These three questions are interesting, and most qualitative researchers  have faced at least one of these issues at some stage. The second one - about  avoiding coding traps has been extensively canvassed in  earlier QUALSOFT discussions.  Several papers came out of this  discussion.  I wrote one, which has just been published on the  web in the  current issue of  journal of the association for qualitative research.  (address for AQR is http://www.latrobe.edu.au/aqr/   Note that current issues are  accessible to  AIR members only. I think an earlier version of the paper might be on the CAQDAS site).    

On the second one, I think there is an emerging view (not consensus, but strong interest ) in the desirability of 'theoretical' sampling.  This means  thinking carefully about your research question and going for  participants/respondents whose  accounts will  best illuminate the  question.  You might decide that your understanding of an issue will be best helped by  comparing  men and women, or older and younger.  So you would  aim to get a sample (size depending on the type of question and on practical issues of time and money) structured in particular ways.

Or you might feel that a  general phenomenon  in which you are interested (like  altruism) will be illuminated by people who have a particular characteristic (e.g. have been live donors of  a body organ )  in particular circumstances (IE have responded to an appeal from a stranger).  


I think the 'theoretical' part is more important in most qualitative projects than considerations like the  source of the sample.    'Bias' might arise  with issues like how many participants have a relationship with the researcher, but  the kind of randomness  desired in quantitative studies is often not an issue.

I'm particularly interested in the  issue underlying the third question, about how best to use  illustrative  quotes- a recent  Australian text with some interesting material on writing up qual research is  Ezzy, D 2002 Qualitative analysis Practice and Innovation Allen and Unwin: Crows Nest .  (The use of an Oz title is deliberate; we have a lively qualitative research scene here and I  want the rest of the world to know more about us!)

My personal preference as a reader  is this:

a) Give me a short juicy quote that tells me what you see as the essence of the category/code/theme. I like to feel I  can make up my own mind about the data here - agreeing or disagreeing that the  quote represents the theme.  
Then give me a sense of its importance.   You  may do this with some  numbers.   I prefer  simple  ratios or  percentages, qualified perhaps by a comment on whether you are describing the frequency of utterance or the  number.   It may be  however that  although the theme doesn't  doesn't occur  very often you think it is a really important.  Tell me why you think this.  I like to be directed a bit here: I probably don't want illustrations.  

Next, I like to know  about the complexity - what  variations occurred?  what do you make of them?  This is where the quotes can get longer , and where I would expect to see some analytical comments - notes on use of particular words, comparisons between participants, links between themes etc. It's often here that little comments about how the analysis was done are useful.  (something like ' "One of the memos I wrote  at this stage included the note ......" maybe with  a reference to the methodology  chapter or the appendix).  Here you move between illustrating (giving me the sense that I check your comments against data) and directing ( telling me in what way I should read the data if I am to mimic your reading).  

When  someone reports in this  organized way, showing me what they think, why they think it and something about the process that has led to the conclusion, I respond with a sense that this is a qualitative researcher I can trust.  

where there is nothing but  conclusions and  the illustrative quotes (even if there are pages of them)   I end up asking  rhetorically "has this person psychically channeled their conclusions?"  I pinched the complaint from my  historian colleague John Murphy, who made it about the  quarterly reports of an Australian  social science commentator.  The commentator  makes his living  by  quarterly reports on issues of community concern.  The reports are incisive, fascinating etc, but because they don't tell me enough about how the conclusions were reached I end up feeling dissatisfied. On the other hand, the commentator goes on  making a living, so I might be offering  advice that is relevant  only to the most academic  qualitative research.  Differences in standards for writing up data  between say,  and evaluation projects done by  a centre for profit and a PHD project would be interesting to explore.
Helen Marshall 
 

Dr Helen Marshall
Senior Lecturer, Coordinator  of context Curriculum and Acting Honours  Coordinator 
School of Social Science and Planning
Royal Melbourne Institute of Technology

phone 61+03+99253016
fax      61+03+99251087
email [log in to unmask]


Double click on www.rmit.edu.au/tce/ssp 
to visit the School of Social Science and Planning's Website

Or click http://www.rmit.edu.au/tce/ssp/socialscience/honours
to visit the honours program page

Or click www.rmit.edu.au/tce/ssp/ap/context
to visit the  CCpage