PART B EXERCISES B1. For Exercises B1 to B5, you will be using the SPSS dataset Polit2SetB. Begin by running a crosstabs (Analyze ➜ Descriptive Statistics ➜ Crosstabs) for the variables bmicat and...

PART B EXERCISES

B1. For Exercises B1 to B5, you will be using the SPSS dataset Polit2SetB. Begin by running a crosstabs (Analyze ➜ Descriptive Statistics ➜ Crosstabs) for the variables bmicat and hlthlimit. The first is a variable that uses the women’s body mass index (BMI), computed from their height and weight, to form four groups: those classified in the normal BMI range (values under 25.0), overweight (values from 25.0 to 29.99), obese (values from 30.0 to 40.0), and morbidly obese (values over 40.0). The second variable is dichotomous responses to whether or not the woman had a health condition that limited her ability to work (yes is coded 1, no is coded 0). In the first dialog box for the Crosstabs, enter hlthlimit as the row variable and bmicat as the column variable. Click the pushbutton for Cells, and select Observed, Expected, and Row and Column percentages. Click on Continue, then open the dialog box for Statistics. Select Chi-square and also Phi and Cramér’s V. Then click Continue and OK to initiate the analysis. Answer the following questions about the output: (a) How many normal-weight women were expected to have a health limitation, if BMI and health limitations were unrelated? How many actually did have a limitation? What are the two values for women who were morbidly obese? (b) What percent of all women had a health condition that limited their ability to work? (c) What is the null hypothesis being tested in this analysis? (d) What is the value of the chi-square statistic? (e) How many degrees of freedom are there? (f) Are the results statistically significant At what level? (g) Comment on the nature of the relationship. (h) What is the value of Cramér’s V? (i) Based on this effect size value, approximately how much power was there to correctly reject the null hypothesis in this analysis? (You will need to consult Table 4 of Appendix: Tables for Power Analyses.)