The Code Application by Descriptor field charts (including the dynamic descriptors which are designed to support analysis of change over time in you qualitative data) have a number of options and can be some of the most useful visualizations for analysis, interpretation, and communication/presentation of research findings. Essentially, these charts represent the number of excerpts that have been associated with a particular code separately for each sub-group within a descriptor field. In the example below we see charts for each code showing the relative number of excerpts tagged with the specific code broken out by the ‘Mother Primary Language’ descriptor field groups: ‘Bilingual,’ ‘English,’ and ‘Spanish.’
The drop-down menu, allows for an immediate reconfiguration of the charts by selecting any of the list-type or grouped number or date/time descriptor fields in the project.
Other important controls for these charts can be found in the panel sub-header showing radio buttons next to the drop-down menu to switch the chart from relative excerpt count to average weights applied (where used) and check boxes for ‘Hit/Miss,’ Sub-code Count,’ ‘Normalize,’ and ‘%.’ By default, the ‘Normalize’ and ‘%’ boxes are checked.
Hit/Miss option (which defaults to ‘off’) toggles the chart to a display representing the number of cases in each sub-group with one or more excerpts tagged with the particular code
Sub-code Count option (which defaults to ‘off’) essentially serves to ‘collapse up the code tree,’ thus including excerpts tagged with all root codes AND subordinate codes (child, grandchild, …) in the visual. For example, if the Parent-Child code had child codes associated with it in the tree, all excerpts coded with Parent-Child Talking OR any of the child codes would be included in the chart for the Parent-Child Talking code
Normalization option (defaults to ‘on’) adjusts each bar based on the relative number of cases in each sub-group (see below for the normalization procedure). Simply, a graphical representation for code application frequency by sub-group is relatively meaningless if there are unequal numbers of individual cases across each sub-group. For example, in this study, the ‘Spanish’ group for ‘Mother Primary Language’ is disproportionally large (representing 64% of the total sample). Turning off the normalization adjustment results in a possibly misleading visualization. Below is the same chart as above with normalization turned off:
This ‘non-normalized’ chart, as compared to the original, appears to suggest a markedly high frequency of ‘Letter Recognition’ coded excerpts for the Spanish group. Hence, normalized charts provide a more unbiased perspective of the underlying data
The ‘%’ check box converts the chart from a raw count presentation to a percentage basis presentation, as shown in the following snapshot of the same chart with the percentage view deactivate