Updated: Dec 15, 2020
Yet again this week I was asked the red flag question in a CAQDAS workshop: “Coding’s done. Now what?” This flags the inappropriate use of CAQDAS: no analytic planning done before plunging into helter-skelter coding. In this post and the next I’ll deal with the two underlying problems: starting to code without thinking about its purpose, and thinking of coding as an event rather than a process. Taken together these can result in a mass of codes that don’t lead to a thoughtful response to the research question. First: how to think about the purpose of coding.
Analytic planning is not inconsistent with the spirit of qualitative research
Planning does not mean pre-determining. All forms of qualitative analysis are open-ended and unpredictable to varying degrees. But this doesn’t mean we don’t plan what we do. It’s true a whole qualitative analysis can’t be planned in one fell-swoop at the outset, but we still make plans. It’s just that we’re open to and expect the plans to change as the analysis unfolds. So planning is essential.
Qualitative analysis is a journey and so we need a map
An analytic plan is the strategy, a way of knowing what we’re going to do before starting to do it. Just like you don’t start driving without knowing where you’re going or a plan for getting there, every qualitative analysis needs an overall plan. This doesn’t mean a fixed step-by-step recipe. On any journey there are bends and turns in the road, and half way you might decide it makes sense to detour and visit a long lost friend – the process isn’t linear. But there still has to be a map and a plan, it just gets adjusted often. In qualitative research the analytic plan is some kind of conceptual framework that leads to specific analytic tasks that develop iteratively as the analysis proceeds.
Planning means thinking about what you anticipate is next
There are always alternative routes to a destination. Choosing the next road to take depends on where your earlier choices have brought you, and where you need to get to next. Planning a qualitative analysis is informed by what has happened up to this point, and crucially, what you anticipate doing next. These are what Nick and I call the ‘context of analytic tasks’ that determines the planning for how and what to code.
Coding before planning?
“Coding” isn’t a single thing, there are many different approaches (see for example, Johnny Saldana’s The coding manual for qualitative researchers). The mechanical process – what we call the tactics – is simple and straightforward. Coding can be manual, with highlighter pens and scribbled margin notes on hard copy transcripts as the tactics, or accomplished with general purpose software likes MS Word or Excel, or with a dedicated CAQDAS program. You highlight a segment of data and label it according to what it represents. The label might be a keyword, a descriptive phrase, an abstract concept. These are very different, and deciding what the codes represent – as well as what to code, when to code, and why to code – depends on the purpose. This is not straightforward and depends on the strategies – what has happened to this point, and what we anticipate doing next.
Why would researchers code before planning?
Understanding why this happens will help the teachers amongst us work out how it can be addressed. Is it common? Is it related to the use of CAQDAS packages or digital tools in general? We have a responsibility to prepare the current generation of researchers that works in an increasingly digital context. Here are three thoughts.
Coding as an end rather than a means
Coding is a crucial analytic process in many methodologies, but I disagree with Johnny Saldana who says that “coding is analysis” (2016:9). Saying that the coding is ‘done’ rarely means that the analysis is ‘done’. Coding is a means to an end, not an end in itself. We still need to plan what we're going to do with codes – coding is a process, not an event. Much more on this big issue in the next post.
Mis-reading grounded theory
The influence of grounded theory is profound. Even researchers not conducting a grounded theory study are influenced by its principles, particularly of being open to seeing what is in the data rather than being unduly influenced by existing theoretical constructs. But can this lead unconsciously to failing to plan? Researchers’ tell me they’re “waiting to see what happens” because they want to “be led by the data”. Nothing wrong with that, but if it leads to not planning an analysis, then there’s no wonder the red flag question comes up – “Coding’s done. What’s next?” Knowing what to do next doesn’t miraculously appear before us when we’re ‘done’ with coding. Analytic planning is crucial.
The CAQDAS package will take care of it
MS Word doesn’t tell us how best to structure a thesis or report; Excel doesn’t tell us which graph best illustrates a finding; and SPSS doesn’t tell us which is the most appropriate statistical test to perform. Likewise, a CAQDAS package won’t tell us what we should do with our codes once they’re created and applied to our data. But the “what next?” question implies that it should, a sense that “if only I knew the software better I’d know how to proceed.” CAQDAS programs are not methods of analysis but a suite of tools to pick and choose from as appropriate for the method being adopted. Analytic planning is what lets you know how to proceed. However, CAQDAS programs help here by making any lack of planning more visible and obvious.
Five-Level QDA as a framework for analytic planning
Nick and I developed the Five-Level QDA method to enable new users of CAQDAS programs to quickly develop the expertise they need for their idiosyncratic projects without going through the years of trial and error that we did!
The method unpacks what expert users unconsciously do when they translate their analytic strategies into software tactics.
We developed Analytic Planning Worksheets to provide an adaptable framework for analytic planning and learning to quickly master how to accomplish analytic tasks using CAQDAS tools. The worksheets are designed as a learning tool, and needn’t be used once the skill of translation between strategies and tactics has been developed. But feedback when we tested different iterations of the worksheets in our CAQDAS workshops over the past few years suggests that many students and researchers find them valuable for planning and documenting their analysis even after they have mastered the process.
Now that our books are out, we’re starting to think more deeply about the implications of this, and would love to hear what others think.