Economics 421/521

Winter 2006

Homework #2

[PDF FILE]

**
**

**Part I. True or false? If false, explain why.**

1. If you wish to use a set of dummy variables to capture six different categories of an explanatory variable, you should use six different dummy variables and a constant in the regression model.

2. The coefficients in the log-linear model can be interpreted as elasticities.

**
**

**Part II.**

3. Consider the following model:

log(Y_{i}) = α + β_{1}(1/X_{i})
+ ε_{i}

What is the elasticity of Y with respect to X? What is the slope, (dY_{i} /dX_{i})?

4. Using the **NFL data set** posted on the class web site, consider the
following equation for the number of games won by teams during the 2004 regular
season:

Win_{i} = α + β_{1}YDPG_{i} +β_{2}FMDi
+ u_{i},

where Win_{i} = the number of games each team won during the 2004 regular
season, YDPG_{i}= the average total yards per game, and FMD_{i} = the number of
successful fourth down attempts.

a) Use Eviews to estimate the equation with OLS.

b) Test the hypothesis that total yards per game has a positive impact on the
number of games won (use a 5% level of significance).

c) Is the coefficient on the FMDi variable statistically significant at the
10% level of significance?

d) What is the interpretation of the coefficient on FMD_{i}? Does the estimated
value make sense?

5. Use Eviews to estimate the following equation:

log(Win_{i}) = α + β_{1}log(YDPPi)
+β_{2}log(PENYDSi) + β_{3}log(TFi)
+ u_{i}

where YDPP_{i} = yards per play, PENYDS_{i} = total penalty yards for the season,
and TF_{i} = total number of fumbles.

a) What is the interpretation of the coefficient on YDPP_{i}? Explain
mathematically.

b) Is the coefficient of PENYDS_{i} statistically significant at the 10% level
of significance?

**
**

**Part III.**

1. Chapter 6, problem 6.14. [**Data set**]

2. Chapter 9, problem 9.5.

3. Chapter 9, problem 9.21. [**Data set**]