Displaying Descriptive Statistics for Missing Values
Univariate Statistics
Univariate statistics can help you identify the general extent of missing data. For each variable, the following are displayed:
- Number of nonmissing values
- Number and percentage of missing values
For quantitative (scale) variables, the following are also displayed:
- Mean
- Standard deviation
- Number of extremely high and low values
Indicator Variable Statistics
For each variable, an indicator variable is created. This categorical variable indicates whether the variable is present or missing for an individual case. The indicator variables are used to create the mismatch, t test, and frequency tables.
Percent mismatch. For each pair of variables, displays the percentage of cases in which one variable has a missing value and the other variable has a nonmissing value. Each diagonal element in the table contains the percentage of missing values for a single variable.
t tests with groups formed by indicator variables. The means of two groups are compared for each quantitative variable, using Student's t statistic. The groups specify whether a variable is present or missing. The t statistic, degrees of freedom, counts of missing and nonmissing values, and means of the two groups are displayed. You can also display any two-tailed probabilities associated with the t statistic. If your analysis results in more than one test, do not use these probabilities for significance testing. The probabilities are appropriate only when a single test is calculated.
Crosstabulations of categorical and indicator variables. A table is displayed for each categorical variable. For each category, the table shows the frequency and percentage of nonmissing values for the other variables. The percentages of each type of missing value are also displayed.
Omit variables missing less than n % of cases. To reduce table size, you can omit statistics that are computed for only a small number of cases.
To Display Descriptive Statistics
This feature requires Statistics Base Edition.
- From the
menus choose:
- In the main Missing Value Analysis dialog box, select the variable(s) for which you want to display missing value descriptive statistics.
- Click Descriptives.
- Choose the descriptive statistics that you want to display.