Hypothesis Building and Testing
The foundation of hypothesis building lies in discovering links between elements of your data. CAQDAS can help you to discover these links and with graphic facilities display them for you. Two basic ways of organizing and linking your data are available. The first of these involves a hierarchical organization. This is where your data may be classified in a few broad themes then each theme classified into a number of subcategories each of these into sub subcategories and so on. This is similar to the coding used for quantitative data analysis. For example in figure below you may classify your data under the broad themes of values concepts and strategies on reviewing all the data under strategies you may wish to divide them into mature and immature or long term and short term. Your values theme may fall naturally into say internal and external and then you may identify several subcategories such as duty trust and liking.
An alternative approach is to code each section of data as you work through it and later on to discover ways of gathering together like categories that can be grouped together under common headings which can in turn be grouped together in broad themes and so on. For example in the case above you might identify the concepts of duty trust and liking from your reading of the data and then decide to group all these under a common heading of values later you may wish to have distinct subheading under values which you call internal and external allocating to them your original categories accordingly.
The second way of organizing and linking your data involves a network organization. This is where your categories are more flexibly linked to one another without the rigidity implicit the hierarchical theory building. Networks can be rearranged in whatever way suits your developing research project. Some software offers the opportunity to work with both networks and hierarchies with tools for moving between them.