mixed methods workshop unc -- the odum institute february 12-13, 2015 lisa pearce ([email protected])
TRANSCRIPT
Mixed Methods WorkshopUNC -- The Odum Institute
February 12-13, 2015
Lisa Pearce ([email protected])
Chitwan Valley Family Study (CVFS), 1995-2015perl.psc.isr.umich.edu
• Survey methods• Hybrid calendar methods• Environmental measures• Observation• Semi-structured interview methods
National Study of Youth and Religion (NSYR), 2002-2013youthandreligion.org
WAVE 1 (2002/3: Ages 13-17)
• RDD Telephone Survey, National Sample (n=3290)
• In-Person, Semi-Structured Interviews
WAVE 2 (2005: Ages 16-21)
WAVE 3 (2007/8: Ages 18-24)
WAVE 4 (2012/3: Ages 23-30)
• Follow-up Telephone Survey
• Follow-up In-Person, Semi-Structured Interviews
Workshop Plan
• Defining Mixed Methods Research• Standard Definitions• Paradigms• Mixed Research Approach
• MMR & The Research Process• Research Questions• Research Design & Data Collection• Data Analysis & Interpretation• Presentation of Results• Publishing• Research Funding
• Being a Mixed Methods Researcher
Standard Definitions (Johnson
et al. 2007)
• Burke Johnson and Anthony Onwuegbuzie:• Mixed methods research is the class of research where the researcher
mixes or combines quantitative and qualitative research techniques, methods, approaches, concepts or language into a single study or set of related studies.
• Jennifer Greene:• Mixed method inquiry is an approach to investigating the social world
that ideally involves more than one methodological tradition and thus more than one way of knowing, along with more than one kind of technique for gathering, analyzing, and representing human phenomena, all for the purpose of better understanding.
(Johnson et al. 2007)
Paradigms
“Qualitative”Research
“Quantitative”Research
OntologyWhat can we know?
EpistemologyHow do we know it?
AxiologyThe role of values
MethodologyTechnical approaches
The Use of Methods Through Time (Pearce 2012)
1900s 1930s 1950s 1960s -70s-80s 2000s
ArchivalLife HistoryObservation
Sample Surveys/Quantitative/Positivist
Qualitative/Interpretive/
Constructivist
Pragmatism/Mixed
Methods
Thomas & ZnanieckiE. Franklin Frazier
Middletown Studies
Union Democracy Boys in White
Triangulation
TABLE 2 A Pragmatic Alternative to the Key Issues
in Social Science Research Methodology
Qualitative Approach
Quantitative Approach
Pragmatic Approach
Connection of Theory and Data
Relationship to Research Process
Inference from Data
Induction
Subjectivity
Context
Deduction
Objectivity
Generality
Abduction
Intersubjectivity
Transferability
Morgan (2007)
Mixed Research
Philosophical PositionsEpistemology•Subjectivity•Objectivity• Intersubjectivity
Ontology•Relativism•Absolutism•Critical Realism
Axiology•Value-Bound•Value-Free•Value-Conscious
Practical Positions
Logic•Research Problem•Research Question/s•Project Design
Locations on Axes of Inquiry•Subjective to Objective• Inductive to Deductive• Idiographic to Nomothetic•High to Low Reflexivity
Methods•Data Collection•Analysis• Interpretation
Figure 1. Facets of Social Inquiry (Pearce Forthcoming)
• Formulate research problem/objective• Develop purpose/question/hypothesis• Select a research design/method• Collect data• Analyze data• Interpret/validate data• Communicate findings
Research Process (Onwuegbuzie & Leech 2005)
Mixed Methods Research
• …is that which integrates the collection and/or analysis of two or more types of data to answer a set of interrelated research questions
Principles Behind Good Mixed Methods Research (Axinn & Pearce 2006)
• Method counterbalance• Strengths and weaknesses• Investigator involvement• Flexibility in design and application
• Comprehensive empirical record• Multiple kinds of evidence
• Reproduce findings with multiple methods• Discover mechanisms• Highlight temporal ordering
MIXED METHODS RESEARCH & THE RESEARCH PROCESSResearch Questions--Research Design & Data Collection--Data Analysis & Interpretation--Presentation of Results—Publishing--Research Funding
Research Questions (Alford 1998)
Research Problem
RESEARCH QUESTION/S
Theoretical
Empirical
Action Agenda
Mixed Methods Research Questions(Clark & Badiee 2010)
• Level of Integration• Separate questions• Broad general questions• Hybrid question• Procedural/mixing question• Combination
• Relationship Between Questions• Independent/Dependent
• Relationship of Questions to the Process• Predetermined/Emergent
Surveys Semi-Structured Interviews Focus Groups
ObservationHistorical/Archival/
MaterialsExperimental
Components of Research Design & Data Collection
Figure 1. Comparison of Structure, Interviewer Involvement, and Researcher Involvement among Data Collection Methods (Axinn & Pearce 2006)
Data Collection Method Level of Structure
Interviewer Involvement
Researcher Involvement with Study Population
Surveys High Usually Low
Semi-Structured Interviews
Low Always High
Focus Groups Low Always Medium
Observation Low Usually High
Historical or Archival Methods
Out of researcher’s
control
Out of researcher’s
controlLow
Figure 2. Comparing Sample Size and Coding among Data Collection Methods (Axinn & Pearce 2006)
Data Collection Method
Data on Large Numbers of People Could Be Coded As
NumbersCould Be Analyzed As
Text
Surveys Usual approach X X
Semi-Structured Interviews
Possible X X
Focus Groups Possible X X
Observation Possible X X
Historical or Archival Methods
Possible X X
Features of Design
• Method balance• Equal or unequal
• Method order• Sequential• Parallel/simultaneous/concurrent
• Method integration• Data collection
• Sampling (Teddlie & Yu 2007)• Choice of methods• Design of instruments
• Data analysis, inference, and interpretation• One set of findings guides another analysis• Two sets of findings compared/contrasted
• Multiple points of integration
Focus Groups Survey RESULTS
Less structured methods used to help develop survey measures and
instruments
EXAMPLE:Zeller, Richard A. 1993. “Combining Qualitative and Quantitative Techniques to Develop Culturally Sensitive Measures.” in David G. Ostrow and Ronald C. Kessler (Eds.) Methodological Issues in AIDS Behavioral Research. New York: Plenum Press, pp. 95-116.
SurveySemi-Structured
InterviewsRESULTS
Survey methods used to set the stage for less
structured investigation
EXAMPLE:David J. Harding. 2009. “Violence, Older Peers, and the Socialization of Adolescent Boys in Disadvantaged Neighborhoods.” American Sociological Review 74(3): 445-464.
Semi-StructuredInterviews &Observations
Survey RESULTS
Survey data used to test findings from a less
structured method
EXAMPLE:Small ML, Jacobs EM, and Massengill RP. 2008. "Why Organizational Ties Matter for Neighborhood Effects: A Study of Resource Access through Childcare Centers." Social Forces. 87(1).
Experiment/Audit Study
Observation &Fieldnotes
Results
Less structured investigation used to help
explain quantitative findings
EXAMPLE:Pager, Devah, Bruce Western, and Bart Bonikowski. 2009. “Discrimination in a Low-Wage Labor Market: A Field Experiment.” American Sociological Review 74: 777-799.
Semi-StructuredInterviews with
Children & Mothers
Survey of Teachers
RESULTS
Parallel use and equal balance of methods
EXAMPLE: Timberlake (1994) from Tashakori & Teddlie (1998)
Systematic Anomalous Case Analysis (Pearce 2002)
• QUANT• Theory-informed survey analyses• Regression diagnostics to identify anomalous (and matched) cases• Sampling
• QUAL• Fieldwork• Findings spark new & revised hypotheses
• QUANT• Additional survey analyses
• Latent Class Analysis • Types of Religiosity
• Predicted by survey data• Using semi-structured interviews to confirm and/or flesh out
types
Typology Development(Pearce & Denton 2011)
Five Latent Classes of Religiosity
Hybrid Methods
• Event history calendars• Axinn & Pearce 2006
• Life histories from survey data & records• Singer, Burton H., Carol D. Ryff, Deborah Carr, and William
Magee. 1998. “Linking Life Histories and Mental Health: A Person-Centered Strategy.” Sociological Methodology 28:1-51.
• Q-sort methods• Shemmings, D. 2006. '"Quantifying” qualitative data: an
illustrative example of the use of Q methodology in psychosocial research’. Qualitative Research in Psychology 3 (2). pp. 147-165.
• Grid design• Columns = time periods (years, months, etc)• Rows = domains of interest
• Visual, cooperative process• Allows for memory cues• Allows flexibility
• Question ordering• Probing
Calendar Methods
Life History Calendar
Neighborhood History Calendar
Questionnaire Calendar
• Gives a visual structure to interviewing• A tool for reflection
Using LHCs in SS Interviews