The Code Co-Occurrence Table provides information about how the code/tag system was used across all project excerpts. These symmetric, code by code, matrices present the frequencies for which all code pairings were applied to the same excerpt and, by default, overlapping excerpts. Such a display can expose both expected and unexpected patterns in which two codes were (or were not) used together. These patterns illuminate how concepts related to the research questions and represented by the code system are combined in the natural schema (i.e., cognitive frameworks that help organize and interpret information) activated by study participants as they report on the topic represented by project codes.
What does it tell us about our data, research questions, and how respondents naturally connect the concepts we are examining when we discover that that codes ‘A’ and ‘B’ co-occur at relatively low rates compared to codes ‘A’ and ‘C’? Dedoose facilitates the process of addressing questions like these quickly and with a variety of attributes to suit the needs of different researcher preferences. It is also important to note that when including counts for overlapping excerpts, the cell values represent ‘hits’ for excerpts coded with both the associated codes AND excerpt with one of the codes that overlaps with an excerpt coded with other code. This feature can be deactivated by clearing the ‘Include Overlapping Excerpts’ box in the upper right corner of the panel.
For example, the highlighted cell in the table above indicates that 16 excerpts or overlapping excerpts were coded with both the ‘Reading by Others’ and ‘Letter Recognition’ codes. This pairing’s relatively high frequency indicates that as participants are thinking and reporting on one of the concepts, they often discuss thoughts about the other. Such a combination suggests that an overarching schema which includes both concepts are being activated as participants formulate their responses. Drilling down to the underlying qualitative data (by clicking the cell and reviewing the excerpts) provides a deeper understanding of participant reports and the naturally occurring patterns in their thought processes.
Observation of patterns in how the code system was applied can illuminate a wide variety of connections within (a) the nature of the conceptual framework represented by your code/tag system and how it was applied and (b) the nature of the data themselves. Patterns like these are often unlikely to be noticed or understood in the midst of coding activity. Yet in the analysis stage, these patterns can be extremely valuable in discovering and understanding how respondents naturally discuss these concepts in combination and how researchers will discuss these organizing principles as they understand and present their findings.