# run a series of statistical tests to see if missingness is related to other characteristics of the women in this sample, beginning with t tests.

run a series of statistical tests to see if missingness is related to other characteristics of the women in this sample, beginning with t tests..

If you have access to the SPSS Missing Values Analyses, run the MVA with the same set of variables and use Little’s MCAR test to see if you reach different conclusions based on the results of that test than you did based on results from Exercises B6 and B7.

Exercise B7

We will continue to test whether missingness on the CES-D scale is related to mothers’ characteristics, this time with chi-square tests to test differences in proportions on categorical variables. Select Analyze ➜ Descriptive Statistics ➜ Crosstabulations. Use cesdstat as the column variable, and the following variables as row variables: race/ethnicity, educational attainment, current employment status, and marital status. Select the chi-square test as the statistics option, and request observed column percentages in the Cells option. Answer the following questions based on the output: (a) What is the extent of missing values for the four variables in these analyses? (b) Were there significant differences between women who did and those who did not have a CES-D score on any of the variables in these analyses? (c) Based on our findings in Exercise B6 and B7, would you infer that the pattern of missingness is MCAR, MAR, or MNAR?

Exercise B6

Next, we will run a series of statistical tests to see if missingness is related to other characteristics of the women in this sample, beginning with t tests. Select Analyze ➜ Compare Means ➜ Independent Samples T Test. Move the following quantitative variables into the Test Variable list: Age at the time of the interview; Age at first birth; # of children living in HH past month; Family income, all sources; Number of types of abuse of 4 mentioned; SF-12 Physical Health Component Score; and SF-12 Mental Health Component score. Then, enter cesdstat as the Grouping Variable and click the Define Groups pushbutton. Enter the value 1 for Group 1 and the value 2 for Group 2, then click Continue. Click the Options pushbutton and make sure that Missing Values is set to: Exclude cases analysis by analysis (i.e., pairwise deletion for these tests of missing versus nonmissing CES-D values). Then click Continue and OK to run the analysis. Answer the following questions based on the output: (a) We know from the previous exercise that there are no missing values for the new variable cesdstat. What is the extent of missing values for other variables in these analyses? (b) Looking at the column labeled Sig. (2-tailed) in the Independent Samples Test table, were there any significant differences between women who did and did not have a CES-D score for any variables in the analysis? If so, for which variables?

run a series of statistical tests to see if missingness is related to other characteristics of the women in this sample, beginning with t tests.

Posted in Uncategorized