To begin to show more concretely how coding is done, we can use an example from a field study of gang life that relied on ethnographic interviews as a source of data (Decker and Van Winkle, 1996). The study began with some theoretical ideas regarding what attracted young men to join gangs, especially that certain instrumental benefit of joining a gang, such as the ability to get protection from others or to join in the sale of drugs, were a major attraction. Then the researcher goes through the transcribed interviews to find words or phrases that seem to represent these behaviors as reasons for joining gangs. Table 16.2 presents some excerpts from the interviews with the key words that indicate a particular coding category highlighted. The people in these interviews were responding to questions about why they got involved in gangs. The researchers then use these codes to assess the extent to which, throughout all of the coded interviews, there is support for their theoretical ideas about the role of instrumental benefits.
There are a couple of things to notice in the coding in both Tables 16.1 and 16.2. First, the coding does not focus on counts of how often things happen but rather descriptions of what is applicable to a particular person or context. The data analysis produces words, phrases, and description as meaning rather than using numbers to extract meaning. Second, there is a close link between the codes and the data: When a code is applied, it is linked to a particular section (word, phrase, sentence, or paragraph) of the raw data. This is important in terms of assessing validity, which will be discussed later in the chapter.