Beruflich Dokumente
Kultur Dokumente
Research
Mixed Method
Research
Mixed Method
Research
Neither of the two styles of research can fully deliver on its promise to establish the truth
about phenomena of interest to nurse researchers. However, the strengths and weaknesses
of quantitative and qualitative data are complementary.
Combined shrewdly in a single study, qualitative and quantitative data can supply each
others lack. By using multiple methods, researchers can allow each method to do what it
does best, with the possibility of avoiding the limitations of a single approach.
Why Use
Multi
Method?
Incrementality
Qualitative methods are well suited to exploratory or hypothesisgenerating research early in the development of a problem area.
Quantitative methods are needed as the problem area matures for the
purposes of verification.
However, the evolution of a theory or problem area is rarely linear and
unidirectional.
The need for exploration and in-depth insights is rarely confined to the
beginning of an area of research inquiry, and subjective insights may
need to be evaluated early and continually.
Enhanced Validity
For example, Ersek, Ferrell, Dow, and Melancon (1997), in their study
of quality of life in women with ovarian cancer, used qualitative data
to validate their quantitative quality-of-life measures.
APPLICATIONS
OF
MULTIMETHOD
RESEARCH
Instrument Development
Hypothesis Generation
MULTIMETHO
D RESEARCH
DESIGNS
1.
Multimethod Component Designs
Expansion design, in which different methods are used for distinct inquiry
componentsas might be the case in an evaluation that involved both a
process and impact analysis. The results from such studies are often
presented in a side-by-side fashion, rather than woven together into a single
story.
1. Iterative designs involve a dynamic in which the findings from one method are
used as a basis for moving forward with further research using the alternative
method (as is typically the case with instrument development and refinement).
2. Embedded designs (or nested designs), one methodologic approach is
embedded in the other, interlocking contrasting inquiry characteristics in a
framework of creative tension
3. Holistic designs feature the essential interdependence of alternative methods for
gaining a full understanding of complex phenomena. In holistic designs, the methods
are integrated simultaneously rather than hierarchically.
4. Transformative designs, the emphasis is on blending the value commitments of
different research traditions to arrive at a better representation of the multiple
interests in the larger social context.
Sandelowskis scheme makes clear that most multimethod studies involve decisions about
how to order data collection. In some cases (especially in component studies), data
collection for the two approaches occurs more or less simultaneously.
In others, however, there are important advantages to timing the approaches so that the
second phase builds on knowledge gained in the first.
TIP: Many multimethod studies are conducted in two or more phases, such as conducting indepth interviews with a subsample of patients from whom biophysiologic data were
obtained after analysis of those data has been done. If there is a possibility that you might
go back to study participants to obtain more data, be sure to structure your consent form in
such a way that they are aware of any potential future demands on their time. Also, be sure
to obtain contact information to facilitate finding them at a later date.
STRATEGIES
FOR
MULTIMETHO
D RESEARCH
Mass (2000), for example, believes that the quest for meaning and
the quest for measurement are incommensurable Nevertheless,
examples of multimethod research abound. Although it is not
possible to develop a catalog of multimethod strategies,
Clinical Trials
Evaluation Research
Qualitative data collection methods are especially useful when the researcher is
evaluating complex interventions. When a new treatment is straightforward (e.g., a
new drug), it is usually easy to interpret the results: post-treatment group
differences usually can be attributed to the intervention.
However, many nursing interventions are not so straightforward. They may involve
new ways of interacting with patients or new approaches to organizing the delivery
of care. Sometimes, the intervention is multidimensional, involving several distinct
features. At the end of the evaluation, even when hypothesized results are
obtained, people may ask,
What was it that really caused the group differences? (If there were no group
differences, then the important question would be, Why was the intervention
unsuccessful?) In-depth qualitative interviews with subjects could help to address
these questions. In other words, qualitative data may help researchers to address
the black box question understanding what it is about the complex intervention
that drove observed effects.
Surveys
Ethnographies
The methods used in ethnographic field studies usually yield a rich array
of data amenable to qualitative analysis, such as notes from qualitative
observations, indepth interviews, and narrative documents such as diaries
and letters.
Ethnographers can, in some cases, profit from the collection of more
structured information from a larger or more representative sample than is
possible in collecting the qualitative data.
OBSTACLES
TO
MULTIMETHO
D RESEARCH
1. Epistemologic biases
2. Costs.
3. Researcher training.
4. Analytic challenges.
5. Publication biases.
1. References: