Possible Ways to Blend Qualitative and Quantitative Research Methods

1
Possible Ways to Blend Qualitative and
Quantitative Research Methods

• Ian McDowell
• Seminar Presentation
• 1997

2
Overview of Talk

• Research methods gradually evolving in
  recognition of inadequacies in current methods
• Two paradigms positivist / quantitative vs.
  subjectivist / postmodern
• There are strengths in each
• Types of blend of the two, in
• study design
• data collection
• analysis
• comunicating results

3
Styles of Thought(how do we know that we know
what we think we know?)

• Many dualisms throughout history of thought
• Quantitative vs. qualitative
• Deductive vs. inductive
• Right brain and left
• Yin and Yang
• Apollonian and Dionysiac
• Male and female
• Reductionist vs. systems thinking

4
Changing philosophies of knowledge

• 17th 18th centuries order, logic and science,
  world seen through senses. Mechanical world.
  Realism and logical positivism
• 19th century - social revolution can we analyze
  behaviour logically? Idealism the human mind as
  source of knowledge people, not logic, crucial
  in explaining reality
• 20th century - phenomenology qualitative research

5
Two paradigms

• Challenge biological variability should we
  focus on the general or the specific?
• General public health epidemiology deduction.
  Nomothetic
• Specific clinical medicine psychology
  induction. Idiographic

6
Quantitative approach

• Describes and imposes external structure on data
• Gives parsimonious summary of results
  reductionist (for example, shared variance is
  attributed to one variable hence it is
  reductionistic)
• Seeks to isolate systems from their enviornment
  and to generalize
• Efficient, but incomplete view of
  interconnectedness of reality
• Asks the How? question
• Externally valid

7
Qualitative approach

• Interprets, explains generates concepts
• Seeks to be open, flexible
• The investigator is the instrument art versus
  science
• Sampling becomes a crucial issue (in data
  collection and in analysis)
• Somewhat magical approach to analysis
• Asks the Why? question
• Particularizes internally valid

8
Blending Qualitative and Quantitative

• Metaphor of binocular vision
• Seeks to array strengths of one against
  limitations of the other
• Nature of the balance may depend on stage of
  study qualitative in a process evaluation,
  quantitative in outcome study, for example

9
Five blends

• Hierarchical model one method takes the lead
• qualitative leads
• quantitative leads
• Partnership model equal but contrasting
  contributions
• sequential
• cyclical
• simultaneous application (triangulation)

10
Applying these Types of Blend

• In different stages of research
• Conceptualizing the study
• Collecting data
• Analysing data
• Interpreting the data

11
Stage 1 Conceptualizing the Study

• Hierarchical model, quantitative leading, in
  hard studies
• Hierarchical model, qualitative leading, in
  soft topics
• Partnership model applicable in mixed studies or
  in broad programme of research that involves
  sequence of individual studies
• Sequential partnership in formulating study
  qualitative leads into quantitative

12
Stage 2 Collecting the Data

• Goal of blending approaches is to compensate for
  limitations in each approach
• Hierarchical model illustrated by data
  supplementation (e.g., qualitative interviews
  highlight responses to a standardised
  questionnaire)
• Partnership sequential model illustrated in
  qualitative work to develop questionnaires

13
Stage 3 Data analysis

• Generally hierarchical determined by design of
  study. Orientation of funding agencies often
  makes it hard to achieve a true balance
  (disciplinary racism)
• Hierarchical, quantitative leading illustrated
  by analyses of outliers
• Hierarchical, qualitative leading case studies
  are followed by secondary analysis fo
  quantitative data (e.g. a survey) to estimate
  representativeness of insights gained from the
  case study
• Iterative analyses in partnership model, but this
  will probably be criticized from both camps.

14
4 Interpretation Dissemination

• Hierarchical, quantitative leading
• Use case histories or quotations to illustrate
  quantitative results
• Use qualitative results to comment on exceptions
  to the rule
• Hierarchical, qualitative leading use
  quantitative results to validate what people
  suspected all along

15
Future Directions

• Theres increasing awareness in funding agencies
  of the importance of qualitative research. Its
  a start, but.
• However, the paradigms are sufficiently different
  that its very hard to blend them attempts
  rapidly lead to criticism that you are perverting
  the tenets of each approach
• Disciplinary purity seems remarkably important to
  academics a fundamental part of personal
  identity so conflicts will be common
• A successful blend will be truly
  transdisciplinary