# How to Conduct Column Comparisons between variables in a Number - Multi question

The Column Comparisons statistical testing is not available for crosstabs where the question that is selected in the Brown menu is a Number - Multi question, and it is not possible to create a SUMMARY table for a Number - Multi question such that the variables are shown along the columns to facilitate the Column Comparisons. The methods below work around these restrictions by converting the question to a Pick One - Multi and showing the Column Comparisons and Average in the Statistics - Below.

Note that if your crosstab has the Number - Multi question in the Blue menu and a categorical question in the Brown menu then you can use the column comparisons as normal and you do not need to consider the methods discussed here.

## Column Comparisons for a Number - Multi SUMMARY

- Highlight the variables for the Number - Multi question in the Variables and Questions tab.
- Right-click and select
**Copy and Paste Questions > Exact Copy**. - Click in the
**Question Type**column for one of the copied variables and select**Pick One - Multi**. - Make a new table and show the new question in the Blue drop-down menu.
- Ensure that the variables are placed in the columns of the table by right-clicking on the row labels and selecting
**Swap Rows and Columns**. - Right-click on the table and select
**Average**,**Column Comparisons**, and**Column Names**from the**Statistics - Below**. - Highlight all but one of the row labels in the table, right-click, and select
**Hide**.

The Column Comparisons below the table show the results of the comparisons between the Averages shown below the table.

## Column Comparisons for a Number - Multi Crosstab

It is not possible to have all of the results in a single table, and they must instead be spread across several tables where each table has an appropriate filter applied.

- Perform the steps above.
- Make a copy of the table for each category you would ordinarily place in the rows of the table.
- Apply the appropriate filter to each table.