ConnectedText; any case studies?

< Next Topic | Back to topic list | Previous Topic >

Posted by Miles Taub
Mar 26, 2012 at 08:05 PM

This is a terrific post and provides me with several excellent ideas on a project I’m working on now. I’m working with an RTF file of an annual financial report (10-K) and the ideas in the good Dr.‘s post help enormously.

Miles

Posted by Daly de Gagne
Apr 2, 2012 at 12:09 AM

Dr Andus and Steve, thank you for your posts. I apologize for not being more timely in replying.

OK - based on what you both have written, and Manfred’s piece on how he uses ConnectedText, I have just purchased the program.

I am somewhat scared - because I have never been able to figure it out before, but am convinced if I can stick with it, it may be the one-app solution for my note-taking, writing, research, personal data etc.

If you gentlemen can bear with me as I learn how to use CT, both Steve and I may end up with some good material for posts on our respective blogs.

Thanking you in advance,

Daly

Stephen Zeoli wrote:
> >
>Dr Andus wrote:
>>Then came the next phase, which meant actually starting the work
>in CT, without
>>spending too much time reading the Help file. And I’m progressing
>step-by-step,
>>looking up the Help file or the CT forum or actually asking for help as
>and when I need it.
>>And so learning about CT has been incremental and totally in
>parallel with inventing
>>my own version of it (i.e. my own desktop layout,
>arrangement of tools, creating
>>logical relationships between documents
>(topics) and categories etc.). So I
>>recommend 1) finding a problem you need to solve
>(a writing or organisational
>>problem), 2) get stuck into CT, and 3) learn about the
>features as you go along and ask
>>for help when needed.
> >Daly, I agree with Dr Andus’s
>approach for learning or growing into CT—especially number 1. If you have an idea of
>where you want to get to, that does probably make it easier.
> >But there is also the
>approach of just starting out thinking of CT as a collection of note cards. Create
>notes as needed and don’t worry about the wiki part at first. You can open and dock the
>Topics list, which gives you a flat index of all your topics. Now you’ve got what is a
>pretty conventional note manager with editor and index of topics. Also, the search
>functions are pretty robust, so I don’t think you’ll be in any danger of not being able
>to get to your information. You can then just slowly teach yourself some of the
>features of CT, making your database more sophisticated as you go along. You can also
>use the Auto Link feature, which sniffs out phrases that match Topic titles and
>creates links for you.
> >Steve Z.

Posted by Dr Andus
Sep 18, 2012 at 12:36 PM

Dr Andus wrote:
>Most recently I’ve started using CT as a qualitative data analysis solution, to code
>textual data, for which I used NVivo in the past. This is what I do:
> >1. I take a 20,000
>word document (a transcript of an interview) and paste it into CT as a new ‘topic’
>(document).
>2. I dock the table of contents window on the left, and have the edit view
>of the document on the right of it.
>3. I start reading through the document and “code”
>it by adding in headings (up to 5 levels).
>4. As headings are added, they start showing
>up in the TOC pane in the left, so I can see the hierarchy of the themes (codes).
>5. When a
>large enough thematic group emerges (under a top-level heading), I use the “cut to new
>topic” command to remove that chunk of text from the current topic, so it becomes a
>topic of its own. This way the text I’m working on is gradually reducing in size, and
>eventually becomes the central (home) page from which the coded topics become
>linked.
>6. I open the Navigator pane to see the relationship between the ‘home page’
>and the associated coded pages (between 5-10 documents).
>7. Then I open the Topics
>and Categories panes and dock them to the right-hand side of the CT window. Then I
>proceed adding the newly created topics (documents or pages) to the relevant
>categories.
>8. I then review each newly created coded topic and write a conclusion
>section, which contains the conclusions drawn from the given material, basically
>the findings of the research.
>9. Once I’ve done that for each new topic, I return to the
>‘home page’ of this group of topics (which was the topic I started out with but which now
>only contains the links to these coded topics) and I use the “including parts of
>topics” command to incorporate all the conclusion sections from the coded topics.
>Essentially I’m extracting (or abstracting) the findings of the various
>sections.
>10. Once my topic home page contains the extracted findings, I then
>consolidate these findings into a final set of findings (another level of
>abstraction).
>11. As a final step, I use the “including parts of topics” command to
>extract this final set of findings and include them in my “Findings” topic, which
>should be the top level findings page for the entire research project.
> >So basically
>what I have done here is I have carried out a qualitative analysis of textual research
>data, by “coding the data” (thematising it), and then carry out several operations of
>abstraction, by drawing out and consolidating the research findings. I like to think
>about it as a “bubbling up” process, as I’m going from the particular text (the
>interview transcript) and I gradually move to a more abstract (higher) level, by
>dragging out the findings, reaching eventually the top level of abstraction, which
>will constitute the theoretical contribution of my study.

I will eventually get started with the promised CT tutorials, but in the meantime, before I forget, there is another (perhaps less daunting) way of doing qualitative analysis in CT, which might be quicker than the above process.

1. Have Outline pane docked on the left (or right, if you prefer), and the View pane on the right (or left), roughly half-and-half.
2. Use the View pane to review the documents (topics) you want to analyse.
3. Use the Outline pane to start recording your codes and analysis in an outline hierarchy. If necessary, use the Outliner’s “Link to Topic” feature to create a link between the “code” (outline item) and the topic.
4. Organise the resulting codes into a meaningful hierarchy.

This is a manual way of doing something that NVivo can do more or less automatically. The advantage of doing it in CT is that it allows you more control with organising the outline (which will become the depository of your codes). E.g. you can export it as OPML and import it into a more feature-rich outliner such as Natara Bonsai, where you can analyse and organise it further and then even re-import it back into CT. The CT outline can obviously be used as an outline for writing, and you can e.g. drag it onto a blank CT page (topic) where the outline items immediately become a hierarchy of headings.

I believe Eduardo is considering some further improvements to the Outliner in CT v. 6, which would make the navigation of very large outlines even easier and might even make the extraction of headings as “codes” possible. I haven’t seen the beta yet but I’m hoping these changes will be implemented.

Finally on the CT forum I heard of some plans to implement a new mark-up which would allow for the coding of passages and extracting these passages, which would present all new opportunities for qualitative data analysis.

Posted by Alexander Deliyannis
Sep 18, 2012 at 12:57 PM

Dr Andus wrote:
>Finally on the CT forum I heard of some plans to implement a new
>mark-up which would allow for the coding of passages and extracting these passages,
>which would present all new opportunities for qualitative data analysis.

I know that this is not the right place for CT feature requests (I will eventually subscribe to the CT forum) but my hopes are that eventually CT will also support MarkDown…

Posted by Dr Andus
Sep 18, 2012 at 01:39 PM

[Moved from another thread]

Alexander Deliyannis wrote:
>Dr Andus wrote:
>>Yes, this is a common complaint
>>about working with wikis and CT in
>particular. However, I think some of this is down to
>>perceptions. A string of
>command like that looks like intimidating gibberish (which
>>disappears in view
>mode though). But actually once you get used to the look of it, it
>>might be quicker or
>at least the same as doing this in Word
> >Quite true.
> >In practice, I spend most
>time in “edit” mode, so complex markup gets in the way. But I could probably get used to
>it indeed.

Here is one thing that helps me with ignoring mark-up in the text in CT edit mode. It’s a bit counter-intuitive but if you assign different colours to specific mark-ups, somehow the eye recognises that they are mark-ups and jumps over them, so it becomes easier to read and write in edit mode. Go to Tools > Editor > Colours and select different colours for the various elements. E.g. Headings for me are light blue, links are URL-type blue, Commands are green, Comments are orange, and Includes are red.

Pages: ‹ First < 6 7 8 9 10 11 12 >

Back to topic list