- Where Salary Is The Annual Salary Of A Ceo Measured In Thousands Of Dollars And Sales Is Annual Firm Sales Measured In 1 (429.13 KiB) Viewed 182 times
AT THE FIRST QUESTION ASKING ABOUT INTERPRETING THE SLOPE
COEFFICIENT. I HAVE SEEN SOME ANSWERS SAY FOR MODEL i. EVERY $1
MILLION INCREASE IN SALES INCREASES SALARY BY $15000. IS THIS
CORRECT? Is this because 0.015 x 1000000? Which equals
$15000?
What about for Model ii. I saw someone use e function.
How did they get that answer? And why did they use e
function?
In another answer I saw someone wrote for Model iii. For
every 1% increase in sales. Where did they get the percentage from?
Why didn't they use the same for every increase in $1 million as
the other models? Why wasn't the e function used in model iii and
iv? How did they solve for log(sales)?
For question b. How do you distinguish if it's a linear
model?
And for question c what are the different types of
causal effects? Can it be economy or something?
where salary is the annual salary of a CEO measured in thousands of dollars, and sales is annual firm sales measured in millions of dollars. Using a random sample of 209 CEOs for the year 1990, four different functional forms has been estimated using OLS, producing the following results: į- salarý = 1174.5 + 0.015sales ii- log (salary) = 6.846 + 0.000015sales 111- salarý= -898.928 + 262.901log (sales) iv- log (salary) = 4.821 + 0.256log (sales) a) (20 marks) What are the interpretations of the slope coefficients in the estimated models above? Explain in words. b) (10 marks) In which of the regression models above, the relationship between variables (SALARY & SALES) is linear? c) (10 marks) Name one other possible factor that can have a causal effect on CEO salary. Explain the intuition for the causal effect of this factor on CEO salary.