Hi Anne,
>I would still tend to say its a software package which is better at handling
>not too large datasets, and better with an approach which does not require
>too much intensive coding -(if you are going to rely on the code mapping
>facility a lot) - re this, John Seidel woudl probably agree with me (in fact
>he probably said it first).
Sorry but I have to disagree with you here. v5 codes very easily. There
are several alternatives but the basic is to simply block a segment of text
and attach a code word to it. So coding is quite simple and easy.
Further, while a single data file is limitted to 9999 lines (a one hour
interview is about 2000 to 3000 lines so this should be more than
adequate), the number of data files is unlimitted.
In regard to code mapping, I'm not sure what you mean. If it is that your
coding is displayed on the screen embedded in the text, so that you can see
the work that you have already done, then I don't think this works against
intensive coding. Yes, If you have a lot of short segments, and a lot of
code words per segment, things can get a little crowded. But when you do a
search, you can see the output segments with or without the coding.
Further you can expand the context of the search output with or without the
embedded coding.
>
>I have feeling he has deliberately resisted 'autocoding' - (e.g. after a
>text search) - since his own approach has always been dare I say it more
>careful; and his coding processes are always generated by what he reads and
>interprets from the data - rather than any 'shortcutting' (well in terms of
>the initial coding process anyway) - do correct me if I am wrong John.
>
Auto coding is something that I have not favored, but eventually will find
its way into The Ethnograph. My position is that if you search for a word
in the text such as "pain", the results will give you both "He has pain in
his back" and "He is a pain in the ass". Further, there are many times
when "pain" will be talked about without using the word "pain" as in "It
was just unbearable". Finally, if you autocode, you also have to select
arbitrary numbers of lines before and after the occurrence of a word, or
arbitrary units such as a paragraph. Both of these present serious
problems. Is 10 lines before and after enough or too much? If one selects
paragraphs there are two problems: First, have long speeches been broken
down into paragraphs? Second, if the speech has been broken down into
paragraphs, what if you really need the preceeding or following paragraph
to interpret the talk about pain?
While you are coding a data file in The Ethnograph, you can to a search for
words in the text. Thus you could do a search for "pain". You are brought
to the first occurrence and then you make a coding decision: Is this
occurrence conceptually meaningful, and how much context is needed to
convey the conceptual meaningfullness of this occurrence.
Given all this, I understand that there are circumstances when
"auto-coding" can be useful. But I don't think you can ever escape the
problems inherent in auto-coding.
John
[log in to unmask]
P.O. Box 3356
Salt Lake City, UT 84110
Ph: 801-532-3090
Fax:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|