Data was collected on the dependent variable baseball salari

Data was collected on the dependent variable baseball salaries, salaries, in the early 1990s and a multiple regression model was estimated with the variables in the model defined as:

variable name variable label

-----------------------------------------------------

salary 1993 season salary (in dollars)

years years in major leagues

gamesyr games per year in league

bavg career batting average

hrunsyr home runs per year

allstardum ever an all-star (dummy variable)

The result of one of the estimated models is:

a) Based on the statistical analysis and what we discussed in class, which estimated coefficients should we pay attention to and which should we not pay attention to – and why?

b) What is the expected difference in salary for two players, one of whom has one more home run per year than the other player, but is otherwise identical? State your numeric answer in a sentence.

c) What is the effect of a player ever having been an all-star compared to one who never has, ceteris paribus? State your numeric answer in a sentence.

d) Interpret the R-squared number in a sentence.

The result of one of the estimated models is: Dependent variable: salary years gamesyr bavg hrunsyr allstardum constant (intercept) C Std. Err. 23541.92 16120.85 1.46 8878.69 2010.583 4.42 1978.478 1365.283 1.45 55773.88 9687.426 5.76 1022091145145.97.04 constant (intercet 810691 347937.6 -2.33 Number of obs = 353 R-squared =0.5747

Solution

a) Based on the statistical analysis and what we discussed in class, which estimated coefficients should we pay attention to and which should we not pay attention to – and why?

we pay attention to the coefficients which are significant. In this case the variables gamesyr, hrunsyr and allstardum are significant at 5% level, which are to be considered.

b) What is the expected difference in salary for two players, one of whom has one more home run per year than the other player, but is otherwise identical? State your numeric answer in a sentence.

The expected difference in salary for two players, one of whom has one more home run per year than the other player is $55773.88.

c) What is the effect of a player ever having been an all-star compared to one who never has, ceteris paribus? State your numeric answer in a sentence.

A player ever having been an all-star has $1022091 more salary than to one who never has .

d) Interpret the R-squared number in a sentence.

R square =0.5747.

57.47% of variance in salary is explained by the regression model.