Model SelectionMethod Sticking with cashforclunkers imagine

Model Selection/Method

Sticking with cash-for-clunkers, imagine you have a data set that includes 957 U.S. metropolitan or metropolitan statistical areas (CBSAs) for which you have a number of economic and demographic outcomes (income, employment rate, new car purchases, car fleet data, population,...) available from January 2004 through June 2010. Imagine you can also construct a \"clunkers per capita\" variable which gives the number of registered cars with MPG of 17 or less per capita in the city. Furthermore assume you can imagine you can divide your cities into \"clunker heavy\" and \"clunker light\" cities where \"clunker heavy\" cities are those with a lot of clunkers per capita and the \"clunker light\" cities are those with very few clunkers per capita. Cash for clunkers ran from July - August in 2009. Describe how you might determine if cash for clunkers increased new car purchases per capita. What are some things that may create problems in your estimation?

Solution

a)

I will determine if cash for clunkers increased then I will doing a regression analysis to find how much money do they earn per year

b) we are not very sure that the data have very good relationship as a linear regression for exaplem