BAN 304

Homework #5 (Linear Regression) Due: See Blackboard

Problem Solving – Enter answers on Spreadsheet

1. A human resources manager wants to predict the annual salaries of given employees by reviewing
several variables (see spreadsheet).
a. Given the data provided and using the backward elimination process, what estimated
regression equation would you recommend to predict an employee’s salary at a 5% level of
significance? Provide regression equation. (10 points)
MUST SHOW WORK, including ALL regression tables
Put regression tables on this worksheet starting in cell I14 and then put subsequent tables
underneath. Clearly delineate between tables by shading the cells between tables.

Y = 23,177.4726 + 672.325 * Years Employed + 1,916.489 * Years of Education

After 1st regression, remove “# of Prior Jobs” since it has highest p-value (> 5%). After 2nd
regression, remove “Shift” since it has highest p-value (> 5%). Need to move “Number
Supervised” over so remaining variables are in adjacent columns as this is required by Excel.
After 3rd regression, remove “Number Supervised” since it has highest p-value (> 5%). Stop
since all p-values are < 5%.

b. What percentage of the total variation is explained by this model’s variables? Answer as a % to
1 decimal place. (2 points)
74.0%

c. Suppose the HR manager wants to estimate the salary of an employee. The employee has had
3 prior jobs, 12 years of experience, 4 years of post-HS education, supervises 15 people and
works the day shift. What is the expected salary based on the analysis in part a?
Round to nearest whole dollar. (2 points)
$38,911 = 23,177.4726 + 672.325 * 12 + 1,916.489 * 4

d. What other possible factors might you want to include in the model as independent variables?
(2 points)
Any logical answer receives credit

e. A job candidate is asking for a salary of $50,000. The candidate has had 5 prior jobs, 20 years of
experience, will supervise 9 people and work the night shift. Based on the analysis in part a,
how many years of post-HS education should this candidate have to expect this salary?
Round answer UP to nearest whole year. (2 points)
Hint: Use your regression equation. You know Salary (y-variable) and x-variables from your
final regression model. Solve for remaining x-variable.
7 ⇒ ($50,000 - 23,177.4726 – 672.325 * 20) / 1,916.489

1 of 4
BAN 304
Homework #5 (Linear Regression) Due: See Blackboard

2. A Marketing Manager believes sales regions are an important factor in predicting the number of units
sold. Data has been collected for sales from 4 regions (see spreadsheet)
a. Run a Regression Data Analysis based on the regions. (6 points)
i. Select x and y ranges, including headers
• y range will be “Units Sold”
• x range will be Dummy Variables for the regions
ii. Select “Labels” (no other boxes should be checked)
iii. Make Output Range $G$9
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.338424765
R Square 0.114531322
Adjusted R Square 0.051283559
Standard Error 3785.435184
Observations 46

ANOVA
df SS MS F Significance F
Regression 3 77845226.89 25948408.96 1.81083594 0.159872556
Residual 42 601839820.3 14329519.53
Total 45 679685047.2

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 15261.04762 826.0496794 18.47473342 8.81594E-22 13594.01188 16928.08336 13594.01188 16928.08336
Region 1 -3247.247619 1454.410542 -2.232689825 0.030956629 -6182.366923 -312.128315 -6182.366923 -312.128315
Region 2 -1993.547619 1572.751733 -1.267553917 0.211940056 -5167.489115 1180.393877 -5167.489115 1180.393877
Region 3 -892.6190476 1652.099359 -0.540293804 0.591848156 -4226.690535 2441.45244 -4226.690535 2441.45244

b. Based on your answer for part a, what would be the estimated regression equation? (1 point)
Y = 15,261.005 - 3,247.25 * Region 1 - 1,993.55 * Region 2 - 892.62 * Region 3
Where Region 1, Region 2 and Region 3 are binary variables, i.e., either 0 or 1

c. What percentage of the total variation is explained by this model’s variables?

Answer as a % to 1 decimal place. (1 point)
11.5%

d. What would be the estimated sales for Region 2 based on your regression model from part a?
Answer to nearest whole number. (2 points)
13, 268 = 15,261.005 - 3,247.25 * 0 - 1,993.55 * 1 - 892.62 * 0

2 of 4
BAN 304
Homework #5 (Linear Regression) Due: See Blackboard

3. Sample data is provided showing personal income and personal expenditures (see spreadsheet).
a. Construct a Scatter Chart. Chart requirements: (6 points)
i. Include x-axis title – “Personal Income ($)”
ii. Include y-axis title – “Personal Expenditures ($)”
iii. Include chart title - “Personal Expenditures vs. Income”
iv. No vertical or horizontal lines inside chart
v. No decimal places on x-axis or y-axis labels
vi. Commas in values on x-axis or y-axis labels
vii. x-axis should range from $20,000 to $45,000 in increments of $5,000
viii. y-axis should range from $15,000 to $35,000 in increments of $2,500
ix. x-axis and y-axis labels should be expressed in dollars
x. Insert Tick Marks on x-axis and y-axis (Major Type = Inside)
xi. Insert a Linear Trendline and include the equation
Personal Expenditures vs. Income
$35,000

$32,500 y = 0.8983x - 2551.8

$30,000
Personal Expenditures ($)

$27,500

$25,000

$22,500

$20,000

$17,500

$15,000
$20,000 $25,000 $30,000 $35,000 $40,000 $45,000
Personal Income ($)

b. What does this chart indicate about the relationship between profit and market capitalization?
i.e., No relationship, exponential, positive linear, negative linear, etc. (1 points)
Positive Linear

c. Run a Regression Data Analysis. Regression parameters: (3 points)

i. Select x and y ranges, including header
ii. Select “Labels” (no other boxes should be checked)
iii. Make Output Range $E$22
iv. Verify Intercept and Personal Income coefficients are the same as part (a) Trendline

3 of 4
BAN 304
Homework #5 (Linear Regression) Due: See Blackboard

Regression Statistics
Multiple R 0.997690803
R Square 0.995386938
Adjusted R Square 0.995032088
Standard Error 352.8229381
Observations 15

ANOVA
df SS MS F Significance F
Regression 1 349188284.6 349188285 2805.08509 1.4237E-16
Residual 13 1618292.333 124484.026
Total 14 350806576.9

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0%Upper 95.0%
Intercept -2551.809487 550.227559 -4.6377348 0.00046474 -3740.5039 -1363.1151 -3740.5039 -1363.1151
x, Personal Income ($) 0.898258552 0.016960097 52.963054 1.4237E-16 0.86161849 0.93489861 0.86161849 0.93489861

d. Calculate the INTERCEPT and SLOPE using the Excel formulas. Verify Slope and Intercept are the
same as Trendline in part (a) and Regression Analysis in part (b). Answer to 2 decimal places.
(1 point)
Intercept -2,551.8095
Slope 0.8983
Yes, these are the same as part a and in the regression table results in part c.

e. Using your regression model, what are the predicted expenditures for an income of $38,500?
Round to nearest whole dollar. (1 point)
$32,032 = 0.8983 * $38,500 - $2,551.8095
Answer may be +/- $1 depending on rounding used for Slope and Intercept.

4 of 4