Conference PaperPDF Available

Problems & possibilities: Exploring paradigms for mixed methods research

Conference Paper

Problems & possibilities: Exploring paradigms for mixed methods research

Abstract

Mixed methods research is (increasingly) popular in education and other social sciences, partly because of the perceived benefits of exploring a phenomenon using both quantitative and qualitative approaches. Mixed methods studies can sometimes be driven by practical rather than philosophical considerations - what might work, what is achievable, what seems useful, or even what is most likely to get funded. However, the underpinning paradigm/s that guide mixed methods studies warrant careful thought, particularly because the relationship between research methods and paradigms is a contested topic among mixed methods researchers. This presentation will be in two parts. First, I will review the various stances that scholars have taken on method-paradigm relationships in mixed methods research, highlighting issues that researchers need to consider. Second, I will share an example of a mixed-methods study with an unusual method-paradigm combination, showing how an interpretivist stance was woven into all aspects of my doctoral research on teachers' experiences of professional development in Abu Dhabi. This presentation contributes in response to calls for more explicit consideration of paradigms in mixed methods research as well as calls for more specific exemplars illustrating the possibilities of mixed methods research.
Problems &
possibilities: Exploring
paradigms for mixed
methods research
KATRINA McCHESNEY
UNIVERSITY OF WAIKATO
SPECIA L ISSUE:
A k a l e i d oscope of pe r spec t i ves on the p ote n t i a l, contr i b u t ions, a n d grand
visi o n o f a mix e d met h o d s a p p roach to e d u c a t ional i n quir y
4 relevant problems
Unresolved debate re relationships between methods & paradigms in mixed-methods research
(Greene & Caracelli, 1997; Johnson & Onwuegbuzie, 2004; Morgan, 2007; Bazeley, 2009;
Bergman, 2010; Creswell & Plano Clark, 2011; Mertens, 2012; Shannon-Baker, 2016;
Fetters & Molina-Azorin, 2017b; Hathcoat & Meixner, 2017)
Mixed methods research seen as “insufficiently rigorous” (Bergman, 2010)
Reporting of studies often omits discussion of paradigm / positioning (Alise & Teddlie, 2010)
Lack of exemplars of robust mixed methods research (Bryman, 2007)
Paradigms & methods
Paradigm = “a worldview, together
with the various philosophical
assumptions associated with that
point of view”
(Teddlie and Tashakkori, 2009)
Method/s = “tools and techniques”
for data collection and analysis
(Guba & Lincoln, 1989)
Paradigm /
philosophical
positioning
Methods
Methodology
(Creswell, 2010; Guba & Lincoln,
1989; Mackenzie and Knipe 2006)
Combining methods & paradigms:
5 stances
1. Binary stance particular methods belong with particular paradigms
Problems for mixed-methods research:
Incommensurability (Kuhn 1970; Lincoln 1990) mixed-methods impossible?
Lack of genuine qual-quant integration (Yin 2006; Tashakkori & Creswell 2007b).
Combining methods & paradigms:
5 stances
2. A-paradigmatic stance side-stepping paradigms in favour of ‘what works’
“Paradigms could be important for methodology but should not be used to inform the
inquiry process’ (Shannon-Baker, 2016)
Problems for mixed-methods research:
Credibility
Validity
Coherence
Combining methods & paradigms:
5 stances
3. Dual / dialectical stance combining 2 paradigms (1 for Qual, 1 for Quant) in one study
Problems for mixed-methods research:
Incommensurability again
Reduction of one paradigm in favour of the other
Two parallel but distinct studies on same topic (qual / quant)
Combining methods & paradigms:
5 stances
4. Pragmatist stance (subset of dual / dialectical) paradigms are logically independent so can
be mixed and matched harmlessly based on what suits the research question/s
“The pacifier in the paradigm war” (Bergman, 2010)
“It is one thing to endorse pluralism [or dialecticalism] … but it is quite another to build a social science
on a what-works pragmatism. It is a mistake to forget about paradigm, epistemological, and
methodological differences between and within QUAN/QUAL frameworks. These are differences that
matter.” (Denzin, 2012)
Problems for mixed-methods research:
Incommensurability
Reduction of one paradigm in favour of the other
Two parallel but distinct studies on same topic (qual / quant)
Combining methods & paradigms:
5 stances
5. Holistic / single-paradigm stance one paradigm for the whole of a mixed-methods study
“If it suits their purposes, any of the theoretical perspectives could make use of any of the
methodologies”
(Crotty 1998, 12; see also Guba & Lincoln 1994; Johnson & Onwuegbuzie 2004;
Mackenzie & Knipe 2006; Gray 2013)
Not a new idea BUT to date most mixed-methods research using this stance has taken an
overarching positivist or post-positivist stance
(Giddings 2006; Alise and Teddlie 2010; Hesse-Biber 2010; Denzin 2012; Torrance 2012)
My doctoral research
“Investigating teachers’ experiences of professional development within a major education reform
in the Emirate of Abu Dhabi” (2017; http://hdl.h&le.net/20.500.11937/57566)
Interpretivist paradigm underpinning all aspects of a mixed-methods study
Interpretivism (Willis 2007):
understanding (not explaining, generalizing or critiquing) the fundamental aim of research
understanding-in-context situatedness of knowledge
Research objectives
Focus on teacher perceptions
Verbs investigate, examine, compare (not measure, evaluate, determine …)
Extra objective added in response to teachers’ voices
Interpretivism: “Approaches that leave open the opportunity to discover things as the
research progresses” (Willis, 2007)
Data sources
Teacher perceptions = data from teachers (no attempt to “triangulate” via other stakeholders)
Qual & quant interwoven
Sample Quant Qual
393 teachers Survey × 2
786 sets of responses
Open response item
96 comments
35 teachers Survey × ~8.5
297 sets of responses In-depth interviews
Gathered
together
Gathered
together
Sampling
Main focus = capturing the diverse range of experiences in the wider target population
Larger sample:
Invitation to all teachers in target population
Targeted data gathering to improve representation
Smaller sample:
Purposive + snowball sampling
Interpretivism: seeking to capture “the multiple perspectives that are inherent in most human
endeavors” (Willis, 2007)
Data analysis, interpretation, &
reporting (1)
Most objectives separate qual & quant analyses then brought together to draw holistic
conclusions (Bryman, 2007).
Quant: Means, standard deviations, confidence intervals, correlation, multiple regression, ANOVA
Qual: Thematic analysis, constructivist grounded theory (Charmaz)
Reporting for individual objectives some sequential, some interwoven
Cross-study reporting / synthesis
“In genuinely integrated studies, the quantitative and qualitative findings will be mutually informative. They
will talk to each other … and the idea is then to construct a negotiated account of what they mean
together … Mixed methods research … is about forging an overall or negotiated account of the findings
that brings together both components of the conversation or debate.” (Bryman, 2007)
Data analysis, interpretation, &
reporting (2)
Alert to possible impact of
cultural differences in
communication (e.g.
Hofstede et al., 2010)
Key idea: all data reflected
teachers’ constructions of
meaning not absolute
truth
Example: Arab vs Western
teachers’ perceptions of
different forms of PD
Quality considerations
Existing quality criteria for mixed-methods research (Howe 2012; Torrance 2012):
Post-positivist worldview
Qual & quant being used to develop a single definitive / generalisable understanding of the research topic
Instead: Specific practices to enhance quality & validity in interpretivist / social constructivist
research
Creswell & Miller (2000)
Willis (2007)
E.g. member checks; extended researcher experience in the research environment; peer review;
researcher journaling; audit trails; disconfirming evidence; thick, rich qualitative description …
Zooming out paradigms &
methods
Whichever is chosen first (methods or paradigm/s) does not predetermine the other
Not all method/paradigm combinations may make sense but certainly interpretivism can work
for single-paradigm mixed method studies
3 recommendations (McChesney & Aldridge, 2019):
1. That the paradigmatic or philosophical underpinnings of any research study be explicitly stated
2. That both the paradigm(s) and methods selected be suitable to allow
the aims and objectives of the study to be met
3. That researchers demonstrate how the research methods and the
overall conduct of the study reflect or acknowledge the chosen
paradigm(s), making explicit and justifying design decisions
“Using multiple and diverse methods is
a good idea, but is not automatically
good science … Lacking justification and
planning, mixed method decisions may
not be defensible.”
GREENE AND CARACELLI, 1997
... This article has emerged from two prompts. The first was my contribution to the 2020 Cancelled Conference Conversations series (McChesney, 2020). My presentation centred on the question of paradigms for mixed methods research and drew a lot of interest from postgraduate students in particular. ...
Article
Full-text available
Mixed methods research is increasingly popular both within and beyond education because of the advantages offered by combining qualitative and quantitative methods. Some mixed methods research, however, does not fully harness the potential or depth that mixed methods has to offer. In this article, I consider some of this potential in terms of how mixed methods research can contribute to addressing "wicked problems," theory generation, and culturally responsive research. I then discuss two important considerations for quality mixed methods research: appropriate paradigmatic foundations and the genuine integration of qualitative and quantitative components. The article is intended to provide both provocations and resources for those learning about, teaching about, considering, using, or contributing to mixed methods research in education.
ResearchGate has not been able to resolve any references for this publication.