Jon,
I'm not sure whether I understand your problem well. If you are only
interested in your group of volunteers, but are unable to convince them to
participate all in your study, you have the problem of the effect of
self-selection on the representativity of what basically is an census, not
a sample. That is in analogy more an issue of non response than of
sampling. The only thing you can do is comparing formal characteristics of
your participants with formal (and known!) characteristics of
non-participants a try to find a plausible argument to generalise to all
your volunteers.
However, if you are not only interested in your own volunteers but in
volunteers in a more general sense, then the reasoning should be a logical
and theoretical rather than a statistical one: see to what extend the
volunteers you studied have characteristcs common with or different to
other types of volunteers and find a plausible reasoning for
generalisations yes or no.
Quite another situation would be a qualitative research on voluntary work
(in general or in a particular field) and finding a relevant purposive
sample to be able to make generalisations. Then, my strategy would be to
develop a theoretical reasoning of either different "types of volunteers"
or different "types of situations volunteers work in" (or a combination of
both) to define probable sources of difference and commonality. Then, as a
rule of thumb, find at least 2 (and preferably more) examples of each type
(ideally: for all permuted combinations; if that would be too much: at
least 2 cases of each specific type). Not 1, since then you never know
whether what you find is characteristic for the type or idiosyncratic for
the case. The drawing of a property space of all typological dimensions
declared relevant will help to see what purposive sample is most efficient,
given time and means. (Allen H. Barton: "The concept of property-space in
social research", in: Paul Lazarsfeld & Morris Rosenberg (eds.), The
language of social research; a reader in the methodology of social
research, The Free Press of Glencoe 1964: 40-53. !! Know your classics!,
still very useful.).
Is this something like an answer to your query?
Emiel Droogleever Fortuijn
At 12:09 22-11-01 +0000, you wrote:
>
>
>
>
> Hello Emiel still crop up in purportedly small scale qua
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>litave
>studies.
>
>
>
>
>
>
>
>
>
>
>
>sernse
>of
>self.
>I
>would
>be
>interested
>to
>know
>how
>you
>nstrue
>a
>sample
>composed
>of
>volunteers
>whe
>all
>are
>included
>in
>the
>study.
>Woulyou
>argue
>that
>some
>further
>theory
>bas
>selection
>should
>take
>place?
>best
>Jon
>berts
>
>
>
>>From:
>Emiel
>Droogleever
>Fortuijn
>
>
>>Reply-To:
>qual-software
>
>
>>To:
>[log in to unmask]
>
>
>>Subject:
>Re:
>Haphazard
>sampling
>
>>Date:
>Thu, 22 Nov 2001
>12:53:39
>+0100
> >
>>I
>even
>don't agree
>with
>being
>sympathetic to
>the
>idea
>that
>qualitative
>
>>strategies
>are
>not
>representative.
>The
>eternal
>difficulty
>is to
>unlearn
>>that
>there
>is
>only one
>type
>of
>represenativity,
>i.e.
>statistical
>>&
>Strauss
>invented
>the
>very
>apt
>>term
>"theoretical
>representativity"
>which
>is
>arrived
>at
>by
>adequate
>
>>"theoretical
>sampling":
>for
>theoretical
>reasons
>your
>are
>purposefully
> >looking for
>cases that
>fit
>a
>thoretical
>idea.
>Sometimes
>it seems to
>be >more
>adequate
>to
>call
>it
>"conceptual
>representativity",
>since
>one is
>
>>looking
>for
>relatively
>pure examples in
>reality of
>a
>concept,
>for
>instance
>>"an
>almost
>ideal-typical bureaucracy"
>to
>do
>in
>in-depth
>research into
>the
>
>>phenomenon of
>bureaucracy
>(or
>whatever).
>Michael
>Patton gives a
>nice
>
>>practical
>elaboration of
>this
>idea in
>his
>Qualitative
>evaluation and
>
>
>>research methods
>(2nd
>edition),
>Sage 1990:
>182-183.
>He calls it -
>like
>
>>others
>-
>"purposive sampling",
>which comes
>close
>to
>Rowland's
>comment.
>Uwe
>
>>Flick gives a nice
>elaborion
>when
>he
>pleads
>for
>purposive
>sampling
>
>>strategies
>in
>ethnographic
>fieldwork
>(selection
>of
>location,
>settings or
>>scenes;
>selection of
>time;
>selection
>of
>activities or
>things
>happening;
>>selection of
>parts
>of
>selected
>research material
>for
>close inspection;
>>selection
>of
>examples
>for
>presentation
>in
>publications)(see his "An
>
>>Introduction
>to
>Qualitative
>Research,
>Sage
>1998, Table 7, p. 63).
>The
>issue
>>of
>sampling
>is
>undertheorised in
>qualitative research.
>>
>
>
>>Emiel
>Droogleever
>Fortuijn > > > >At
>09:37
>22-11-01
>+0000,
>you
>wrote:
>>>Hi all
>>> >>I was wondering
>if
>anyone else
>has
>reservations
>about
>the use of
>the term
>>>haphazard
>sampling
>to
>describe
>sampling
>strategies used
>in
>qualitative
>>>reserach.
>It
>stkes me that
>even if one
>is
>sympathetic
>with the
>idea that
>>>qual
>strategies
>are
>not
>respresentative (which I
>am) the
>last thing
>that
>>>they are
>is
>haphazard
>- is this a form
>of
>epistemological
>superority?
>All
>>>qual
>research has
>to carefully
>plan
>and uncover key
>actors
>and
>contacts to
>>>generate data
>which
>makes the
>term seem
>something
>of a
>hangover from a
>time >>when
>quant
>methods
>were seen as
>having
>all the
>answers. >>
>>>Im intersted
>in
>what
>others think,
>>>
>>>cheers >>
> >>Rowland
>
>>>
>
>>>
>
>
>>==========================================================================
== >Emiel Droogleever Fortuijn >Afdeling Sociogie en
Antropologie/Department of Sociology and Anthropology >Oudezij
Achterburgwal 185 >1012 DK Amsterdam,
T========================================================================
>Get your FREE download of MSN Explorer at http://explorer.msn.com
>
===========================================================================
Emiel Droogleever Fortuijn
Afdeling Sociologie en Antropologie/Department of Sociology and Anthropology
Oudezijds Achterburgwal 185
1012 DK Amsterdam, The Netherlands (NL)
Tel: *31 (0)20 525 2228 (thuis/home: *31 (0)75 617 0379)
Fax: *31 (0)20 525 3010
Email: [log in to unmask]
===========================================================================
|