Example of Tables with Missing Data
main topic     interpreting results     session command     see also 

You would like to summarize weight data for individuals in your study by gender and eye color.

1    Open the worksheet EYECOLOR.MTW.

2    Choose Stat > Tables > Descriptive Statistics.

3    In For rows, enter Gender. In For columns, enter Eyecolor.

4    Click Associated variables, enter Weight.

5    Under Display, check Means and Standard deviations. Click OK in each dialog box.

Session window output

You realize there are missing data. You decide to include the missing data into your table calculations to see how the marginal statistics are affected.

1    Choose Stat > Tables > Descriptive Statistics.

2    In For rows, enter Gender. In For columns, enter Eyecolor.

3    Click Associated variables, enter Weight.

4    Under Display, check Means and Standard deviations. Click OK.

5    Click Options, check Include displayed missing values in calculations. Click OK in each dialog box.

Session window output

Interpreting the results

By default, Minitab displays all missing values in your tables, but does not include them in calculations unless you check Include displayed missing values in calculations in the Options sub dialog box. When a data point is missing, by default, the entire worksheet row (observation) is omitted from the calculation.

In this example, you are missing eye color data for 1 female and 2 males. If you do not include missing data in your calculations, the mean weight of females in your study is 122.2 pounds. If you include the weight data when eye color is missing, the mean weight is 122.1 pounds. For males, the mean weight is 158.4 when missing data are not included and 158.9 when missing data are included.

You have to decide which statistic is more meaningful for your application. If any part of the data for an individual is missing, some researchers would eliminate the entire point from the study, while others would use the data they can.