from a bivariate regression analysis is

very high, then it means that changes in the independent variable cause changes in
the dependent variable.

Ans: False
Response: See page 526

43. The variation explained by the regression equation is called SSR or sum of
squares due to the regression, and the variation not explained by the regression is
called SSE or the error sum of squares.

Ans: True
Response: See page 527

230
44. The dependent variable in a regression analysis is always plotted on the vertical
axis.

Ans: True
Response: See page 518

45. Bivariate analysis is the same thing as multivariate analysis.

Ans: False
Response: See page 518

46. The Spearman rand-order correlation is appropriate for variables that are not
measured with metric data.

Ans: True
Response: See page 518

47. The least squares estimation procedure often results in a perfectly straight line.

Ans: False
Response: See page 521

48. The R2 statistic ranges from -1 to 1.

Ans: False
Response: See page 526

49. ANOVA is used to test the significance of a regression analysis.

Ans: True
Response: See page 529

50. It is possible to run a regression analysis with more than 1 independent variable.

231
Ans: True
Response: See page 518

Essay

51. Some of the results of a regression analysis are as follows:

SSE = 52,500 SST = 485,200

Compute the Coefficient of Determination (R2) and interpret.

SUGGESTED ANSWER: R2 = (SST -- SSE) / SST

R2 = .8918 which means that over 89% of the variation in
the dependent variable can be explained by the independent
variable.
Response: See page 527

52. Interpret the correlation coefficient for the following data:

Sales Sales Training Score

45,000 98
34,500 74
23,750 57
51,450 99
41,000 85

The resulting correlation coefficient is +.98 and the probability of insignificance

is 3%.

SUGGESTED ANSWER: There is a strong positive correlation between sales and

sales training scores. Hence, higher sales are associated
with higher sales training scores and vice-versa.
Response: See page 530

53. A college professor is trying to predict which students will go snow skiing over
spring break. He hypothesizes that students with higher grades will be more
likely to go snow skiing. From past years he has compiled data which has
indicated that students with higher grades are more likely to go snow skiing.
Y = number of snow skiing trips during spring break

232
 X = grade point average

His resulting model is as follows: Y = -8.032 + 3.893(X)

R2 = .884
P = .000
F = 61.1 significant at .000

How would you interpret the professor’s model? Does grade predict the frequency of
skiing?

SUGGESTED ANSWER: The regression model is significant, with a strong positive

R2 of .884, meaning that 88.4% of the variation in the
number of ski trips taken is explained by the variation in
grades. The F-statistic also indicates a significant linear
relationship. Hence, the professor’s model is a good
predictor of the frequency of student snow skiing.
Response: See pages 525-529

233