Thread by @JMooneySomers

Important process differences within #thematicanalysis family: small q (deductive, coding reliability concerns, qual data in quant thinking) vs big Q (data driven, open flexible organic coding), middle q (training with @ginnybraun @QRNHub)

But what’s a theme @ginnybraun? Domain summaries - cluster responses to a q or issue but lots of meaning variation within eg a single theme that covers ‘risks and benefits of x’ #thisisbad

Vs meaning based themes - an underlying idea or concept that holds the data together (data may look superficially different but is united by the idea) #thisisgood @ginnybraun @QRNHub

Major insight for me (but only 930am): is a theme something you know in advance what you’re looking for (you look for risks and benefits) vs something you evolve through engagement with the data (could not have anticipated the output). #thematicanalysis @QRNHub @ginnybraun

Themes as inputs is a major issue for #thematicanalysis in medicine and health - I see the inputs approach a lot (eg look for risks and benefits). My students definitely struggle bc they are so solution focused. Is this bad #qualitative? @QRNHub @ginnybraun

Something I need to understand better - #thematicanalysis having an experiential vs critical orientation and potential conceptual clash of doing both. Also critical orientation feels like why you’d use #discourseanalysis

Themes as discovery (archeologist) vs crafting (sculptor) analogy is familiar. Helpful to think of final analysis as a complete product (found everything there is to find) vs more like a defensible analysis that you’ve critically interrogated but is not final...

But not as simple as knowing your assumptions - I speak with a lot of HDR who are navigating the assumptions of supervisors / chief investigators + publishing expectations - they make pragmatic decisions not necessarily in line with their personal assumptions

And so to some practice - we wrote down our assumptions and reactions to the topic of the dataset first. And really spent time on familiarising: what assumptions underpin MY analytic observations, is what was familiar/ unfamiliar. Really useful

Great insight from my table mate @kari_lancaster : it’s hard to see something new or unexpected in the data when you are so familiar with the topic. Any tips twitterverse? #thematicanalysis

Things you were excited about in familiarisation may turn out to be nothing much when you actually code - @ginnybraun #rubberhitsroad #thematicanalysis @QRNHub

Coding captures something of interest to the analyst -> code more then once (radical!) as you may see new interesting things @ginnybraun #thematicanalysis #neverpassive

Interesting process experience - individually generated codes then worked as a group on those codes. Our codes got richer, more productive #teamwork #thematicanalysis @ginnybraun @QRNHub

Coding is a discipline - don’t jump ahead, make yourself stick with it. Risk only seeing your initial impressions, better to engage in a close interrogation. @ginnybraun No doubt this is hard! #thematicanalysis @QRNHub

Conversation at our table about who does the coding in #thematicanalysis - you, a collaboration or your research assistant? Coding is not an analytically neutral process so it matters who does it. How do people who use RAs manage / conceptualise this? @ginnybraun @QRNHub

open organic coding all your data, never a coding frame #thematicanalysis. I tell my students to do this for some transcripts to develop a flexible coding frame hmm. this undoubtedly driven by large datasets, tight timelines, audit-ability culture #challenged @ginnybraun @QRNHub

You are not delivering your data set as your output - a great reminder from @ginnybraun that your job as an analyst is to pick a story to tell of your data (one of many, maybe you’ll tell other stories later) #thematicanalysis @QRNHub

Gee that was a super useful #thematic workshop - thanks @ginnybraun for sharing your experience and wisdom. And thanks @QRNHub for facilitating