Analytical Coding: Performing Qualitative Data Analysis Based on Programming Principles

In this paper, we argue that qualitative data analysis software lacks a tool that can be used to fulfill an algorithmic evaluation of conceptualization carried out in qualitative studies. We propose the context-oriented models of coding that conjugate single codes, that is, brief denotations made in natural language, by unusual local relationships called context-fixed elucidation (CFE). CFE is a local relationship between miscellaneous aspects of a case under study. The set of separate CFEs, originated by the analyst during conceptualization and called thesaurus, represents the case as a whole. On the basis of CFE structure and using the thesaurus' single codes as data, there is proposed an algorithm which calculates, without the involvement of the expert, whether there is or not global coherence of single codes used by analyst within the thesaurus. The tool thus obtained emulates for the codes originated in qualitative study the relationships known in the object-oriented programming, such as polymorphism, visibility, encapsulation, inheritance. A probe application of the new tool is demonstrated by the conceptualization of textual evidence. The application was performed with the help of a pilot computer package which architecture is based on the context-oriented models. Thanks to the models, QDAS can obtain special tools that would make researchers' analytical work more intelligible and coherent. The models proposed can find applications outside of research discourse including computer technologies used in various social spheres where people communicate in natural language.