NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/bit/listserv/qualrsl/2076

Text File | 1993-01-22 | 3.9 KB | 84 lines

Newsgroups: bit.listserv.qualrs-l
Path: sparky!uunet!usc!howland.reston.ans.net!spool.mu.edu!torn!nott!cunews!tgee
From: [e-mail] (Travis Gee)
Subject: Re: Coding in qualitative analysis
Message-ID: <tgee.727717727@cunews>
Sender: [e-mail] (News Administrator)
Organization: Carleton University
References: <[e-mail]>
Date: Fri, 22 Jan 1993 15:48:47 GMT
Lines: 74

In <[e-mail]> Monika Ardelt <[e-mail]> writes:
>Travis:
>yes, I think your interpretation of my explanation is in principle
>correct.

>> "Even though every person in the society is in principle able to
>> conduct the analysis, we need other researchers from the same
>> sociocultural background to to validate our analysis, because we
>> cannot assume that the interviewee has done the analysis of latent
>> structures himself or herself."
>>

That's a relief...this stuff is pretty abstract, and easy to
misinterpret!

>However, not only can we not assume that the interviewee has done the
>analysis him- or herself, in general,it is very difficult to use the
>method of structural hermeneutics to analyze one's own behavior.
>The key is that one starts the analysis by trying to find as many
>different meanings for a particular statement as possible.

The point of my previous postings, on reflection, was to corrode the
privileged position of the researcher with respect to what is
happening in someone else's mind.  This is a key presumption of any
approach that purports to be "objective,"  and has traditionally
resulted in the distinct possibility that the researcher's pet
hypotheses find their way into the interpretation of the data.
Grounding the data with respect to the people who generate it is an
approach which I feel lessens the static discharge that commonly
accompanies the publication of a qualitative study.

  My previous posts were not meant to say that interjudge reliability
should be banished, because we *do* want other researchers to agree
with us. The problem lies in the fact that we are looking for
intersubjective agreement. We want other researchers to come to the
same conclusions as us. But what is our standard for validity? You
observe that

>the interviewee, more often than not, is very likely to have already
>some idea what his/her particular statement meant and, hence, is not
>open enough to alternative interpretations.

   We each draw on our unique experiences to derive all possible
meanings *for us*. Researchers share many experiences, so the domain
of meanings which are generated will by and large be shared a-priori.
This means that it will be *constrained*.  People who do not share the
training of a researcher are *more* likely to arrive at different
meanings than people who do share that training. That is why in my
study, I took my conclusions not only to some of the people that I
interviewed, but some people who had *not* been interviewed. It
controlled for the possibility of both Type I and Type II errors.
  Both of these types of error are enhanced when we do not refer back
to the source of the data because a) our unique experiences as
researchers can result in latent "meanings" that aren't really there,
and b) the unique experiences of the interviewees cannot be brought to
bear on our conclusions.

>  I agree that an example
>would probably help to clarify matters.  And even though I was a
>little bit reluctant to give one since the method is very elaborate,
>I'm currently working on one.  It will just take me some time to
>complete it.

I look forward to the post!

((((((((((((((((((((((((((((((((((((((((((((((((((((((((((
Travis Gee ()  [e-mail]     ()
           ()  [e-mail]  ()        ()()()()
           ()                           ()              ()
           ()                           ()()()()()()()()()
Recent government figures indicate that 43% of all statistics
are utterly useless.