Scribd
Upload
382 views4 pages

Case Assignment 6

This analytics report analyzes the relationship between starting career salary and mid-career salary using data from 151 state schools. Four non-linear regression models were tested: quadratic, lin-log, log-lin, and log-log. The log-log model provided the best fit with the highest adjusted R-squared value of 0.770 and lowest standard error. The log-log model shows that a 1% increase in starting salary corresponds to a 1.08% increase in mid-career salary on average, holding all else constant.

Uploaded by

api-661653599
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
382 views4 pages

Case Assignment 6

This analytics report analyzes the relationship between starting career salary and mid-career salary using data from 151 state schools. Four non-linear regression models were tested: quadratic, lin-log, log-lin, and log-log. The log-log model provided the best fit with the highest adjusted R-squared value of 0.770 and lowest standard error. The log-log model shows that a 1% increase in starting salary corresponds to a 1.08% increase in mid-career salary on average, holding all else constant.

Uploaded by

api-661653599
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

ANALYTICS REPORT

TO: BNAD277 PROFESSORS


FROM: ASHLEIGH BURKE
SUBJECT: CASE ASSIGNMENT #6
DATE: 5 NOVEMBER 2022

Introduction
The data used was collected from the Wall Street Journal on 151 state schools measuring both
starting career median salary and mid-career median salary. The intentions of this assignment are
to understand how much of an impact starting career salary has on mid-career salary. We tested
the data through four of the non-linear regressions: quadratic, lin-log, log-lin, and log-log. After
interpreting all four of the non-linear regression equations, we formulated an Adjusted R2
comparison to see which model provides the best fit. We had to calculate the Adjusted R2 in
order to be able to compare the Quadratic R2 to the Log-log R2 ∈dollars.

Data Analysis

Quadratic Regression Equation

^
MCS=−5347.76+1.89 ( SMS ) +6.62 E - 07 ( SMS )
2

Interpretation
^
Marginal Effect =1.89+2(6.62 E - 07) ( 55 000 )
As starting median salary increases by $1000 from $55,000 to $56,000, mid-career median salary
increases by $1,960.69, on average and all else constant. The shape of the quadratic model is
concave up because the 6.62E-07 is greater than zero.
Lin-Log (Logarithmic) Regression Equation
^
MCS=−860837.96+87937.60 ln ( SMS )

Interpretation
As starting median salary increases by 1%, mid-career median salary increases by $879.38, on
average and all else constant.

Page 1 of 4
Log-Lin (Exponential) Regression Equation

^
ln(MCS)=10.22+2.39 E - 05 ( SMS )
Interpretation
As starting median salary increases by $1000, mid-career median salary increases by 2.4%, on
average and all else constant.
Log-Log Regression Equation

^
ln (MCS)=−0.32+ 1.08 ln ( SMS )

Interpretation
As starting median salary increases by 1%, mid-career median salary increases by 1.08%, on
average and all else constant.

150
Adjusted R2: 1-(1-0.77) ( ) = 0.770267
149
For the log-log regression equation, we had to calculate the new adjusted R2 because the
dependent variables are not in the same terms (one is in terms of y and the other is in terms of
ln(y)). After interpreting all four of the non-linear regression equations, we formulated an
Adjusted R2 comparison to see which model provides the best fit. We had to calculate the
Adjusted R2 in order to be able to compare the Quadratic R2 to the Log-log R2 in dollars.
2 2
R Interpretation: R =¿0.7718
We are 77.18% of the way toward perfectly predicting mid-career median salary.

Standard Error Interpretation: Se =0.058


Our predictions of ln(mid-career salary) are off by 0.058 log units

Prediction:
^
ln (MCS)=−0.32+ 1.08 ln (55000 )
^ ln(MCS) 2
11.47 +0.058 /2
e =¿ e ¿
^
MCS=$ 95,959.55

Conclusion
In conclusion, the model that provides the best fit for the data is the Log-Log regression model
because it has the lowest standard error and the highest adjusted R2. We determined this by
formulating an Adjusted R2 comparison to see which model provides the best fit. We had to then
calculate the new Adjusted R2 in order to be able to compare the Quadratic R2 to the Log-log R2
in dollars. From our best model, we found that as starting median salary increases by 1%, mid-

Page 2 of 4
career median salary increases by 1.08%, on average and all else constant. For further questions
of inquiry, email [email protected].

Appendix
Quadratic Regression Output

Lin-Log Regression Output

Log-Lin Regression Output

Page 3 of 4
Log-Log Regression Output

Adjusted R2Comparison

Page 4 of 4

You might also like