Auto Code Dataset - Filter and Group


This dialog box is displayed when you click the Filter and Group button to filter or group row values that will be used to create nodes by the Auto Code Wizard.

For example, if you have a column containing ages, you can use this dialog box to limit the values used, so that only age values between 18 and 65 are used. You can also group the values, so that actual ages are grouped into age ranges. For example, you could create age groups for 18-30 years, 31-49 years, and 50-65 years.

Options on this dialog


For column  This displays the column you selected in the Auto Code Wizard. You can choose to limit or group the row values in this column that will be used to create nodes. You cannot change the column—to select another column, you must click OK or Cancel to return to the wizard.

Under Using, you can select:

  • All values in the column (Default)  All values in the column are used to create nodes.

  • Only values  Select this option if you want to limit the values in the column that are used to create nodes. If you select this option, you must specify your selection criteria on the right.

Under Code contents at Nodes matching you can select:

  • Each unique value (Default)  Nodes are created for each unique row value.

  • Values grouped by  Select this option to group values together and create a node for each group. If you choose this option, then you must also specify how you want to group the values. You can select:

Grouping Description
Equal Interval The difference between the minimum and maximum values in each group is equal. For example, if your survey has respondents aged between 18 and 61, you can use this option to create four groups of equal interval:
  • 18-28 (11 values)
  • 29-39 (11 values)
  • 40-50 (11 values)
  • 51-61 (11 values)
If the values cannot be equally divided, then NVivo adjusts the grouping to make the intervals as equal as possible.
Standard Deviation The values are grouped based on how they vary from the average value (mean). For example, if the average age of your respondents is 30 (with a standard deviation of 9), then you could use this option to create the following groups:
  • 18-20 (up to two standard deviations below average)
  • 21-29 (up to one standard deviation below average)
  • 30-38 (up to one standard deviation above average)
  • 39-47 (up to two standard deviations above average)
  • 48-56 (up to three standard deviations above average)
  • 57-61 (more than three standard deviations above average)
User-defined Interval Select this option, if you want to create your own groupings and specify the start and end values for each group

If you select Equal Interval, you can:

  • Adjust the Start and End values, to create groups that cover values outside the range of values in your dataset. For example, if you are grouping survey respondents by age, and have values between 18 and 59, you can set the start value to 16 and the end value to 65, to create age groups to cover anyone aged 16-65. This could be useful if you later get responses from younger or older respondents

  • Adjust the number of groups that will be created.

  • Click Calculate to show your groupings in the groups preview area. You can change the names of the groups by entering your preferred names in the Nodes column.   If you are not satisfied with the groupings, change the settings and then click Calculate again.

If you select Standard Deviation, you can:

  • Adjust the number of groups that will be created, by changing the Group range.

  • Click Calculate to show your groupings in the groups preview area.  You can change the names of the groups by entering your preferred names in the Nodes column. If you are not satisfied with the groupings, change the settings and then click Calculate again.

If you select User-defined Interval, you can create your own groupings and specify the start and end values for each group:

  • To create a group, specify the Start and End values of the group, and then click Add to add the group into the groups preview area.

  • You can change the names of the groups by entering your preferred names in the Nodes column in the preview area.

  • You must create groups that cover the range of values in the column you are using.

  • You can optionally create groups that cover values outside the range of values in your dataset column. For example, if you are grouping respondents by age, and have values between 18 and 59, you can set the start value to 16 and the end value to 65, to create age groups to cover anyone aged 16-65. This could be useful if you expect to get responses from older respondents in the future.

  • You can remove a group from the groups preview area, by clicking Remove.