In qualitative analysis, it is essential that a researcher knows how to link different categories to see:
1) who (or which documents) say relatively same things
2) what factors tend to be 'determinants', and what are 'dependants', according to the informants' accounts
So, linking categories help us to :
1) map out the net of meanings related to the issue under study
2) test and build theoretical model to explain the issue
Nvivo allows you to link a text, or a few lines of text to other sources of information in four ways:
1) link the selected text with a annotation - that is, you just make an internal note, to record something interesting about this document. In this case, the link does not actually 'go out'. The text is linked to your thinking only.
2) link the selected text with a file in a set of files you are working on. Let us say you want to link it with a another document you think says same thing about one issue.
3) link the selected text with a website. This is in case you want to resort to that website later to take some more evidences for your arguments
4) link the selected text with another selected text in another documents. Interesting! In this case, you must have already made a node somewhere else. Then link the text with it.
But what to do with all the links now? Later when you build your model, you will need to import nodes into it. Because the nodes are accompanied with links, all the links can go with them.
Open model explorer in Nvivo, create a New, and then import the nodes for your new model.
Here we see a network of things, thanks to the 'links'. Amazing!
But, that is quite descriptive. Then you should decide the direction for each link. Remember, you have to resort to informants' accounts to make direction of links. Otherwise, the model is contaminated by your pre-defined theoretical ideas. The same rule should apply when you make the links in the earlier step. Just let the informants determine factors they think important for their behaviors.
When you 've done with making link directions, what will you see: A model describing factors related to each other; some factors tend to be the 'core' ones.
Still, what you have done is purely descriptive.
if you are just reviewing a topic, then relax now.
If, however, you are writing a paper/thesis, the above is just the first, but extremely important, step.
What you have to do next is to interpret the 'results'.
Now, go back to your theoretical thinking, sit conveniently in a nice sofa, and ask yourself why the model appears in that way.
If you are thinking in a feminist way: Is this because of gender norms?
If you are thinking in a functionalist way: Is this because each 'factor' in the model has some 'function' in the life of people you are studying?
If you are thinking in a materialist way: Is this because of unequal access to resources?
So on and so for.
Now, go back to your theoretical chapter (if you have written one). You should rewrite it basing on your reflections with the model. Mandatory! The following chapters are just to describe the model components and 'variations' (using Assay function to see differences in ideas between different groups in regard to age, sex, occupation, so on so for. It is like to make a cross-tab table in SPSS. But before that, you should have already grouped different age groups into different 'sets'. Set is synonym for 'group'). Nvivo will, of course, not produce numbers, but texts.
The Assay will help us to decide if there is any 'pattern' according to informants' attributes (being men/women, old/young, unemployed/employed). It even helps you to decide which kind of 'ideas' are more often repeated by showing you the percent of each idea.
Then come to the discussion chapter, write what you have reflected in above task. Check grammar, print it out, and submit it!
Keep discovering Nvivo!
