How to Perform Column Comparisons With Missing Data in Grid Variable Sets

Pairwise inclusion of missing values is the default option for comparing columns in a grid question. This article shows you the different options for column comparisons.

Requirements

• One or more tables structured as follows:
  • A SUMMARY table using one of the following types of variable sets: Numeric - Grid, Binary - Grid, Nominal - Multi, where the variables are in the columns of the tables.
  • A crosstab where the variable set in the Columns is a Binary-Multi.

Method - Pairwise inclusion of missing values

This is the default option. What this means is that when columns are compared, only respondents with complete data on all the variables corresponding to the columns are used. Casewise deletion is employed if an ANOVA-Based Column Comparison has been specified.

Method - Casewise exclusion of missing values

This involves comparing only respondents with complete data across all columns being compared. This occurs automatically when using ANOVA-Based Column Comparison. It can be made to occur with any table by application of a filter.

Method - Dependent tests

These tests use all available data to perform the test. 

Dependent samples tests are statistical tests explicitly designed for the problem in this example. Two notes of caution about dependent samples tests:

• They are not recognized in the statistical literature. That is, while a number of market research software programs provide these tests, there is no body of published work supporting their validity.
• The tests assume that the data is missing completely at random and, as this assumption is often not appropriate in survey research, these tests should generally not be applied without first checking this assumption. A way to test the assumption is to compare the responses of people with missing data with those without. This comparison is outlined in Example 2 of this article: How to Perform Column Comparisons with Missing Repeated Measures Data.

To apply dependent tests:

1. Select the table or tables in which you wish to modify the statistical testing assumptions.
2. Select Object Inspector > Appearance > Significance > Advanced.
3. On the Test Type tab, change the Proportions test > Survey Reporter Proportions or, if testing means, Survey Reporter Means.
4. On the Column Comparisons tab, change Overlaps > Dependent. 
5. Click Apply to Selection or Set as Default.

Next

How to Apply Significance Testing in Displayr

How to Investigate Your Statistical Significance Testing