3. In considering the difference in smoking between men and women, what is the null hypothesis? What is the alternative hypothesis? Is the alternative hypothesis simple or composite?

4 Suppose that we wish to compare the health status of two groups of people. What variable might we use to measure the status? What variables might we wish to control in order to draw the appropriate inferences?

5 If someone reports that the mean weight for fourth-grade boys is 80 pounds and for fourth-grade girls is 78 pounds, what must you know to test hypotheses using the difference of means?

6 If we are trying to relate output to labor inputs and capital inputs using regression analysis, would we expect the coefficients of the regressions to be positive or negative? Why?

7 What are dummy variables? How are they useful in identifying differences among groups?

8 Suppose that you used regression methods to estimate the demand curve for physician visits and found a positive relationship: that is, you found that the higher the price is, the more visits are demanded. What problem has likely arisen? Explain the problem in words. Why might it make statistical inference difficult?

9 Rich people consume more health care services than poor people. Explain two ways one might test this hypothesis.