This article describes how to go from a crosstab with a time series variable...

...to a table with a *lagged* rolling average:

## Requirements

- A table with a date/time variable in either the rows or columns

## Method

- Select your table.
- In the toolbar go to
**Automate > Browse Online Library > Modify Whole Table or Plot > Replace with Rolling Averages**. - Set the
**Number of periods to roll**. - OPTIONAL: Tick
**Dates are in the rows**if applicable. - Press
**OK**.

Please note the following:

- A lagged rolling average is computed. For example, if January has a score of 3, February 4, and March 6, the 3-period rolling average for March is (3 + 4 + 6) / 3 = 4.33333333.
- Where values cannot be computed due to insufficient data, the value is shown as
*NaN*. For example, if January has a score of 3, February 4, and March 6, the 3-period rolling average for the three months, respectively, are NaN, NaN, and 4.33333333. - This rule is only applied to the first statistic selected in each of
**Statistics - Cells**and whichever is appropriate of**Statistics****- Below**and**Statistics****-****Right**. - You may not have a NET in the variable containing the dates.
- The statistical testing results on the table are not changed by this rule. That is, they reflect the results for the original, un-rolled statistics.