## Introduction

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.

## Method

There are a variety of different assumptions that can be made regarding missing values when conducting tests on SUMMARY tables of the following type of grid tables

- Number - Grid
- Pick Any - Grid
- Pick One - Multi where the variables are in the columns of the tables. (NB: prior to Q 4.3.1, casewise exclusion was automatically applied.)

## Pairwise inclusion of missing values

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

## Casewise exclusion of missing values

This involves only performing comparisons using respondents with complete data on all the 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.

## Dependent tests

These tests use all the data that is available for performing the test. To apply dependent tests, make the following changes in Statistical Assumptions:

- Change the
**Proportions**setting to**Survey Reporter Proportions**. - Change the
**Means**setting to**Survey Reporter Means**. - Where a Pick Any question is in the Brown Drop-down Menu or analyzing a Number - Grid, Pick Any - Grid or Pick One - Multi SUMMARY set
**Overlaps**to**Dependent**.

## NEXT

How to Modify Significance Tests

How to Show Statistical Significance in Q

How to Modify Significance Tests Using Rules

How to Apply Significance Testing in Q

How to Interpret Column n with Missing Data

How to Ignore Missing Data in the Column Sample Sizes of a Table

How to Apply Significance Testing to Grid Tables with Lots of Missing Data

How to Recode Missing Values in Q

How to Perform Column Comparisons with Missing Repeated Measures Data

## Comments

0 comments

Article is closed for comments.