A NET is a combination of categories on a Categorical Question. A NET is always computed in the same way: it represents the proportion of respondents (i.e., cases) to have selected one or more of the categories and have no missing data.
NET with mutually exclusive categories
With mutually exclusive categories, such as gender, a NET is the sum of categories: 51% males + 49% females = 100% NET. Thus, with Pick One and Pick One - Multi questions, a NET is in essence a total.
NET with categories that are not mutually exclusive
The following table shows awareness data, where people were permitted to be aware of 1 or more brands (it is referred to as a Pick Any or Binary - Multi question). Note that the NET here is not a total. Where the categories overlap, the NET will always be less than or equal to the sum of the categories, except where there is missing data (this is discussed in more detail below).
NET with categories that are not exhaustive
A NET is only 100% if the categories it represents are exhaustive. The example below shows the ages of children in families. The NET is not the sum of the categories above, because people can have more than one child. And, it is not 100% because people in the sample have no children.
NET with missing data
Cases with any missing data are excluded from the NET. In the table on the left, all the brands have a sample size of 327 except for Pepsi Light, which has a sample size of 314 due to 13 cases with missing data (i.e., Missing n). In this example, the NET is computed using only the 314 cases that have complete data (i.e., Pepsi Light and all the other brands).
In the table to the right many of the brands have missing data, and as a total of 21 cases have some missing data, the Base n (or Sample Size in Displayr) used in the computation of the NET is 306.
By definition, the Base n/Sample Size for the NET must always be less than or equal to the smallest Base n/Sample Size shown in a SUMMARY table. For the same reason, the Column n (Column Sample Size in Displayr) shown for the NET row in a crosstab and shown in Statistics - Below must always be less than or equal to the smallest Column n/Column Sample Size in that column.
See the following articles for a discussion of options when you wish to display a different sample size:
- How to Investigate a Sample Size or NET that is Too Small in Displayr
- How to Investigate a Sample Size or NET that is Too Small in Q
Creating a NET
NETs are automatically created when a project is set up in Q and Displayr. Further NETs can be created by selecting multiple categories, right-clicking, and selecting Create NET (see Manipulating Data).
The NETs that are automatically created (i.e., which, by default, are shown at the base and to the right of tables) are no different from NETs that are manually created using Create NET.
See also
How to Investigate a Sample Size or NET that is Too Small in Displayr
How to Investigate a Sample Size or NET that is Too Small in Q