So, what could we say instead?
I am tempted by Paul T H's (sorry for abbreviation) use of 'selection' -
i.e. purposeful selection of participants rather that purposeful
sampling....
what do others think, and what would happen if you used this in a research
proposal?
Actually - here's an opportunity for some empirical shtuff, I have to write
a Ffud (phonetic spelling for PhD) proposal. I'm going to ask my
department!
Opinions from interested parties on possible new terminology would be
gratefully received.
Actually it would make quite a nice email collaborative new-fangled 21st
century paper for us all?
Sarah Delaney
Research Officer
Health Services Research Centre
Department of Psychology
Royal College of Surgeons in Ireland
The Mercer Building
Mercer Street Lower
Dublin 2
00-353-1-4022121
[log in to unmask]
> ----------
> From: Paul ten Have
> Reply To: qual-software
> Sent: Friday, November 23, 2001 11:19 am
> To: [log in to unmask]
> Subject: Re: Haphazard sampling
>
> At 10:49 23-11-01 -0000, you (=Sarah Delaney) wrote:
>
> >sometimes I wonder whether even using the term 'sampling' with all it's
> >connotations can be construed as attempting to 'live up' to a quant
> >standard, when really qual should be setting it's own. What does
> 'sampling'
> >mean? How relevant is it to the qual endeavour when it is so associated
> >with the positivistic tradition that you have explain for hours to
> someone
> >why qualitative sampling is different?
> >
> >does this make any sense?
>
> Yes, I think it does. 'Sampling' suggests a population of researchable
> 'objects' with variable 'properties' (i.e. variables) which is too large
> to
> be studied in full, so one studies a sample from that population in order
> to estimates of distributions of and statistical relationships between
> those variables. Now a lot of 'qualitative' research uses this rather
> 'quantitative' language in whole or in part. For instance, in the
> 'grounded
> theory' approach the language of variables is used without any discussion
> of its specific features and limitations. Similarly for 'sampling' in
> other
> qualitative contexts. For many qualitative studies, the 'population' that
> is studied is made up of individuals, for others it is 'organizations' of
> a
> specific type. But does it make sense to talk about a 'population' if the
> object of study is 'a culture', or 'language use' or 'solving software
> problems'?
>
> To my mind, the conventional language of variables, samples etc. does not
> fit all types of qualitative research. It depends on the concepts,
> frameworks, etc. one uses. Ultimately, one's 'theoretical object' should
> be
> the starting point for one's choice in these matters. Taking off from that
> point, various kinds of data selection may be considered, for instance to
> maximize variation within the set, or to limit variation to specific
> contrastive subtypes. Furthermore, the treatment of deviant cases deserves
> a well-reasoned choice of strategy.
>
> Etc. !
>
> My 2 euro cents...
>
> Paul
>
>
>
> Paul ten Have, Associate Professor, Department of Sociology &
> Anthropology,
> Faculty of Social and Behavioural Sciences, University of Amsterdam.
> O.Z. Achterburgwal 185, 1012 DK Amsterdam, the Netherlands
> http://www.pscw.uva.nl/emca/index.htm
>
>
|