Picking up the pace: PhD workflows and finding what fits

I’m past the halfway mark of my PhD, and lately have been trying to pick up the pace alongside my other professional commitments.

There’s been a variety of tools and approaches I’ve used (and at times rejected) across my PhD – everything from Power BI, Data Studio, RStudio, BigQuery, spreadsheets, handwritten notes, Scrivener, and Obsidian.

I sometimes find that attempts to optimise workflows can actually reduce efficiency. (The same can be said of spending time writing about workflows – this blog post being a case in point). However, whether efficient or not, experimenting with different workflows, systems, and tools has, over time, helped me find a balance between efficiency and joy in my research approach.

So, I’ve taken some time to reflect on the different approaches I’ve taken in my PhD so far – across knowledge organisation, data analysis, and writing – and considered what I hoped for from them, what actually happened, and what each taught me.

1. Handwritten notes

The start of a new project – and one of my biggest yet. How to approach this with absolute perfection? Overly analytical, handwritten notes, of course. In reality, this was a tried-and-tested approach from the honours part of my undergraduate era, and I probably already knew it wasn’t going to be sustainable. We’re thinking multiple highlighter and pen colours based on the type and category of notes.

It was good while it lasted; however, it didn’t last long. As someone who adores starting new projects, I know that I throw myself into them with a lot of intensity and then need to assess which parts to pull back from as the bigger picture becomes clearer and greater efficiency is needed.

The handwriting of notes itself wasn’t the problem, but the perfect structure I expected myself to bring every time was, especially when working across many other professional projects and responsibilities.

I found I needed a different approach to help me see both the detail and the bigger picture all at once, and be less time-intensive.

2. Literature review spreadsheet

As I set aside (some of) my written notes, I started using a spreadsheet in the early stages of my literature review. Several tabs contained records of research articles, search strategies, grey literature, and drop-down values for each category (and sub-category and sub-sub-category) I was tracking. The ID column for each article was mapped to different search strategies, and vice versa, with fields to record relevance to both the research design and the literature review.

Looking back, it gave me the (editable) structure I needed. However, it may have been an overly convoluted approach for someone who wasn’t conducting a literature review that required such a structured methodology. It did serve its purpose, though, and helped me to understand and track where I was heading. I set aside this spreadsheet after I was well into my literature review and realised my time was now better spent finessing my research questions, engaging with my theoretical framework, and making a start on data collection.

The spreadsheets (and sticky notes) have, however, made a big return as I’ve started the coding stage of my qualitative data analysis.

While I’ve used NVivo for analysis before, I usually switch to other approaches and tools partway through, as it doesn’t align with how I like to think through reflexive thematic analysis coding (and feels too clunky). The less eloquent way of saying this is that NVivo has made me feel like chucking my laptop out the window partway through previous research projects.

For me, there needs to be the messiness of sitting with a pile of transcripts, notes scattered, and designing my own system to just get on with it.

3. Obsidian

I turned to Obsidian a couple of years back and took it well beyond my PhD, thinking it would make an excellent mapping tool for everything. It worked well, especially as I started creating templates for to-do lists. However, I didn’t stick with it for as long as I’d hoped. I suspect this was in part because I was moving interstate at the same time. I needed other types of automation in place for research notes, processes, and literature.

I still use Obsidian, but it’s not currently built into my workflow in any substantive way. Perhaps that will change. It reminded me how much I appreciate seeing structures, ideas, concepts, and systems mapped out visually, though. That works well for someone with an interdisciplinary research project who thinks in a very associative, structural, and systems-type way.

4. BigQuery and Power BI

On the quantitative data side, Power BI was a big part of my first academic library (and leadership) role. I had the chance to try Tableau during my LIS degree, and those skills served me very well when I landed a fabulous position that required data visualisation. I’ve since returned to Power BI in my PhD (and for reporting in my current role too).

After using BigQuery (for SQL queries of the COKI dataset), I initially tried a dashboard with Data Studio. I quickly put that aside because the workflow and design didn’t feel intuitive. After some experimentation – seeing if I could connect BigQuery with RStudio (which I succeeded at, though it was purely as a challenge to see if I could) – I returned to Power BI, which has worked splendidly ever since.

It was a nice reminder of how much I enjoy the process of data visualisation and the clarity it brings not only to data outputs but also to my interpretation and understanding of them. I think the process of creating visualisations itself is where much of my own joy, understanding, and interpretation come from.

5. Scrivener

Most recently, I found myself staring at an unwieldy Word document, with multiple versions, comments, and tracked changes that I could no longer keep track of. Perhaps a Word document for each chapter?

Thanks to Kate Tickle’s excellent recommendation, I tried Scrivener instead.

I spent an evening (actually, well into the night) breaking down my thesis document into manageable sections in Scrivener.

And with that, I found myself revisiting my chapter structure, drafting how my methodology contributed to a practice-theory governance model, refining both conceptual and theoretical frameworks, and realising I had a much clearer mindset.

While I do a lot of writing, I think visual approaches are the key here. They leave me feeling like I’m building something – which I always enjoy – and help me to clarify and communicate how all the connections I encounter through my writing and workflows do actually intersect (and matter). It’s also been about finding systems that match how I think things through and channelling my energy into them.

Seeing each chapter (one at a time) without entirely separating them helped me to better visualise the narrative across them, which is (again) helpful for an interdisciplinary, mixed-method project that needs to be coherent on the other side. I can see myself writing the rest of my thesis in Scrivener going forward (and I’m sure I still have a lot to learn about its functionality).

For now, though, I have some interview transcripts to finish coding before the end of the year.