|
Qualitative
analysis is real work. Do not try to analyze badly
gathered data. It takes too much time for too
little knowledge gained. If the data are bad enough,
you may reach erroneous conclusions. Usually a
pre-existing data set gathered for some other
purpose is not useful. Gather your own data using
a suitable sampling framework for your research
question.
A challenge,
a joy, or possibly a nightmare is the proper analysis of qualitative
data. Many ethnographers have spent years trying to make sense
of her or his field notes. In health care we do not have that
luxury or agony. I will introduce four analysis techniques particularly
suited to symptom research.
1)Content
Analysis
The
simplest and most popular qualitative method for
health care research is content
analysis. (Click here for a brief bibliography
with more information about content analysis).
Content analysis is the systematic description
of behavior asking who, what, where, where and
how questions within formulated systematic rules
to limit the effects of analyst bias. It is the
preferred technique for analyzing semi-structured
interviews and cognitive testing interviews. Content
analysis is comfortably self-taught and analyses
progress quickly. Those are big advantages. The
disadvantages are the analyses can be "dirty"
and I am sometimes not convinced by results that
seem marred by analyst bias. For focus group and
narrative data, I am almost never convinced that
it was worth the trouble to gather the data if
content analysis is the only technique used. Content
analysis is a good beginning but often not sufficient
by itself.
2)Grounded
Theory
This is the
classic and still standard technique for analyzing health data
and for lots of other data too. Grounded theory uses a systematic
hierarchical set of procedures to develop inductively derived
theory grounded in data. Glaser
and Strauss invented Grounded Theory in the 1960s to
analyze data on caring for dying patients. Their books are classics
and in wide use today because of the insights they provide, particularly
in the field of palliative care. To visit a website devoted to
Grounded Theory which includes an audio interview with Glaser
click here. Gifford,
1999, and Lewis,
1997, are examples of analyses conducted with
Grounded Theory.
3)Narrative
Summary Analyses
This technique
was invented later by Carol Gilligan, partly as a response to
Grounded Theory and also in response to the movement in literature
and history called Deconstruction. Gilligan reminded us that after
taking our data apart to get at essences, we also could gain valuable
insights by putting the data back together, not in their raw form,
but in re-ordered form to tell stories from the points of view
of different participants. Narrative Summary Analysis Technique
is also called "threading".
4)Triangulation
The standard
meaning of triangulation in health research is the strengthening
of both qualitative and quantitative analyses by combining insights
from both. For an excellent article on triangulation from the
Journal of Advanced Nursing click
here.
|