Welcome, Guest. Please Login
Tinderbox
  News:
IMPORTANT MESSAGE! This forum has now been replaced by a new forum at http://forum.eastgate.com and no further posting or member registration is allowed. The forum is still accessible via read-only access for reference purposes. If you wish to discuss content here, please use the new forum. N.B. - posting in the new forum requires a fresh registration in the new forum (sorry - member data can't be ported).
  HomeHelpSearchLogin  
 
Pages: 1 2 3 
Send Topic Print
Setting up TB 6 for qualitative research project (Read 8273 times)
NZT-48
Full Member
*
Offline



Posts: 24

Setting up TB 6 for qualitative research project
Nov 10th, 2015, 9:16am
 
Hi,

I'm working on a qualitative research project (i.e., a investigative reporting project), and was considering using the model for doing this in the web video, "Processing Qualitative Data with Tinderbox."

Quick background...  I use DEVONThink Pro to store and organize my documents, and Scrivner for writing.  I have a number of text files for which I need to segregate and organize names, dates, associations (e.g., overlapping government agencies) etc.  

The web video, "Processing Qualitative Data with Tinderbox" seems to provide a means by which I can achieve this efficiently, while even automating some of the segregation and organization of this data, e.g., through Tinderbox's use of agents.  Is that a fair assessment?  If so, does the "Processing Qualitative Data with Tinderbox" provide a good model?

I've tried to follow the instructions in that video to explode text and then process the data, but I haven't been able to figure out how to set up Attributes first (per the instructions in the web video, which seems to use an earlier version of TB).

How can I set up Attributes for this process in TB 6?  Which Attributes would one suggest I set up, given the categories I've described:  names, dates, associations (e.g., overlapping government agencies)?  I assume that the next step is setting up alias from there, correct?

I've got more questions, but I feel I need to first understand the basic parts of this before I get in too deep.  Thanks so much for your help...
Back to top
 
« Last Edit: Nov 10th, 2015, 10:11am by NZT-48 »  
  IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #1 - Nov 10th, 2015, 11:08am
 
To set up new user attributes, use the 'User' tab of the Document Inspector (the linked page explains the interface).

The number and data type of attributes you want will depend on how many things you have to track - which is hard to gauge without seeing your data. Any facet of the data you want to refer to in your encoding process, and for which there isn't an an existing system attribute, you'll want to add an attribute.

Explode in Tinderbox v6 is much more powerful than previously. See more on Explode.

Most likely you'll want to set all/some of your user attributes as Key Attributes in your exploded data notes. That's most easily set up via use of prototypes. Note that the Explode function (see link above) can help set up prototypes for exploded notes assuming you've configured them before you import.

Tom Webster's video explains the process pretty thoroughly and likely sounds a reasonable fit for your task. Don't forget that just because he shows the data triage happening in a Map, you can still use other views to look at the sorted work - especially once you've done your sorting.

If that all seems a lot to do, break it into small iterative experiments:
  • adding new user attributes.
  • setting up key attributes.
  • make and apply prototypes.
  • explode imported text and apply a prototype to the notes created.
  • experiment with Tom W's map-based sorting method.
Then, you can roll that together into a file suitable customised for your process.  If you're going to to this several times, consider making a fully configured blank document and using a copy for each separate set of work.

Lastly, if you get stuck on some of the detail - just ask.
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
NZT-48
Full Member
*
Offline



Posts: 24

Re: Setting up TB 6 for qualitative research project
Reply #2 - Nov 12th, 2015, 2:31pm
 

Thanks so much for this robust post, and answers to my questions.  I greatly appreciate all of the help.  Truly.  I have more questions, but first...  I'd like to more clearly explain what I'm seeking to do before I put this in motion, and ask follow up questions...

As I said, I want to process text containing qualitative data.  By "processing" I'm wondering if TB can do the following (either through the steps outlined in "Processing Qualitative Data with Tinderbox" video or via another set up):

-segregate and organize: names (e.g., names of individuals, names of units and agencies, etc.), dates, and associations (e.g., where individuals and their respective units overlap), etc.  

-Set up this process in such a way so that, once names and dates are ID'd, TB can automatically "tag" them (or create aliases for them) in appropriate containers and adornments (I assume through a combination of attributes and agents, correct?).

-Once set up, I'd be able to see the segregation of this data (in containers) and links between associations, dates, etc. as shown in TB's Map view.

If possible, I'd also like to do the follow...

-Import this data into Scrivner so that I can access the TB's processed notes/data -- and therein the text that's lined to the segregated and organized data.  It seems that the beauty of TB is it's ability to help one process, then visualize and map out notes for myriad projects. But as researcher and a writer, the final output is in my writing.  So, it seems like porting over processed notes/data from TB to Scrivner, so that I can access (i.e., copy, paste, cite and link to) the notes/data is a key part of this process.  

(By the way, I use Devonthink Pro to create tagged notes from annotated documents, and also port these notes over into Scriner where I use them as reference material for the same purpose.  I've often wondering if there's a way to take these Devonthink tagged notes and process them within TB, and then export/import them into Scrivner, but I'm not sure if that's advisable -- or even viable.  Maybe others have thoughts on that as well -- happy to discuss it if others are interested.)

One concern / quandary about this (and my other, related work processes -(e.g., exporting Devonthink Pro tagged notes into Scrivner )...  

I've read that TB and Scrivner work together harmoniously, and that's great.  But my research grows as I process more data (e.g., annotate & tag more documents), and so I'm concerned about how best to ensure that I'm properly updated everything -- that is, processing all the data, and ensure that everything is getting ported over into Scrivner for the final output.  

Right now, it seems that the marriage between TB and Scrivner works through a process of exporting from TB and importing into Scrivner.  I suppose I wish there were some kind of cloud based solution for such data processing and saving, in order to ensure that everything could be properly synced between these apps.  Does this make sense?

Anyway, I'm curious to hear thoughts about what I've written here.  I want to primarily ensure that what I'm doing is feasible, and advisable (or not), before I ask more technical questions about the kind of TB set up that seems best for my purposes.

Thanks very much for your assistance -- I greatly appreciate it.


Back to top
 
 
  IP Logged
Mark Bernstein
YaBB Administrator
*
Offline

designer of
Tinderbox

Posts: 2871
Eastgate Systems, Inc.
Re: Setting up TB 6 for qualitative research project
Reply #3 - Nov 12th, 2015, 3:59pm
 
Nope -- no cloud sync as yet.  That would be very costly to implement, and for projects like yours I don't think it's absolutely essential.

Do you anticipate a need to return from Scrivener to Tinderbox often?  If the flow is (more or less) unidirectional, repeated export should be effective, perhaps using styles to automate formatting changes.
Back to top
 
 
WWW   IP Logged
NZT-48
Full Member
*
Offline



Posts: 24

Re: Setting up TB 6 for qualitative research project
Reply #4 - Nov 12th, 2015, 4:04pm
 
I'm guessing it would be uni-directional -- at least for the stage I'm in right now. For now I'm thinking the flow would look like:

DATA-> Processed via Devonthink (tags) or Tinderbox -> written up in Scrivner

Like I said, there might be a future edition that could include DATA-> Processed via Devonthink (tags) --> Tinderbox -> written up in Scrivner

I defer to you about whether that's advisable and viable...

Thank you!  I look forward to responses about the other parts of my post...
Back to top
 
 
  IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #5 - Nov 12th, 2015, 4:51pm
 
Quote:
-segregate and organize: names (e.g., names of individuals, names of units and agencies, etc.), dates, and associations (e.g., where individuals and their respective units overlap), etc.  

In what context? If you've lots of tabular data, you'd do better to export it from source as tabular data (extracted from or in addition to the basic text). That way you can import it into Tinderbox and have Tinderbox map the data to attributes for you. That said, the idea this is based on is using sentence or paragraph-length notes. how much per-sentence data do you expect to have. If the data is all mixed into text, it will boil down to your skill with regular expressions as used in String.contains() queries.

Quote:
-Set up this process in such a way so that, once names and dates are ID'd, TB can automatically "tag" them (or create aliases for them) in appropriate containers and adornments (I assume through a combination of attributes and agents, correct?).

Yes, this is as explained in the original video.

Quote:
-Once set up, I'd be able to see the segregation of this data (in containers) and links between associations, dates, etc. as shown in TB's Map view.

Yes. Though if you use agents you can't use linkTo/linkFrom to link with the original note, so you'll need to plan your workflow carefully. A more detailed description relies on better knowledge of the actual data and would e too long an answer to fit here.

Quote:
-Import this data into Scrivner so that I can access the TB's processed notes/data -- and therein the text that's lined to the segregated and organized data.

I think 'Yes', but again we're talking at such a conceptual level it's not clear quite what you're expecting to happen. Tinderbox can open a Scrivener file and use (most of) the data. However, Scrivener can only import OPML - which Tinderbox can create. So the constraint on what you can shift from Tinderbox to Scrivener is limited by the OPML format rules no by Tinderbox. Can someone who actually does Tinderbox -> Scrivener data transfers chip in here as to how well this process works?
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #6 - Nov 12th, 2015, 4:58pm
 
I meant to add that it might help if you linked to an example* of the sort of data you're going to import (ideally at least 2 records) and then describe where/how the record split is found and what (attribute) data, such as names or dates, you hope to extract from the record.

* by all means alter any sensitive real data. The point is to lets us see how structured the data is. When talking conceptually, it's easy to miss one another's unspoken assumptions.
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
B.Moreira
Full Member
*
Offline



Posts: 71

Re: Setting up TB 6 for qualitative research project
Reply #7 - Nov 16th, 2015, 7:34am
 
Hi. I didn't knew T. Webster video, and it was very informative. I liked the way the drag and drop flow is smooth, but I cant seem to make it work in TBX v6.
Having 2 TBX windows opened, one in outline and other in map view, I cant drag the item from the outline window to the map window - the latter does not "gain focus" and nothing happens.
Is there a way to make this drag and drop functionality work between TBX windows?
Back to top
 
 
  IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #8 - Nov 16th, 2015, 7:55am
 
v5 and v 6 have different UIs (changed for good reasons I won't elaborate here). In v6 you can't drag notes between windows. Instead, you need all the notes you want to work with (or aliases to them) in the same map. Thus you may need to adapt the Webster v5 version slightly for the new UI.
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
B.Moreira
Full Member
*
Offline



Posts: 71

Re: Setting up TB 6 for qualitative research project
Reply #9 - Nov 16th, 2015, 8:05am
 
Thanks for the reply. Yes, thought so as well... In many ways, v6 was great UI revolution. But I'm sad it lost this capability - would make working between views simpler... I guess, with windows / tabs changes in v6, it must be a bit complicated to get this back :/
Back to top
 
 
  IP Logged
NZT-48
Full Member
*
Offline



Posts: 24

Re: Setting up TB 6 for qualitative research project
Reply #10 - Nov 28th, 2015, 5:01pm
 
Sorry for the very late reply - work beckoned, and I've been consumed. Pleas see answers below...

Quote:
Quote:
Quote:
-segregate and organize: names (e.g., names of individuals, names of units and agencies, etc.), dates, and associations (e.g., where individuals and their respective units overlap), etc.  


In what context? If you've lots of tabular data, you'd do better to export it from source as tabular data (extracted from or in addition to the basic text). That way you can import it into Tinderbox and have Tinderbox map the data to attributes for you. That said, the idea this is based on is using sentence or paragraph-length notes. how much per-sentence data do you expect to have. If the data is all mixed into text, it will boil down to your skill with regular expressions as used in String.contains() queries.


Actually, it's not. It's a text file, filled with names, dates, unit designators, etc.

Quote:
Quote:
Quote:
-Set up this process in such a way so that, once names and dates are ID'd, TB can automatically "tag" them (or create aliases for them) in appropriate containers and adornments (I assume through a combination of attributes and agents, correct?).


Yes, this is as explained in the original video.


Ok. Just making sure - in part because the version that was featured in that video is different than today's version (e.g., it seems the drag and drop features aren't available any more, correct?).

Quote:
Quote:
Quote:
-Once set up, I'd be able to see the segregation of this data (in containers) and links between associations, dates, etc. as shown in TB's Map view.


Yes. Though if you use agents you can't use linkTo/linkFrom to link with the original note, so you'll need to plan your workflow carefully. A more detailed description relies on better knowledge of the actual data and would e too long an answer to fit here.


Understood.

Quote:
Quote:
Quote:
-Import this data into Scrivner so that I can access the TB's processed notes/data -- and therein the text that's lined to the segregated and organized data.


I think 'Yes', but again we're talking at such a conceptual level it's not clear quite what you're expecting to happen. Tinderbox can open a Scrivener file and use (most of) the data. However, Scrivener can only import OPML - which Tinderbox can create. So the constraint on what you can shift from Tinderbox to Scrivener is limited by the OPML format rules no by Tinderbox. Can someone who actually does Tinderbox -> Scrivener data transfers chip in here as to how well this process works?


I understand your point that "we're talking at such a conceptual level" here about the dynamic workflow between Tinderbox and Scrivner.  Just to be clear, my thought was to process the data in Tinderbox, and then export/import it into Scrivner -- where I'd take notes and an outline structure, and process it into a written document.  It would certainly be helpful to get a  clearer understanding of how this would work, and would it would actually look like (i.e., what transformed Tinderbox output in Scrivner would consist of, and how that would work).  That would be really helpful.  
Back to top
 
 
  IP Logged
NZT-48
Full Member
*
Offline



Posts: 24

Re: Setting up TB 6 for qualitative research project
Reply #11 - Nov 28th, 2015, 5:03pm
 
Quote:
Nope -- no cloud sync as yet.  That would be very costly to implement, and for projects like yours I don't think it's absolutely essential.


Ok, thanks.  Not absolutely essential, but desirable - esp. since it seems it would make it less necessary to regular update data, correct?

Quote:
Do you anticipate a need to return from Scrivener to Tinderbox often?  If the flow is (more or less) unidirectional, repeated export should be effective, perhaps using styles to automate formatting changes.


I think more unidirectional, as you say - from Tinderbox to Scrivener, as you say.

Thanks!
Back to top
 
 
  IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #12 - Nov 29th, 2015, 12:46pm
 
Tinderbox does support drag-drop import of data.  It doesn't support - as at v6.4.0 - drag-drop between different windows.  so part of Webster's method as described isn't possible as I can't drag items between an outline and a map. In v6 that means dragging between windows.

If your data essentially has embedded  attribute-type data and you don't control the source, consider splitting the input via explode and then using agent actions, or rules, to mine this effectively tabular data to populate user attributes.  This doesn't have anything directly to do with Webster's processing method and might be better explored in a separate thread. Here, an example of the sort of data you have would help.

Scrivener accepts OPML. This is quite limited so you don't get rich text for instance. How things look in Scrivener might more usefully be followed up in the (helpful) Scrivener forums. Essentially, you export all or part of your Tinderbox outline to OPML and that your recreated when you import the OPML data into Scrivener. OPML is however not a format that easily supports transfer of lots of custom fields/attributes so you may be limited as to what you can export other than title and text.
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
NZT-48
Full Member
*
Offline



Posts: 24

Re: Setting up TB 6 for qualitative research project
Reply #13 - Feb 19th, 2016, 5:19pm
 
[quote="Mark Anderson"] Quote:
I meant to add that it might help if you linked to an example* of the sort of data you're going to import (ideally at least 2 records) and then describe where/how the record split is found and what (attribute) data, such as names or dates, you hope to extract from the record


Hi, I just realized I failed to do this and would like to give this a try.  I've just scrubbed a sample text document or two.  What's the best way to upload an example?

Thanks!
Back to top
 
 
  IP Logged
Mark Anderson
YaBB Administrator
*
Offline

User - not staff!

Posts: 5689
Southsea, UK
Re: Setting up TB 6 for qualitative research project
Reply #14 - Feb 19th, 2016, 5:40pm
 
The forum does not support direct uploads. Upload a file to your own webspace and link to that or give a dropbox link or such.
Back to top
 
 

--
Mark Anderson
TB user and Wiki Gardener
aTbRef v6
(TB consulting - email me)
WWW shoantel   IP Logged
Pages: 1 2 3 
Send Topic Print