I have recently started to think of how to overcome the structural distortions (e.g. the influence of societal, cultural or organisational influences such as institutionalisation) that research participations may have. I have not yet found a fully satisfactory way around this. Any advice for related research on this is welcome.
When conducting research from a subjective epistemology data including the results of interviews, questionnaires etc does not need to be simply accepted as true. To do so would also surely be to ignore the potential structural factors that may distort data collected. Given this how would you recommend analysing data in such a way that recognises and minimises the impact of these distortions ?
I am interested to see what other 'seasoned' researchers might say here.
As for me, since your research seems to be in social science:
1. Those "structural distortions" are insurmountable because there is no such a thing as a 'distilled' or 'purified' participant free of societal, cultural or organisational influences;
2. In fact, only those data that are affected by societal, cultural or organisational influences are probably the only worthwhile and interesting ones;
3. Perhaps it is even wrong to call them "distortions".
4. If you have a focus, organize interviews with minimal distancing from the main focus. That is perhaps the only thing you can do to get rid of "out of scope" information.
An extra suggestion: Perhaps it would be better to avoid falling into this enchantment of the word 'epistemology'--most of people use it as a metaphor. It usually bears little solution; make things more fancy/sophisticated but with little substance
You might wish to glance at this methodology and the associated theory chapter
http://opus.bath.ac.uk/31921/1/UnivBath_PhD_2012_J_Pugh.pdf
There is a whole field of lay epidemiology that would not see them as distortions - from this perspective it depends on your standpoint.
Here is a lovely and revealing example of what was said to be a Chernobyl cloud and its fall out over Cumbria and the different knowldeges that surround it
http://www.geos.ed.ac.uk/~sallen/jayne/Wynne%201996%20sheep.pdf
and
http://www.dourish.com/classes/readings/Wynne-Misunderstood-PUS.pdf
De acuerdo, en general con las intervenciones. Agregaría que la epistemología, con respecto a las ciencias sociales, cada vez tiene un mayor reconocimiento en el campo de las "ciencias duras", como las matemáticas o la física. Aún en esos terrenos se tiene en cuenta cada vez más la subjetividad del investigador, no como limitante, sino como condición de la investigación. Del lado de la pregunta, la fantasía, la imaginación, y la mentira, forman parte de la realidad en estudio. No son elementos de juicio del investigador, sino son datos que como tales requieren un tratamiento. Tampoco es fácil determinar qué discursos son falsos o verdaderos, y actualmente no se trata de este carácter de los enunciados, sino de los efectos sociales, culturales, y subjetivos que provocan.
Re: "Given this how would you recommend analysing data in such a way that recognises and minimises the impact of these distortions ?"
By far the best place to minimize things like this is back before you *collect* the data, not during data analysis after the fact.
You can torture data to the point where it says anything and nothing. This appears to be more a rule than an exception these days.
I was about to get bogged down in details here, but you are better off looking for more expert advice when it comes to specifics of design.
Rather than guessing about it, I would be inclined to pilot test question sets with a subset of the target population. If you can't get clear results with a small set you are not likely to get very good data from the larger set.
Anyway, the point here is that correcting bad, noisy data is better done at the time you collect the data, not at the time you analyze it.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology