>I've collected about 35 hours of conversation (2 participants
>each), and am tracking interruptions, length of speaking turn,
>etc., in addition to coding for "content." Not surprisingly,
>transcription has been agonizing, so I am very interested
>anything that might help me skip this step!
>
>But I wonder if coding sound would be easier than
>transcribing/coding text. For example, I'd guess you can't
>search for certain words & code them
>automatically in a sound file as (I gather) you can do with a
>transcription--is this right? And how much memory would 35 hours
>of audio require?
>
Just a quick response - I am sure others will have more substantive
experience with sound files -
but yes I woudl agree - its a big decision to abandon the idea of 'text'
totally
- and there really is no option if you wish to use some autocoding
mechanisms which make use of 'content' oriented searches or word archives.
Paradoxically - while its very strong on audio and video analysis - there's
an awful lot of functions in Code-a-text which enable such content analysis,
but only once its all in text format - (or Rich text) - such as the usage
level graph - and much more stats stuff which is generated in-house - so to
speak.
But if dealing with just sound files in Code-a-text - then the best bet is
to summarise important sections of the file on screen -(as you listen to the
file) -
from then on those summaries - or abstracts - or "interpretations" as I
htink they are called in C-A-T - remain in synchrony with the sound (or with
the sound AND the text if you HAVE actually transcribed) - so when you
scroll thro the sound - the InterpreTations appear on screen - or vice versa
you can search Interpretations as well I believe.
...using this approach - which could be rather nice - you avoid total
transcription -
you keep the original context - but you also have the chance to jot down
keywords which occur to you, into the intepretations- - you can edit the
interpretations -
Alan Cartwright (developer) needs to remind me whether you can then do a
'content' (word or string) search on the Interpretations themselves - not at
all sure about that - maybe you have to generate text files from the
Interpretations.
But access to your sound files - or parts of them - is always going to be
less fluid and instant - than if leafing thro' a pile of paper ( albeit
painfully produced) - but you knew that already!
cheers
Ann Lewins
Ann Lewins
Resource Officer, CAQDAS Networking Project
Dept of Sociology
University of Surrey
GUILDFORD GU2 5XH
email: [log in to unmask]
CAQDAS web site: http://www.soc.surrey.ac.uk/caqdas/
Tel +44 (0)1 483 259 455
Fax +44 (0)1 483 259 551
-----Original Message-----
From: Dr. B. Young <[log in to unmask]>
To: [log in to unmask] <[log in to unmask]>
Date: 03 June 1999 16:35
Subject: code-a-text question
>I've been browsing the qual-software archives and have a question
>based on something posted to the list about code-a-text a few
>months ago (February) on the subject of voice recognition
>software.
>
>Alan Cartwright wrote: "Code-A-Text offers another solution. You
>transfer the audio file to a digitised format (the facilities are
>part of the programme) and then code against the sound file
>itself. You thus always have access to the original recording
>(you can do the same with video). You can then either write
>precis of each segment or just transcribe those sections which
>are important."
>
>I've collected about 35 hours of conversation (2 participants
>each), and am tracking interruptions, length of speaking turn,
>etc., in addition to coding for "content." Not surprisingly,
>transcription has been agonizing, so I am very interested
>anything that might help me skip this step!
>
>But I wonder if coding sound would be easier than
>transcribing/coding text. For example, I'd guess you can't
>search for certain words & code them
>automatically in a sound file as (I gather) you can do with a
>transcription--is this right? And how much memory would 35 hours
>of audio require?
>
>I'd be grateful if any of you could share your
>thoughts/experiences on this subject.
>
>(As you may have guessed, I haven't purchased a CAQDAS package
>yet. Still trying to learn enough about them to make an
>intelligent choice.)
>
>Thanks,
>Beth Young
>
>--
>Dr. Beth Rapp Young Building: LS-616, +1347
>Director, University Writing Center Voice: 407-823-2853
>Assistant Professor, English Fax: 407-823-3007
>U of Central Florida, Orlando, 32816-1347
>Email: [log in to unmask]
>Homepage: http://pegasus.cc.ucf.edu/~byoung
>
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|