Coding is a first-pass interpretation technique. It serves to organize the data at hand, and provides a useful way of entering these interpretations into a quantitative analysis framework. Before anything else, coding requires that the analyst go over the data, delineate individual segments (i.e. “mark” or “highlight” passages that interest her), and assign keywords (“codes”) to each of them.

In its traditional manual form, a researcher would typically separate each segment into a physical entity; traditionally, an index card labeled with a word or short phrase. This label should be formulated so as to most aptly suggest the relation of the associated data segment with the objectives of the research project itself.

Some types of qualitative data (e.g., open-end responses from surveys, tightly defined interview questions) are already highly structured from the outset. Such material, typically, can be coded without further segmenting of the content. Rather, the survey or interview questions themselves (or some abbreviated variations thereof) serve as coding device in these cases.

While coding is a powerful and quite efficient way for a first-pass traversal of even the large amounts of data, we should not overlook that a frequent criticism of the coding method–computer-aided or manual alike– is that it appears to seek to transform its material into “quasi-quantitative” data. In some ways it can be said that it “drains” the data of its variety, richness, and individual character.

It is commonly agreed that an efficient way to counter such criticism is by thoroughly expositing definitions of codes and linking those codes soundly to the underlying data.

Today, contemporary qualitative data analysis is frequently supported by computer programs.

Such programs enhance the analyst’s efficiency at data storage/retrieval and at applying the codes to the data. They do not, however, supplant or alter the interpretive nature of coding as such, nor do they usually assign codes of their own accord. Some programs, however, do offer a high degree of efficiency when it comes to assigning, editing, managing and revising codlings. The most advanced such programs, e.g. ATLAS.ti, also offer great assistance in work sharing, peer reviewing, and recursive examination of data and codings.

QDA programs in particular help the endeavor of expositing code definitions by efficiently linking codes with their primary data. Their main strength is to provide direct access to the original data segments at any time, thus guaranteeing the full “richness” of data that would be absent from a mere list of codes.

In a sense, qualitative analysis really and fully comes into its own by the use of dedicated computer software that makes it possible to work both on the “primary” and “secondary” level of the data at any time.