## Practice Quiz: Regressions

### 1. How are regressions used by marketers?

To develop a brand architecture

To estimate the potential budget options of various marketing ideas

To understand the relationship among many different variables.

To estimate the value of a productCorrect

T-statistic

P-value

Coefficients

R-squaredCorrect

### 3. You decide to run a multiple regression to estimate the price of houses in your area. You include variables like number of bedrooms and bathrooms, lot size, and age of the house. However, you forgot to include a very important variable – the size of the house in square feet. Why is your regression likely to give you unreliable results?

You need to use a different model with just 2 variables

Your results are economically significant but not statistically significant.

You are considering variables that are statistically insignificant.Correct

### 4. A health care company is analyzing blood pressure data and its relationship to patient age and weight. Based on the graph, what is the relationship between age and blood pressure?

The relationship is negative, since blood pressure decreases as age increases.

There is no relationship. With no discernible effect on blood pressure as age changes, the company will need to consider other variables to estimate blood pressure.

Cannot be determined

The relationship is positive, since blood pressure increases as age increases.Correct

### 5. A health care company is analyzing blood pressure data and its relationship to patient age. What should the company expect the sign for the coefficient of age to be when they run a regression with blood pressure as the dependent variable and age as the independent variable?

Negative – Blood pressure decreases as age increases

Positive – Blood pressure increases as age increases

Cannot tell the sign of a regression coefficient from the plot.Correct

T-statistic

P-value

R-squared

Coefficients

58.71

0.95

1.46

0.96

### 8. Based on the output from a regression of age and their corresponding blood pressure, how would you interpret the p-value of less than .01 for the coefficient of age?

There is a 99% chance that a positive relationship between age and blood pressure will be observed in another sample.

There is less than 1% chance that a relationship (positive or negative) between age and blood pressure WILL NOT be observed in another sample.

There is less than 1% chance that a relationship (positive or negative) between age and blood pressure WILL be observed in another sample.

There is less than 1% chance that a positive relationship between age and blood pressure WILL be observed in another sample.Correct

58.71

0

11

6.45

### 10. Based on the output from a regression of age and weight and their corresponding blood pressure, which factor predicts difference in blood pressure?

Neither age nor weight

Age

Both age and weight

Weight

### 11. Based on the output from a regression of age and weight and their corresponding blood pressure, how much will blood pressure change if a person’s weight increases by 1 pound?

The heavier person’s blood pressure would be .33 points lower.

The heavier person’s blood pressure would be .86 points higher.

The heavier person’s blood pressure would be .86 points lower.

The heavier person’s blood pressure would be .33 points higher.

## Week 5 Quiz on Regression Analysis

### 1. What does a regression analysis reveal to a marketer?

The development of a brand architecture

The potential budget options of various marketing ideas

The value of a product

The relationship between a number of variables

The development of a brand architecture

The potential budget options of various marketing ideas

The value of a product

1. Why do marketers typically use regression analysis ?

To estimate empirical relationship between sales and advertising.

### 2. Zodiac wants to sponsor an up-and-coming golf player to endorse its watches. Using regression analysis to predict player earnings as a measure of success, they think player height and earnings might be related. Based on the graph above, what is the relationship between height and earnings–and what should Zodiac look for in a player to sponsor?

There is no relationship. With no discernible effect on earnings as height changes, Zodiac will need to consider other variables when selecting a player to sponsor.

The relationship is negative, since earnings decrease as height increases. Zodiac should look for a short player to sponsor.

Cannot be determined

The relationship is positive, since earnings increase as height increases. Zodiac should look for a tall player to sponsor.

## 2. A sports management company wants to represent the most financially successful basketball players. They turn to regression analysis to identify predictors of players’ salary, but they’re not sure how to interpret the results. What does the graph above show about the relationship between age and salary of basketball players?

The relationship is positive. Salary increases as age increases. The company should look for older players to represent.

### 3. A sports management company wants to represent the most financially successful basketball players. They turn to regression analysis to identify predictors of players’ salary. Using the data plotted above, they run a regression with salary as the dependent variable and age as independent variable. What sign is the coefficient of age?

Negative – Salary decrease as age increases

Cannot tell the sign of a regression coefficient from the plot.

Positive – Salary increase as age increases

increase as age increases

## 3. Zodiac wants to sponsor an up-and-coming golf player to endorse its watches and is interpreting the data about height and earnings in the chart above. What should Zodiac expect the sign for the coefficient of height to be when they run a regression with earnings as the dependent variable and height as the independent variable?

Positive – Earnings increase as height increases

Coefficients

T-statistic

P-value

R-squared

R-squared

1.91

0.559

0.30

0.28

## 5. Based on the output from a regression of golf player hei ghts and corresponding earnings (above), how much does salary increase for a unit increase in height?

0.23

### 6. Based on the output from a regression of basketball player age and corresponding salary (above), how would you interpret a p-value of <.01 for the coefficient of age?

There is 1% chance that a relationship (positive or negative) between age and salaries WILL be observed in another sample.

There is 1% chance that a positive relationship between age and salaries WILL be observed in another sample.

There is more than 99% chance that a positive relationship between age and salaries WILL be observed in another sample.

There is 1% chance that a relationship (positive or negative) between age and salaries WILL NOT be observed in another sample.

## 6. Based on the output from a regression of golf player heights and their corresponding earnings (above), how would you interpret the p-value of .01 for the coefficient of height?

There is 99% chance that a positive relationship between height and earnings WILL be observed in another sample

-12.87

39

0

5.8

## 7. Based on the output from a regression of golf player age and the corresponding earnings (above), what is the predicted value of earnings when age is zero?

1.91

### 8. Based on the output from a regression of basketball player age and points scored per game and corresponding salary (above), which factors predict differences in salary?

Intercept

Points scored per game

Age

Both points scored per game and age

Gender

### 9. Based on the output from a regression of golf player heights and gender and their corresponding earning (above), what is the difference in earnings between male and female golf players?

Male golf players make \$3.9 million more than female golf players

Female golf players make \$3.9 million more than male golf players

Male golf players make about .044 less than female golf players

Female golf players make about .044 less than male golf players

## 9. Based on the output from a regression of basketball player age, points scored per game, and corresponding salary (above), how do salaries change with improvements in points scored per game?

Salaries increase by \$0.27M for every point improvement in points scored per game.

### 10. Based on the charts above, why is height not a significant determinant of earnings once gender is included in the regression?

Gender has only two values, one or zero.

With only 39 observations, not enough data is available to determine the answer.

Height has a non-linear relationship with earnings.

The regression attributes all the variation in earnings to height in the absence of gender.