A suggested basic qualitative tool by combining "summarize highlights" and "Split Document"

It seems some users are very interested to use DT3 as a tool for basic qualitative analysis. The key feature is the ability to tag specific text instead of tagging the whole document. While the exact implementation is impossible in DT3, a similar objective can be achieved without using any custom-script but will involve multiple steps. For example, users can use “Summarize Highlights” to consolidate all selected blocks of the highlighted text, that add a split mark for each block within the summarized doc, then use “Split Document” to break the single document into snippets, put them in a group, then tag the individual documents.

An added function in “tools” may help to encourage users to experiment using DT for basic qualitative analysis in fewer steps.


I suggest DT to consider creating a function (or std script) to combine the function of “Summarize and Highlights” and “Split Document”. When this function is invoked on a selection of documents, all highlights in those documents will be extracted and split into “n” documents, each with a page/paragraph backlink to its source. The function/script can also ask/create a destination group for saving the whole set of the split documents. The naming of each document can be the name of the source + the page/paragraph number of the text in the source.

I believe that only moderate modifications, based on the existing two functions, are required? Just an idea/suggestion, not a pressing need.

This is a good idea ngan!

The one thing I would add would be some sort of capacity to tag each "split mark"ed section as it is highlighted.

In this workflow, if I understand your proposal, one would have to highlight first and then revisit all highlighted content in order to tag second. For those working with large documents in projects with a large number of codes/tags (sometimes I will have 100 or more tags, and I believe many researchers will have more), the idea of revisiting each highlighted section and remembering what code you wanted to apply in the original act of analysis might present an undue hurdle.

I’ve been playing with the annotation pane script you recommended. It /almost/ satisfies my need. If it was just a little bit more intuitive with allowing the user to fetch previously used tags (i.e. with a list you could pull from) rather than requiring one to remember and type the tag in every time (again–tricky when ones dealing with 100+ tags!), it would fit perfectly into my work flow.

I haven’t had a chance to use your stacks script yet, which seems a bit more daunting–but also looks quite promising.

Thanks again for your assistance on these matters.