第一题
加入医疗保险的人(INSURANCE=1)和没有加入医疗保险的人(INSURANCE=0)
分析二者医疗费用支出的差异

where Y is a medical expenditure and X is a vector of other factors.问题（1）这个回归模型可以用OLS做吗？理由是什么？（2）回归模型中包含other factors(X)的理由是什么？另外other factors的选定标准是什么？

第二题

The following logit regression model is from Robinson and Min (2002), “Is the First to Market the First to Fail? Empirical Evidence for Industrial Goods Businesses," Journal of Marketing Research, Vol. 39, 120-128.

where Survive is a dummy variable set to 1 if the business survived for five (or ten) years after the Thomas Register's first listing, 0 otherwise; P a dummy for market pioneers; F a dummy for early followers; Lnlead the logarithmic value of the lead time; and X a vector of other factors for the survival.Write a null and alternative hypothesis which can answer the following questions, using the coefficients in the above model.

问题

(a) When pioneer lead time is held constant, do market pioneers have a greater chance than early followers of surviving?

(b) Dose increasing pioneer lead time increase the market pioneer's chance of surviving?

(c) What is the impact of delayed market entry on an early follower's chance of surviving? Is the relation linear or inverted-U shaped?

(The linear relation means that delayed market entry continuously decreases an early follower's chance of surviving. In constrast, the inverted-U relation indicates that delayed market entry initially increases an early follower's chance of survival, but any additional delay decreases the chance.)

(d) If  (linear), what do you expect about the sign of ?

If  (inverted U), what do you expect about the sign of ?

(e) Write the logit model using the estimates in Table 5.

(f) Based on the estimation results in Table 5, make a conclusion about the arguments in (a)~(c).

(Use a significance level of 10%.)