Vous êtes sur la page 1sur 6

. *a) Find the average salary and the average tenure in the sample.

How many CEOs are in their first


>year (i.e. ceoten==0)? What is the longest tenure as CEO?
.
. sum salary ceoten
Variable |
Obs
Mean
Std. Dev.
Min
Max
-------------+--------------------------------------------------------salary |
177
865.8644
587.5893
100
5299
ceoten |
177
7.954802
7.150826
0
37
. sum ceoten if ceoten == 0
Variable |
Obs
Mean
Std. Dev.
Min
Max
-------------+--------------------------------------------------------ceoten |
5
0
0
0
0
. summarize ceoten
Variable |
Obs
Mean
Std. Dev.
Min
Max
-------------+--------------------------------------------------------ceoten |
177
7.954802
7.150826
0
37
. *a) The average salary of a CEO is $865,860. The average tenure is 7.95 years. There are 5 CEOs
>in their first year. The longest tenured CEO has been at the job for 37 years.

. *b)
>Does
.
. gen
. reg

Estimate and report the regression results of the simple regression model
this model have a reasonable goodness of fit?
logsalary = log(salary)
logsalary ceoten

Source |
SS
df
MS
-------------+---------------------------------Model | .850907024
1 .850907024
Residual |
63.795306
175 .364544606
-------------+---------------------------------Total | 64.6462131
176 .367308029

Number of obs
F(1, 175)
Prob > F
R-squared
Adj R-squared
Root MSE

=
=
=
=
=
=

177
2.33
0.1284
0.0132
0.0075
.60378

-----------------------------------------------------------------------------logsalary |
Coef.
Std. Err.
t
P>|t|
[95% Conf. Interval]
-------------+---------------------------------------------------------------ceoten |
.0097236
.0063645
1.53
0.128
-.0028374
.0222846
_cons |
6.505498
.0679911
95.68
0.000
6.37131
6.639686
-----------------------------------------------------------------------------. * b) R2 is 0.013 which means that the goodness-of-fit is poor and not reasonable as you want a
>measurement closer to one. The goodness-of-fit measures the fraction of the sample variation in y
>that is explained by x. Therefore, when the value of R2 is equal or close to 0, it indicates a
>poor fit as very little of the variation in the yi as captured by the estimated y.

. * c) What is the predicted percent increase in salary from one extra year of tenure as CEO?
.
. reg lsalary ceoten
Source |
SS
df
MS
-------------+---------------------------------Model | .850907024
1 .850907024
Residual |
63.795306
175 .364544606
-------------+---------------------------------Total | 64.6462131
176 .367308029

Number of obs
F(1, 175)
Prob > F
R-squared
Adj R-squared
Root MSE

=
=
=
=
=
=

177
2.33
0.1284
0.0132
0.0075
.60378

-----------------------------------------------------------------------------lsalary |
Coef.
Std. Err.
t
P>|t|
[95% Conf. Interval]
-------------+---------------------------------------------------------------ceoten |
.0097236
.0063645
1.53
0.128
-.0028374
.0222846
_cons |
6.505498
.0679911
95.68
0.000
6.37131
6.639686
-----------------------------------------------------------------------------. * c) The predicted percent increase in salary from one extra year of tenure as CEO is 0.97%.

. * d) Estimate and report the regression results from a similar model as in b) replacing CEO
>tenure with log sales. How does the fit of this model compare to the one in b)?
.
.
. reg lsalary lsales
Source |
SS
df
MS
-------------+---------------------------------Model | 18.1563922
1 18.1563922
Residual | 46.4898209
175 .265656119
-------------+---------------------------------Total | 64.6462131
176 .367308029

Number of obs
F(1, 175)
Prob > F
R-squared
Adj R-squared
Root MSE

=
=
=
=
=
=

177
68.35
0.0000
0.2809
0.2767
.51542

-----------------------------------------------------------------------------lsalary |
Coef.
Std. Err.
t
P>|t|
[95% Conf. Interval]
-------------+---------------------------------------------------------------lsales |
.2242794
.027129
8.27
0.000
.1707372
.2778217
_cons |
4.961077
.1999597
24.81
0.000
4.566434
5.35572
-----------------------------------------------------------------------------. * d) The fit of this model compare to the one in part b) is significantly better as R2 is now
>0.28 compared to it being 0.013 before.
.
.
. * e) What is the predicted percent increase in salary from a ten percent increase in sales?
.
. display (.2242794 *.1)*100
2.242794
.
. * e) Therefore, the predicted percent increase in salary from a ten percent increase in sales is
2.24%.

. * f) Estimate and report the regression results from a similar model as in b) replacing CEO
>tenure with the companys log market value. How does the fit of this model compare to the one in
b)?
.
. reg lsalary lmktval
Source |
SS
df
MS
-------------+---------------------------------Model |
14.987163
1
14.987163
Residual | 49.6590501
175
.283766
-------------+---------------------------------Total | 64.6462131
176 .367308029

Number of obs
F(1, 175)
Prob > F
R-squared
Adj R-squared
Root MSE

=
=
=
=
=
=

177
52.82
0.0000
0.2318
0.2274
.5327

-----------------------------------------------------------------------------lsalary |
Coef.
Std. Err.
t
P>|t|
[95% Conf. Interval]
-------------+---------------------------------------------------------------lmktval |
.257463
.0354271
7.27
0.000
.1875437
.3273823
_cons |
4.677773
.2651797
17.64
0.000
4.154411
5.201135
-----------------------------------------------------------------------------. * f) The fit of this model in comparison to the one in b) is significantly better as the R2 is
>higher at 0.23 in comparison with 0.013.
.
. * g) What is the predicted percent increase in salary from a ten percent increase in market value?
.
. display (.257463*0.1)*100
2.57463
.
. * g) The predicted percent increase in salary from a ten percent increase in market value is
>2.57%.

. * h)
Estimate and report the regression results from a model where you include simultaneously
>all three regressors, CEO tenure, log sales, and log market value. How does the R squared compare
>to the previous models? Which model do you prefer? Why?
.
. reg lsalary ceoten lsales lmktval
Source |
SS
df
MS
-------------+---------------------------------Model | 20.5672434
3 6.85574779
Residual | 44.0789697
173 .254791732
-------------+---------------------------------Total | 64.6462131
176 .367308029

Number of obs
F(3, 173)
Prob > F
R-squared
Adj R-squared
Root MSE

=
=
=
=
=
=

177
26.91
0.0000
0.3182
0.3063
.50477

-----------------------------------------------------------------------------lsalary |
Coef.
Std. Err.
t
P>|t|
[95% Conf. Interval]
-------------+---------------------------------------------------------------ceoten |
.0117054
.0053261
2.20
0.029
.001193
.0222178
lsales |
.1628545
.0392421
4.15
0.000
.0853995
.2403094
lmktval |
.109243
.0495947
2.20
0.029
.0113545
.2071315
_cons |
4.503795
.2572344
17.51
0.000
3.996073
5.011517
-----------------------------------------------------------------------------. * h) The R2 in this model is the highest yet meaning that the sample variation has been
>significantly reduced by measuring all three regressors simultaneously. I prefer the multivariate
model better as it accounts for more causalities giving the numbers being analyzed greater meaning
as they have a much smaller standard deviation and a higher confidence interval.
.
. * i) Compare the impact of CEO tenure, sales, and market value on CEO wages from the last
>multivariate model with the impacts from each simple regression. Which one do you think is closer
>to estimate a causal impact? Why?
.
. * i) I think the multivariate model has a closer estimate to a casual impact as it evaluates more
>independent variables that effect the affect the dependent variables, if any. Therefore, there are
>fewer variables being omitted in determining the causality.

Vous aimerez peut-être aussi