Question

Use the following information for questions 11 through 14. In a paper, “The effect of marquee players on sports demand,” by Todd Jewell, the author regresses US soccer attendance on various attributes of the team, stadium, opponent and city, and also on the presence of some particularly famous players, like Beckham and Blanco. The dependent variable is log attendance; so the coefficient should be interpreted as a percent change. That is, a coefficient of 0.43 means the effect of a one unit change is 43%.

11. The coefficient on Blanco (a dummy variable equal to 1 if Blanco is playing in the game) is 0.120 and the standard error is 0.036. The coefficient should be interpreted as:

13. The coefficient is?

12. The t-statistic is approximately?

14. In a second regression the author distinguishes whether Beckham is playing for the home team or the away team.   His results are

Variable                              Coefficient          Standard error

Beckham Home                    0.082               (0.046)

Beckham Away                      0.388              (0.052)

Which statement best describes the results?

Question 14 options:

	Beckham draws significantly more fans when he is playing for the home team;
	Beckham draws significantly more fans when he is playing for the away team;
	Beckham draws the same amount of fans whether playing for the home or visitor;
	Beckham draws more fans when playing for the away team, but the diffierence is not statistically significant.

15. The paper on effort over the duration of baseball contracts found that players shirk more when they

Question 15 options:

	a. are in their rookie seasons
	b. are playing in the playoffs
	c. have more years left on their contract
	d. (b and c) only

Answer 1

11. The coefficient can be interprieted as the presence or absence of Blanco is changing the attendance 12% and the value can be changed upto +/-0.036. Since, log(attendance)=a+b*Blanco; where Blanco=1 if present & 0 otherwise. So easily we can write, log(attendance)=a; when Blanco is absent. Hence,

log(attendance)_Present - log(attendance)_Absent= b . So the change is b which can be interprieted as 12% change in attendance.

12. Though the question 13 is in advance, I think to answer question 12 first. In regression, the t statistic is measured as coefficient divides by its standard error. Here, our coefficient is 0.120 and its standard error is 0.036. Hence, t= (0.120/0.036)=3.3333 (approx).

13. The coefficient is 0.120.

14. This question could be answered well if we had the sample size. Unfortunately we do not have. Anyway we can calculate the t for both Beckham home and Beckham away. t_home=0.082/0.046=1.78 and t_away=0.388/0.052=7.46. At 5%, 10% level of signifacance we can say "Beckham draws significantly more fans when he is playing for the away team".

15. I dont think it does not have sufficient information.