Training

"Which mean do you mean?"

It’s a common question a trainer must ask in Desktop Reporter classes. A mean or average can be calculated and displayed in two different ways in a data table—as an element on the top or side of a table, or as the cell contents or center of a data table.

As an element on the top or side of a table

When you run a crosstabulation of two variables, a mean can be shown as a row or column. If so, the mean is calculated using the factors or values assigned to other elements (rows or columns) of that variable.

For example, when crosstabulating Age by Gender, you can add a mean as a row in the Age variable.

While editing the variable, you can assign each age category a factor or value to represent the range. In the Figure 1, midpoints of each age range are the factor for the mean calculation:

Figure 1: Mean row and factors are added by editing the Age variable.

If you want to add the mean row only to your current table:

• Select the variable on the Design tab
• Choose Variables | Edit Table Variable
• From the menu, select Categories | Insert Categories
• Move the mean to the "Items to Insert" side of the screen
• If needed, set the number of decimal places

Figure 2 shows how you can add the mean row and change the variable for all future tables:

• Select the variable on the Variables list
• Choose Variables | Edit Variable
• From the menu, select Categories | Insert Categories
• Move the mean to the "Items to Insert" side of the screen
• If needed, set the number of decimal places

Figure 2: Select the mean as a row element in the Age variable.

In Figure 3, the mean row shows a mean for each category of the Gender variable. You can interpret the mean with statements such as "The average age reported by males is 31.7 years" or "On average, females report they are 32.3 years old."

Figure 3: The mean row is shown at the bottom of this table.

As cell contents

You can also show the mean or average as the cell contents of the data table. You can crosstabulate two variables, but instead of showing counts or percentages for each intersection of the side variable (rows) and top variable (columns), you can show a mean of a third variable.

Using the earlier example of Age by Gender, you can change the cell contents to reflect the mean of another numeric variable, Visits.

To change the cell contents:

• Choose Tables | Properties on the menu bar
• Click on the Cell Contents tab
• Move the unneeded cell contents from the right side, Included in Cells, to the left side, Available Items
• Move the mean to the Included in Cells column
• Use the drop-down box to select the variable on which the mean will be based
• Add suffixes or prefixes as needed (i.e., $, %, visits, millions)

Figure 4: The table property "Cell Contents" is set to a mean based on Visits.

In Figure 5, each intersection of an Age and Gender category corresponds to a mean number of visits listed. You can describe these by using statements such as "Males 11-16 years old reported visiting the museum an average of 7.6 times this year" or "Among Females 25 to 35 years of age, the average number of reported visits to the museum this year is 4.9 visits."

Figure 5: Mean numbers of visits for each Age/Gender group are shown in the table.

This mean summary table shows averages for each combination, adding another dimension to your analysis. Other similar analyses could be run using "Total," "Maximum," "Minimum," "Median," or "Mode" as the cell contents providing rich information for analyses.

Summary

Which "mean" do you mean? Remember that there are two options on how and where a mean can be calculated and displayed, either as a table element or as cell contents. You will find each option to be a useful tool in your analyses with Desktop Reporter.