Statistics Homework

Exercises 8.1

1. Calculate expected values for the following two-by-two tables:

(a)

		25
		42
34	33	67

2. Calculate the chi-square values for the following table

(a)

18	7	25
16	26	42
34	33	67

3. Determine whether you would accept or reject the hypothesis of independence for the two variables in each of the tables in (a) of Exercise 2.

4. Calculate conditional and marginal probabilities for the cross-tabulations in (a) of Exercise 2 and give them an interpretation.

Exercises 8.2

3. The data on the Hospital Stay worksheet of Chpt 8-1.xls show a sample of one hundred hospital stays by Sex and LOS. Stays >3 days are coded as Long; those 3 days or less are coded Short.

a. Use the pivot table capability of Excel and the chi-square to test the hypothesis of independence between Sex and Stay as coded Long or Short.

b. Use conditional probabilities to provide an interpretation of the results.

7. The Late Delivery worksheet of Chpt 8-1.xls shows the reason for late delivery of the noon meal in three wards of a hospital to be problems with personnel, the facility, or the patient.

a. Use the pivot table capability of Excel and the chi-square to test the hypothesis of independence between ward and the reason for lateness.

b. Use conditional probabilities to provide an interpretation of the results.

Exercises 11.1

1. Use the data on the Hospital charges worksheet of Chpt 4-1.xls and generate the following xy charts. (The first variable in each case should be considered the independent variable.)

a. Age with LOS

b. Age with Charges

c. Charges with Medicare

2. Which, if any, of the charts in (a), (b), or (c) (preceding) appear to be the best to fit a linear model and why? Which appear to be the worst?

5.(SEE ATTACHED IMAGE FOR EXERCISE 11.1, 5)

7. The data on the U5Mort worksheet of Chpt 11-1.xls to represent female literacy (FemLit) and under-five mortality (U5Mort) for one hundred and fourteen countries of the world, each having more than one million inhabitants. The data were taken from the State of the World’s Children (2001). Use these data to calculate b 1 and b 0 for FemLit as a predictor of U5Mort, following the example in Figure 11.5.

Exercises 11.3

1. Use the data from the Charges worksheet of Chpt 11-1.xls and do the following:

a. Generate the regression analysis by using the Tools --> Data Analysis --> Regression add-in and replicate the analysis shown in Figure 11.10.