What you cannot explain to others you do not understand yourself: producing an account is not something for an audience
only but for ourselves. It is incorporation of disperate elements into a coherent whole. The techniques: diagrams, tables, text.

A good story is like a journey, in which we travel with the characters to arrive at a conclusion.
[Robin Cook's novel "Toxin", John Grisham's novel "The Streetlawyer" given as class examples]

Analysis is also like a journey, and its conclusion can be reached and understood only by travelling on it, but this is not reconstruction of every step.

VALIDITY: a valid account is one, which can be defended as sound because it is well grounded conceptually and empirically.

RELIABILITY: consistency through repetition. If our research is reliable, then others using the same procedures should be able
to produce the same result. (Analytical procedures are typically ill-defined!).

GENERALISATIONS: As a basis for generalising beyond our data, qualitative analysis is more likely to be suggestive than conclusive.

We cannot categorise or link data unless we have first read and annotated it;
We cannot connect categories unless we have first categorised and linked the data;
We cannot produce an account without first categorising and linking the data.

Although qualitative analysis is sequential in this sense, it is more realistic to imagine qualitative data analysis as a series of spirals as we loop back and forth through various phases within the broader progress of the analysis. Thus qualitative data analysis tends to be an iterative process.

The various "stages" of research, which we have presented in logical sequence may be better thought of as  recurrent  "phases" through which the analysis passes.

In qualitative data analysis paradoxes are not infrequent:
- We want to consider data in bits, but also to analyse it as a whole.
- We want to consider data in context, but also to make comparisons.
- We want to be comprehensive, but also selective.
- We want to analyse singularities but also generalise.

Back to home