Practice Assignment (to be discussed randomly April 14)

1) 2.109 (p. 215)

2) (p. 699) 10.11a and your expectation for (b), 10.13 but for the information given in 10.11.

3) (p. 170) 2.53

Homework Assignment (due Friday, April 16)

1) A look at a scatterplot indicates that there are 3 points (out of 100) that are far away from the main clusters of points. What impact, if any, would there be on the correlation coefficient if these three points were removed? Explain.

2) Many people who finish high school do not go on to college. Here we look at the percent of people 18 to 24 years old who were enrolled in college each year from 1980 to 1990. We look separately at blacks, Hispanics, and whites. To simplify the analysis we code 1980 as 0, 1981 as 1, and so on, up to 1990 as 10. When we do a regression analysis with the yearly percentages enrolled in college as the response variable and years from 0 to 10 as the explanatory variable, we get these three regression lines:

Blacks: percent = 26.5 + .42 year (r=.71)

(a) Give an interpretation of the coefficient 0.42 for blacks.

(b) The college enrollment percentages of which of the three groups increases fastest from year to year in this period? Clearly indicate how you are determining this.

3) In Lecture 24 we consider the draft lottery of 1970. Let the explanatory variable be the month of the year e.g. x=1 for Jan, 2 for Feb., up to 12 for Dec. Let the response be the mean of the draft numbers for each month. (In Lecture we looked at all 366 days.)

(a) If you make a scatterplot of the data points (x,y) for the 12 months and perform a regression analysis, what values would you expect to get for the intercept, the slope, and the correlation coefficient for a lottery that was truly random? (Hint: May help to draw sketch.)

(b) What do you conclude about the draft lottery when the analysis gives you the following results:

mean draft number = 230 - 7.1 (month) (r=-0.87)

(c) The value of the test statistic t = -5.50. Could this value have occurred by chance alone? (Tell me if the result is statistically significant, showing lots of detail in your significance test.)

4) (p. 169) 2.51

5) Recall the problem on husband's and wife's ages from PA 9.

(a) Based on the output below, is there a statistically significant relationship between the husband's age and the wife's age?

(b) Extra credit: Carry out a test of significance to decide if the husband ages and the wife ages tend to change at the same rate.

6) (p. 220) 2.116 Calculate percentages and make a segmented bargraph to display the relationship.