What is your research workflow when it comes to writing papers and using software for citations and notes? (primary relation to literature review)

Posted by Alexander Deliyannis
Mar 16, 2011 at 10:47 PM

Dr Andus wrote:
>- capture website info with Surfulater;
>- capture writing ideas in Notes on iPod Touch/iPad, which syncs with Gmail;
>- collect references in EndNote and link to PDFs on PC hard drive;
>- read and annotate PDFs with PDF XChange Viewer on PC or GoodReader on iPod Touch/iPad;
>- copy quotes and notes from PDF and Gmail and paste them into Whizfolders in order to organise them into hierarchical outline; use Whizfolders as main database of ideas and quotes;
>- use NVivo for analysing other collected material (text, image, audio, video).
>- draft final outlines in Natara Bonsai;
>- do the final writing in MS Word.

Dr,

My workflow was* very similar to yours, but I only used two tools in practice: Surfulater for all reference material and Brainstorm for all writing, apart for the final formatting. I made this choice after trying several tools, including Zoot, UltraRecall, Endnote and Maxthink among others; I liked many of them, but most were redundant or even bothersome in my workflow. This included Endnote; it was for me much more convenient to maintain bibliographies in Brainstorm (even though they were in plain text) as I could link them to quotes and reuse them without disturbing the workflow.

Surfulater is focused on capturing and organising web content, so I grabbed most of my references including articles in HTML form, also linking to the PDF versions in my file system. Highlighting and copying content is far more seamless with HTML than PDFs (e.g. no carriage returns to remove), articles can be included in multiple categories as clones or via tags if one prefers. Linking to the original article from other applications, e.g. Word, is easy via the sulkb:// tag, etc.

Brainstorm is excellent for developing textual content bottom-up or top-down complementarily. It also supports cloning (namesakes) and can identify identical text automatically. I note that it now looks rather dated (it started life as a DOS application after all) and it may have some trouble dealing with huge amounts of multi-cross-referenced text, but I have found nothing as powerful. Nevertheless, I’m now using mostly Sense for top-down writing, as it provides a better overview of the content and is actively developed.

*It’s been a year since I last did any academic writing, though I should be getting back to it soon.