At 08:30 17.04.2003 -0400, you wrote:
>This is great. I hadn't thought of using the Network Modeling for directly
>"coding" complex interpretations.
>
>A related question... Is there support in Atlas.ti for managing the codes for
>individuals/instances? Keeping the codes list of named or numbered individuals
>of a specific elementary type could be helped by special display and
>bookkeeping
>tools (e.g., dealing with "plane #1", "plane #2" and "plane #1 isa plane",
>"plane #2 isa plane" etc.). Any suggestions on how best to manage that?
No problem: you would implement such a "typology" with code "families",
which are simple named (and commentable) containers for codes, memos, etc.
So a code family "propositions" would hold codes like "seeing_1" or
"flying_77", a code family instances codes like "John", "Mary", "7 pm". You
can have these families (which can also be seen as dichotomeous variables
assigned to the entities) as nodes in networks too, so they are quite
accessible.
>I agree that the networks can get overly complex, but even simple use can
>reduce
>the number of codes and allow more expressive and precise coding. For example,
>in studying International Trade, it allows the roles "source" and
>"destination"
>of exports to be combined with country codes rather than having multiple
>compound codes like "export from US to UK", "export from UK to US", etc.,
>while
>of course just having "export", "UK", "US" codes wouldn't distinguish
>direction.
Yes, not many of our ATLAS.ti users exploit this capability of assigning
complete statements to incidences in the text. This seems to be one of the
features that one should see in "real time" ...
- Thomas
>Thanks,
>John Hanna
>
>----- Original Message -----
>From: Thomas Muhr <[log in to unmask]>
>To: <[log in to unmask]>
>Sent: Wednesday, April 16, 2003 11:56
>Subject: Re: QDA Software for Coding with Conceptual Graphs
>
>
>Our QDA software - ATLAS.ti - supports network modeling on semantic and
>episodic level via user definable sets of relations. Through semantic
>retrieval (and relation properties like transitivity), you may yield all
>constituents of a proposition. In addition, rhetorical structures can be
>represented on the data level via named hyper links. So it fits a number of
>approaches like action theory, grounded theory or argumentation analysis
>quite well.
>However, there is some drawback with all such representation techniques if
>used beyond browsing or "heuristic" mind mapping. They look really pretty
>for trivial examples but can get quite complex and loose some of their
>expressive power when it comes to modeling real world episodes.
>
>Here is a textual output of a network fragment from ATLAS.ti with both
>semantic and episodic knowledge representing different interpretations of
>the "birds" ("Millers saw the cranes when they were flying over the alps")
>example:
>
>Action_x
> <agens>: Agent_x
> <instr>: Instrument_x
> <loc>: Loc_ationx
............
>____________________________
„Computers, like every technology, are a vehicle for the transformation
of tradition“ (Winograd & Flores, 1987)
Scientific Software Development - Berlin - www.atlasti.de
Dipl.-Psych. Dipl.-Inform. Thomas Muhr - [log in to unmask]
Support: [log in to unmask] Sales: [log in to unmask]
|