DePaul University Emotional and Social Experiences of The Child Essay

2EJAMES H. BANNING, PROFESSOR OF EDUCATION, COLORADO STATE UNIVERSITY
‘Qualitative Data Analysis is an excellent text. Carol has impressively woven detailed
examples of qualitative data with appropriate research theories, which enables a better
understanding of analytical approaches in the research process.’
TEJENDRA PHERALI, SENIOR LECTURER IN EDUCATION STUDIES AND SOCIOLOGY,
LIVERPOOL JOHN MOORES UNIVERSITY
In this fully updated and expanded second edition, Carol Grbich provides a guide
through current issues in the analysis of qualitative data. Packed with detailed
examples, a glossary, further reading lists and a section on writing up, this book is
exactly what you need to get you started in qualitative research.
The new edition covers analytical approaches including:
•
•
•
•
•
•
•
grounded theory
classical, existential and hermeneutic phenomenology
feminist research including memory work
classical, auto- and cyber ethnography as well as ethnodrama
content, narrative, conversation and discourse analysis
visual interpretation
semiotic, structural and poststructural analyses.
Qualitative Data Analysis
‘Grbich’s new edition provides succinct and useful coverage of the newer qualitative
approaches. It is a must read for those seeking an understanding of data analysis from
a perspective that includes an understanding of paradigms, traditions of inquiry and
traditional analytical approaches.’
A one-stop-shop for students new to qualitative data analysis!
www.sagepub.co.uk/grbich2
grbich_qualitative2.indd 1-3
Grbich
Carol Grbich is a Professor in the School of Medicine at Flinders University
in South Australia.
second
edition
Qualitative Data Analysis
An Introduction
Carol Grbich
18/10/2012 12:36
grbich_qualitative2.indd 7
19/10/2012 13:37
00-Grbich Prelims.indd 1
19/10/2012 6:19:47 PM
SAGE has been part of the global academic community
since 1965, supporting high quality research and learning
that transforms society and our understanding of individuals,
groups, and cultures. SAGE is the independent, innovative,
natural home for authors, editors and societies who share
our commitment and passion for the social sciences.
Find out more at: www.sagepublications.com
Connect, Debate, Engage on Methodspace
Connect with other researchers and discuss
your research interests
Keep up with announcements in the field,
for example calls for papers and jobs
Discover and review resources
Engage with featured content such as
key articles, podcasts and videos
Find out about relevant conferences
and events
Q
C
Connecting the Research Community
www.methodspace.com
00-Grbich Prelims.indd 2
07/09/2012 3:39:12 PM
second
edition
grbich_qualitative2.indd 8
19/10/2012 13:37
00-Grbich Prelims.indd 3
19/10/2012 6:19:59 PM
SAGE Publications Ltd
1 Oliver’s Yard
55 City Road
London EC1Y 1SP
SAGE Publications Inc.
2455 Teller Road
Thousand Oaks, California 91320
SAGE Publications India Pvt Ltd
B 1/I 1 Mohan Cooperative Industrial Area
Mathura Road
New Delhi 110 044
SAGE Publications Asia-Pacific Pte Ltd
3 Church Street
#10-04 Samsung Hub
Singapore 049483
Editor: Katie Metzler
Editorial assistant: Anna Horvai
Production editor: Katie Forsythe
Copyeditor: Elaine Leek
Proofreader: Bryan Campbell
Marketing manager: Ben Griffin Sherwood
Cover design: Jennifer Crisp
Typeset by: C&M Digitals (P) Ltd, Chennai, India
Printed and bound by CPI Group (UK) Ltd,
Croydon, CR0 4YY
© Carol Grbich 2013
First edition published 2006. Reprinted 2007, 2009 (twice),
2010, 2011 (twice)
This edition first published 2013
Apart from any fair dealing for the purposes of research or
private study, or criticism or review, as permitted under the
Copyright, Designs and Patents Act, 1988, this publication
may be reproduced, stored or transmitted in any form, or by
any means, only with the prior permission in writing of the
publishers, or in the case of reprographic reproduction, in
accordance with the terms of licences issued by the Copyright
Licensing Agency. Enquiries concerning reproduction outside
those terms should be sent to the publishers.
Library of Congress Control Number: 2011945803
British Library Cataloguing in Publication data
A catalogue record for this book is available from
the British Library
ISBN 978-1-4462-0296-8
ISBN 978-1-4462-0297-5 (pbk)
00-Grbich Prelims.indd 4
07/09/2012 3:39:12 PM
Table of contents
About the Author
vii
Part 1
General approaches to collecting and analysing qualitative data
1
1
Introduction
3
2
Design methodologies, data management and analytical approaches
15
3
Incorporating data from multiple sources: mixing methods
25
Part 2
Traditional analytical approaches
39
4
Classical ethnography
41
5
Critical ethnographic approaches
55
6
Feminist approaches
68
7
Grounded theory
79
8
Phenomenology
92
Part 3
Newer qualitative approaches
105
9
107
Postmodern influences on society and qualitative research
10 Autoethnography
119
11 Poetic inquiry
129
00-Grbich Prelims.indd 5
07/09/2012 3:39:12 PM
12 Ethnodrama and performative art
143
13 Cyber ethnography and e-research
154
Part 4
Analytic approaches for existing documentation
165
14 Structuralism and poststructuralism
167
15 Semiotic structural and poststructural analysis (deconstruction)
176
16 Content analysis of texts
189
17 Content analysis of visual documents
200
18 Narrative analysis
216
19 Conversation analysis
229
20 Discourse analysis
245
Part 5
Data management using qualitative computer programs
257
21 Coding
259
22 An overview of qualitative computer programs
268
Part 6
Interpreting and presenting qualitative data
289
23 Theorising from data
291
24 Writing up and innovative data display
302
Glossary
Index
326
328
Solutions to the exercises and PowerPoint slides for lecturers are available at
www.sagepub.co.uk/grbich2
vi
00-Grbich Prelims.indd 6
QUALITATIVE DATA ANALYSIS
18/10/2012 11:35:31 AM
About the author
Dr Carol Grbich is a Professor in the School of Medicine at Flinders University in
South Australia. She is an Epidemiologist and Sociologist and is the author of a number of textbooks on Qualitative Research including Qualitative Data Analysis: An
Introduction 1st Edition (Sage, 2007), New Approaches in Social Research (Sage, 2004)
and Qualitative Research in Health: An Introduction (Allen and Unwin, 1999), as well as
authoring several texts on the Sociology of Health and Illness.
She is also Foundation Editor of the International Journal of Multiple Research
Approaches.
00-Grbich Prelims.indd 7
07/09/2012 3:39:13 PM
00-Grbich Prelims.indd 8
07/09/2012 3:39:13 PM
PART 1
General approaches to collecting
and analysing qualitative data
The processes of data analysis in qualitative research are complex. It is not simply a
matter of choosing and applying an accepted process such as thematic analysis.
A combination of three key areas is involved:
•• the first is to do with you, the researcher – your views and choices in the research journey and
the impact of these on the data you collect and analyse
•• the second relates to the design and methods used, the quality of the data you have gathered
and how you have managed it, and
•• the third involves your display of findings and your theoretical interpretation of your analysed
data, presented for the reader to assess.
The three Ps – Person, Processes and Presentation – are key issues here.
This book starts from the premise that these three elements are essential to any undertaking of qualitative data analysis, and the examples and strategies presented attempt to
indicate how they integrate.
Part 1 introduces you to the background information required for understanding qualitative research. The first chapter deals with research characteristics, investigative areas of
research and the role of the researcher, together with his/her influence on the data and finishing with a brief review of the major paradigms that have underpinned qualitative research.
The chapter concludes with the important issue of how to evaluate qualitative research.
The second chapter introduces you to the main tools for data collection, transcription
and preliminary data analysis – the analysis you undertake while you collect your data.
The four most common analytic approaches are then discussed and the range of methodologies currently available for you to choose amongst is displayed. The third chapter
deals with one of the newer trends in research: the mixing of qualitative and quantitative
approaches, termed multiple or mixed methods.
Chapter 1 Introduction
Chapter 2 Design methodologies, data management and analytical approaches
Chapter 3 Incorporating data from multiple sources: mixing methods
01-Grbich-Ch-01.indd 1
07/09/2012 3:39:35 PM
01-Grbich-Ch-01.indd 2
07/09/2012 3:39:35 PM
ONE
Introduction
When you undertake a qualitative research study, there are a number of prior aspects that will
need addressing: your proposed topic area, is it suitable for the collection of qualitative data?; then
yourself, what impact will your prejudices have on the research and how will you treat your potential
readers?; and finally, which paradigm would best fit your research question: Realism/postpositivism,
Critical theory, Interpretivism/Constructionism, Postmodernism, Poststructuralism or Mixed/multiple
methods? Then, having dealt with these decisions, it is always useful to know how to evaluate a
piece of qualitative research in order to ensure that your design ticks all the right boxes.
KEY POINTS
•
•
•
•
•
The characteristics of qualitative research
The best topic areas for qualitative research investigations
Issues you need to think about prior to commencing research
Research paradigms
How to evaluate qualitative research
Introduction
Qualitative research is a fascinating topic. It provides detailed information and can
progress knowledge in a variety of areas: it can help assess the impact of policies on a
population; it can give insight into people’s individual experiences; it can help evaluate
service provision; and it can enable the exploration of little-known behaviours, attitudes
and values.
Knowledge is a key term here. This can take a variety of forms, and in most cultures
there are various claims to knowledge:
1 The first is tenacity – this refers to a belief that has been held for a long time, for example doing
good to others is viewed as the right thing to do because eventually this good will be reflected back
to you; there may be no evidence to prove this is true but we still tenaciously claim that this is so.
01-Grbich-Ch-01.indd 3
07/09/2012 3:39:35 PM
2 Intuition or our gut feeling is another source of knowledge – for example, we may feel that in a
particular situation X is the best thing to do or the right answer; there may again be very little
evidence that this is so but if it feels right, we tend to follow that particular path.
3 Authority – in particular religious or legal authority provides directions for the way we ought to
behave in order to lead a ‘good’ life … but good for whom? In addition, would another way be
more beneficial to us as individuals or as part of a group?
Research tries to step back from knowledge claims developed through tenacity, intuition
and authority, by carefully constructing a question and a study design in order to provide
the best views of a particular issue so that conclusions can be derived from available evidence. Sometimes findings will challenge the other three sources leading to conflict; for
example, when scientists said the earth was not flat but round there was a huge uproar as
traditional beliefs about falling off the edge of the world were challenged.
In doing research we try to advance knowledge by aiming to get closer to the ‘truth’ of
the matter while realising that truth is a very elusive concept, which shifts depending on
whose truth is being portrayed and whether that ‘truth’ is:
•• Subjective (your own view)
•• Relative (your view compared to others)
•• Objective (taking a distant perspective)
•• Absolute (as in philosophical arguments).
So rather than getting too caught up in the notion of ‘truth’ and the bases of various
claims to knowledge in research, instead we seek to reduce uncertainty by using the best
and most transparent approaches available.
There are two important aspects to any kind of research: the first is that your data
should be collected from the real world … from situations or people involved in whatever the defined research problem is. This real world evidence is termed empirical data.
Understanding the nature of this data is an ontological process and is related particularly
to the wider structural and cultural issues that influence claims to truth. Then these
understandings need to be further interpreted in a more abstract way using existing
theories of knowledge – epistemology – to explain your findings about the world and to
enable your interpretations to be more globally applied. For example, I might research
the experience of being blind by interviewing people who are blind (empirical data).
Understanding their experiences would require knowledge of the culture and the health
system and other supports available for these people (ontological) while interpreting their
experiences might lead me to use the concepts of stigma or normal versus abnormal (epistemology) to make sense of their experiences.
What are the characteristics of qualitative research?
Qualitative research favours certain styles of design, collection and analytic interpretation. The underpinning ideology or belief system asserts that:
•• subjectivity has value (meaning that both the views of the participant and those of you the
researcher are to be respected, acknowledged and incorporated as data, and the interpretation
of this data will be constructed by both of you (the researcher is not a distant neutral being)
4
01-Grbich-Ch-01.indd 4
GENERAL APPROACHES
07/09/2012 3:39:35 PM
•• validity (trustworthiness) is seen as getting to the truth of the matter, reliability (dependability)
is viewed as a sound research design and generalisability is local and conceptual only
•• power lies predominantly with the researched (who are viewed as being the experts on the
research topic)
•• an holistic view is essential (so the structures impacting on the setting such as policies, culture,
situation and context need to be included)
•• every study is time- and context-bound (so that replication and generalisation are unlikely
outcomes).
Which areas are best for researching?
Qualitative research can best help us explore or assess:
•• culture
•• phenomena
•• structural processes
•• historical changes.
In more detail, culture could involve anything from investigating the behaviours and
rituals of a particular tribe or group of people in a particular setting (street kids, pupils
or staff in a classroom, patients or clinicians in a hospital ward or an individual in a
particular cultural context). Phenomena involves detailed investigations over time of a
particular experience (for example, marriage breakdown, illness etc.). Structural processes might involve investigating policy change and its impact on a specified setting
or group (such as increasing taxes or closure of mental institutions). And historical
changes might involve documented changes in discourses (ways of communicating
over time; for example, changes in treatment of an illness as recorded in medical
journal articles).
The question focus is usually the what, how, when, where or why aspects of the chosen topic.
One important issue the qualitative researcher needs to consider prior to commencing
research is the choice of research paradigm to work within.
Research paradigms
As researcher, you can choose which of the available broad paradigms (worldviews of
beliefs, values, and methods for collecting and interpreting data) that you would prefer
to work within.
There are five options:
1 Realism/postpositivism (expert researcher documenting reality from a centred position).
2 Critical theory (with a focus on class, power and the location and amelioration of oppression).
3 Interpretivism/Constructionism (mutual recognition and use of symbols and signs in reality
construction).
4 Postmodernism and poststructuralism (the questioning of ‘truth’ and ‘reality’ and the sources of
‘knowledge’).
5 Mixed/multiple methods (using the best set of tools for the job).
Let us explore each of these in a little more detail.
INTRODUCTION
01-Grbich-Ch-01.indd 5
5
07/09/2012 3:39:36 PM
1. Positivism to realism (postpositivism)
The eighteenth century in Europe was an era, termed the Enlightenment, when positivism (the School of Philosophy that asserts that reality lies only in things that can be seen
with the naked eye), optimism, reason and progress became the dominant discourses
(ways of thinking, speaking and writing) and all knowledge was believed to be accessible
through processes of reason. The ‘rational man’ was believed to have the capacity to
uncover a singular knowable reality through pure understanding and rigorous intellectual reasoning. These processes of broader reason, needed to gain knowledge, included a
focus on observation in order to gain ‘facts’ via scientific deduction. Scientific knowledge
gained from observation and based in logical thought processes was seen as having the
potential to displace ignorance and superstition, which were the tools of power of the
church. Scientific knowledge was seen as having the capacity to facilitate freedom from
religious influences and to lead the way to a New World built on the notions of progress
and a universal foundation of knowledge.
However, researchers’ ability to provide predictable and replicable outcomes and to control variables came under debate as Einstein’s theory of relativity and later Heisenberg’s
theory of uncertainty challenged these views and postpositivism eventuated … The assumption that a world that could be precisely measured and documented exists independently
just waiting for us to gain sufficiently sophisticated tools to discover it, was questioned, and
the belief that absolute, knowable truth existed became sidelined and provisional truths
became a more likely outcome. The ultimate essence of external reality was also challenged
by Sigmund Freud’s exploration ([1900]1913) of the unconscious mind as a source of reality
construction. He suggested that ‘reality’ was not only constructed from internal as well as
external sources but that this reality changed continually in interaction with the environment, especially in interaction with others, and that what had previously been considered
as externally and objectively ‘real’ was also closely linked to the maintenance of power.
More recently within postpositivism it has been argued that scientists are inherently biased
by their education and life experiences and that their observations are value-laden and fallible, making errors likely. Our ability to know reality with certainty is thus problematic and
no findings can be viewed as absolute or universally generalisable. This has led some positivists to the modified epistemology of realism. Realism asserts that structures creating the
world cannot always be directly observed and when and if they are observable their genesis
is not always clear; thus we also need our creative minds to clarify their existence and
then to identify explanatory mechanisms. For example, we cannot see gravity but we
know it exists and that it requires a mixture of intuition, various intellectual processes, and
the laws of physics in order to clarify the workings of this force. The focus for research in a
realist approach involves the identification of the linking of different realisms, for example
in nursing, the biological and psychosocial models of nursing can be linked to a biopsychosocial model that has bridging links to interpretations of biological mechanisms and to the
psychosocial empirical world as well as to patients’ and researchers’ influences on these.
Types of research
In terms of qualitative research, both postpositivism and realism
draw from positivism in that the researcher is seen as occupying a pseudo-objective distant neutral role where their influence in the construction of reality is seen as minimal.
6
01-Grbich-Ch-01.indd 6
GENERAL APPROACHES
07/09/2012 3:39:36 PM
Classical ethnography (see Chapter 4) and Straussian grounded theory (see Chapter 7)
are sometimes seen as fitting in to this orientation. Careful description, truthful depiction,
studies with clear aims, objectives, a reliable design, a focus on neutrality, objectivity and
theory-testing characterise these approaches.
2. Critical emancipatory positions
Changes in the economic system through industrialisation around the turn of the
twentieth century led to Karl Marx’s critique ([1867]1999) of capitalist exploitation,
profit, power and class conflict, being recognised. The outcomes of such economic change
became viewed as resulting in societal fragmentation. During the 1960s and 1970s social
critics such as feminists identified power imbalances and pointed to the long-term oppression of women by men, while others pointed to inequalities in social justice. Reality was
now being viewed as power directed and multiply constructed. The origins of ‘truth’ were
seen as lying in obscure history and/or layered aspects of the present and access required a
range of approaches including those beyond the scientific. The simplicity of such notions
as the integration of the individual, the power of the author, the universality of knowledge and concepts of uniqueness and originality, came under question.
Types of research
In research, critical positions view reality not as existing ‘out
there’ but as being produced by particular exploitative social and political systems
comprising competing interests where knowledge is controlled to serve those in power.
Issues of race, gender, poverty, politics and culture are seen to shape individual identity. Researchers attempt to identify those who are powerless (usually exploited by
those in powerful positions) in order to document their unequal situation and to bring
about change through an active process of emancipation through knowledge-sharing or
the transformation of society. Any qualitative approach that has taken a critical stance,
including grounded theory, phenomenology, ethnography, hermeneutics, sociolinguistics, narratives, and feminist research (see later chapters for these) can fit into the critical
emancipatory grouping.
3. Constructionism/Interpretivism
These positions assume that there is no objective knowledge independent of thinking. Reality
is viewed as socially and societally embedded and existing within the mind. This reality is
fluid and changing, and knowledge is constructed jointly in interaction by the researcher
and the researched through consensus. Knowledge is subjective, constructed and based on
the shared signs and symbols that are recognised by members of a culture. Multiple realities
are presumed, with different people experiencing these differently. The research focus is on
exploration of the way people interpret and make sense of their experiences in the worlds
in which they live and how the contexts of events and situations and the placement of
these within wider social environments have impacted on constructed understandings.
The understandings researchers construct and impose through interpretation are seen as
limited by: the frames derived from their own life experiences; subjectivity (the researcher’s
own views and how they have been constructed); and intersubjectivity (reconstruction of
views through interaction with others through language and written texts).
INTRODUCTION
01-Grbich-Ch-01.indd 7
7
07/09/2012 3:39:36 PM
Types of research
Qualitative methodologies including grounded theory, phenomenology, ethnomethodology, ethnography, hermeneutics, sociocultural narratives, and
feminist research (see Chapters 4–20 for more on these approaches).
4. Postmodernism and poststructuralism
As we moved through the last decades of the twentieth century, unified, powerful, centred
individuals with an authoritative point of view became rejected in favour of anti-heroes
and complex multidimensional individuals (see Chapters 9 and 14 for more detailed
explanations of postmodernist ideas and applications). Literature began to mirror the
changes in the economy, science, art and architecture by portraying reality as shifting
and uncertain rather than set, and by incorporating multiple perspectives from a
range of disciplines such as music, philosophy, psychology, sociology and drama as
well as including visual possibilities.
Postmodernism views the world as complex and chaotic and reality as multiply constructed and transitional – unable to be explained solely by grand or meta narratives
(such as Marxism and Buddhism, which make universal claims to truth). Postmodernism
is very sceptical of such narratives, viewing them as containing power-laden discourses
developed specifically for the maintenance of dominant ideas or to enhance the power
of certain individuals. The search for reality ‘out there’ is qualified by the understanding
that society, laws, policies, language, discipline borders, data collection and interpretation
are culturally and socially constructed. In recognition of this socially constructed world,
disruption, challenge and a multiplicity of forms are essential in order to pull these constructions apart and to expose them for what they are. Meaning rather than knowledge
is sought because knowledge is limited by ‘desire’ (lack of knowledge or the imperative
to bring about change) and constrained by the discourses developed to protect powerful
interests and to control the population’s access to other explanations. Truth is multifaceted and subjectivity is paramount.
Poststructuralism, with its emphasis on the fluidity of language and meaning, forms an
important subset of postmodernism. It developed as a reaction to structuralism, which
sought to describe the world in terms of systems of centralised logic and formal structures. In structuralism, patterns provided meaning and all words were seen as having
recognised meanings that could be learned. Language was seen as a system of signs and
codes, rules and conventions – and the deep structures that enable a language to operate
within a cultural system – were sought. Poststructuralism (see Chapter 14) seeks the
deconstruction of the discourses (ways of thinking, speaking and writing) that have been
established to control ways of thinking.
Types of research Most forms of qualitative research now have an established postmodern position: for example, ethnography, grounded theory, action, evaluation research,
phenomenology and feminist research. Postmodernism favours descriptive and individually interpreted mini-narratives, which provide explanations for small-scale situations
located within particular contexts where no pretensions of abstract theory, universality, or generalisability are involved. Within structuralism and poststructuralism two
data analytic approaches have become popular and are available for use by qualitative
8
01-Grbich-Ch-01.indd 8
GENERAL APPROACHES
07/09/2012 3:39:36 PM
researchers. The first is discourse analysis, where the dominant ways of writing and
speaking about a particular topic become set in place over time and require historical
tracking back to identify who has benefited from one particular discourse and how
other competing discourses have been marginalised. The second analytic approach is
deconstruction, where words are viewed as containing power-laden discourses with
multiple meanings requiring careful deconstruction in order to break down artificially
constructed boundaries before putting the text back together in transitional form.
5. Mixed/Multiple methods
This is the most recent approach and follows postmodernism’s exhortion to cross barriers and to break down boundaries. The two approaches, qualitative and quantitative – for
decades seen as poles apart – have now become integrated into mixed/multiple method
studies (sometimes called the third wave/third movement). In this situation, they are seen
less as two approaches ideologically poles apart and more as an eclectic set of tools which
you the researcher – very like the bricoleur (creative handyman) of postmodernism – can
use to provide the best answers to your research question. Clearly the issues involved
in utilising these very different approaches can be somewhat thorny but this has not
prevented researchers from tackling these issues head on and providing ways of dealing
with them. The changes in classical physics which provide the underpinning for quantitative approaches, particularly the movement into chaos and complexity theory, have
reflected many of the postmodern thought changes seen in qualitative research (Grbich,
2004) and these changes may have facilitated this cooperation. The ensuing paradigm
has often become termed ‘pragmatism’ – a mix of postpositivism and social constructivism, a leaning toward postmodernism, and an emphasis on empirical knowledge, action,
triangulation and the changing interaction between the organism and its environments
(see Chapter 3 for more detail regarding mixed methods).
Example of paradigm choice
Let us take a research topic, ‘An exploration of the lives of young people who are homeless’, and
see how your position as a researcher would differ in each of the above five paradigms:
•• Realism/postpositivism (expert researcher documenting reality from a centred position). Here an
authoritative researcher would assume that truth can be found by gathering detailed accurate
observational and interview data of the lives of young people living on the street.
•• Critical theory (with its focus on class, power and the location and amelioration of oppression). Here
the interpretation of the data you collect would focus on power – where does it lie? And the assumption would be that the structures of society (education, health and the socioeconomic influences of
the culture) would be determining aspects for a situation where young people became homeless.
Action research – working with the homeless to bring about change – might be an outcome.
•• Interpretivism/Constructionism (mutual recognition and use of symbols and signs in reality construction). Both the aspects of individual choice and lack of choice would be taken into account
here as each individual case is explored by you in conjunction with a homeless person.
•• Postmodernism and poststructuralism (the questioning of ‘truth’ and ‘reality’ and the sources
of ‘knowledge’). Previous explanations would be rigorously questioned and the discourses of
INTRODUCTION
01-Grbich-Ch-01.indd 9
9
07/09/2012 3:39:36 PM
‘homelessness’, ‘begging’, ‘mental illness’ etc. examined and deconstructed. Your assumption
would be that the reasons young people are homeless are individual, complex and always
changing and no one solution will fit all.
•• Mixed/Multiple methods. Both qualitative and quantitative data will be needed to see broader
aspects of individual circumstances within policy, practice and the views of the wider community.
Evaluation of qualitative research
How can we assess the quality of our qualitative research and that of others? The techniques
by which quantitative research are evaluated are not appropriate but sets of guidelines for
evaluating qualitative research have been suggested (Kitto et al., 2008) and these guidelines
are detailed below seven headings to show the essentials that need to be accounted for in
a good piece of qualitative research:
Clarification
•• What is the research question/s?
•• What are the aims of the research?
•• What did the researcher seek to investigate?
•• Does the research question reflect what has been investigated?
•• Have the aims been translated into the design so that all of them have been accounted for?
Justification
•• Why is a qualitative approach the best option to answer this question?
•• Why was the particular qualitative research design chosen?
•• Why was the study undertaken the way it was? Are the questions, aims and design a perfect match?
•• Were any forms of data triangulation evident? For example, multiple sources, i.e., documents,
interviews, survey data, observation; multiple methods, i.e., mixing methodologies such as ethnography and phenomenology; and multiple theories, where multiple theoretical and conceptual frames have been applied to the research to enhance insights into phenomena.
Process
•• Has ethics approval been obtained?
•• Have the techniques of data collection been clearly documented?
•• How were participants/settings accessed?
•• What sampling techniques have been used to answer the research question?
•• Who was interviewed/observed? How often? And for how long?
•• What interview questions were asked?
•• What was the purpose of any observation/s?
•• Which existing documents were accessed? And how were they assessed?
•• How was collected data managed?
•• Are all the forms of data analysis completely transparent?
•• What were the major outcomes of the analytical process in terms of findings?
In more detail, the exposure of what the researcher actually did needs to be very explicit.
•• How were participants accessed?
•• Who were these participants?
10
01-Grbich-Ch-01.indd 10
GENERAL APPROACHES
07/09/2012 3:39:36 PM
•• How was rapport achieved?
•• Were any sampling techniques used?
•• What data collection techniques were used?
•• How did interviews occur? Face to face? Telephone? Focus group? Teleconference? Video
conference? Email? Skype?
•• Who was observed? When? How often? For how long? For what purpose? What existing sets of
documentation were collected?
•• How was data managed?
•• What forms of data analysis were undertaken – transparency of process is essential here.
Representativeness
Notions of comprehensiveness and diversify of results is sought in qualitative research
in preference to conformity and homogeneity. An audit trail, monitoring changes and
decisions taken in the project, should be recorded in the researcher’s diary and made
transparent where applicable. In addition:
•• Have all the results been reported? Display of results is one aspect of this, and hypertexts to the
original data set so the reader can see where your quotes have come from is becoming common.
•• Has a holistic answer to the research question been achieved?
Interpretation
Has a conceptual discussion of the results and linkage to existing theory/new theory/
models of practice been developed to explain the relevance of findings to a targeted audience or discipline?
Reflexivity
•• Has a clear statement of the impact of the researcher’s views upon the data and the methods
chosen been included?
•• How has researcher position and perspectives shaped the vision, slanted the design and questions and affected the interpretation of results? Has the researcher changed previous views on
this topic? And has the researcher provided a critique of her/his self in the research process
regarding their own history, culture, class, experiences and level of empathy?
Diversity of process, capacity to connect and intertextuality (connections with other relevant sources of influence) as well as the researcher’s own epistemological positioning and
ongoing response to research outcomes, should also be evident.
Transferability
•• Has a critical evaluation of the application of findings to other similar contexts been made?
•• How do results match/contradict others on this topic?
•• Has the relevance of these findings to current knowledge, policy, and practice or to current
research been discussed?
•• To what extent are findings applicable to other similar settings, situations and experiences? And
to what extent has this study successfully contributed to knowledge?
INTRODUCTION
01-Grbich-Ch-01.indd 11
11
07/09/2012 3:39:36 PM
Newer ethnographic approaches
The newer ethnographic practices (documented in Chapters 10–13) are very challenging
to evaluate, assess and/or review as few established criteria exist. The simplest assessment
would be a personal one:
•• Do you feel that you as the reader have been brought as close as is possible to the voice or
images perceived or heard by the researcher?
•• OR, Do you feel you have been led into a mish-mash or collage of bits and pieces so that you
are no closer to experiencing the feelings and emotions of others than you would reading a dry
academic text centred wholly in the authoritative voice of the researcher?
Researcher position
Subjectivity is crucial here.
•• What have been the experiences of the researcher? Exposure of who the author actually is (past
influences, beliefs, values and experiences as well as their responses in all situations) should be
available.
•• Has the researcher been highly involved as a participant in his/her own right or what has been
her/his position?
•• How close to the participants’ view, voices, emotions and feelings is the display of data and how
much ‘shaping’ (changing or manipulating) has the researcher been involved in?
Process
If the design involved small-scale mini-narratives where reality is seen as multiply constructed, multiple methods (both qualitative and quantitative) are often needed to present
a holistic view of any situation or experience. Juxtaposition will often be called upon to
identify voices/perspectives that have previously been marginalised or silenced by powerful discourses. The emphasis will be on the complexity of both situations and language – in
particular via double coding, irony, paradox, the longevity of particular discourses, discursive practices and deconstruction. The seeking of multifaceted realities and the exposure of
complex individuals with past lives as well as current issues and experiences, is desirable.
There is no assumption of universality or generalisability or even transferability of any
findings – these are seen as localised and transitory
Truth – has this been viewed as a complex constructed entity? If so, many voices and
many approaches may be required to expose it. Language, discourse and discursive practices obscure truth and need tracking to enable new but transitory representations to
emerge. If truth is sought through one individual then the multifaceted nature of that
individual is important to demonstrate rather than the display of one simple dimension.
The reader
Has the reader been allowed to interpret data rather than have it interpreted for him/
herself by the researcher? The role of the researcher is to take the reader as close to the
experiences under study as possible with minimal or no researcher interpretation so the
reader can share these experiences.
12
01-Grbich-Ch-01.indd 12
GENERAL APPROACHES
07/09/2012 3:39:36 PM
In addition, the theory of constraints (Agar, 2004: 22–3 adapted) can also be applied to
assessing the outputs from the approaches described in Chapters 10–13:
1 The ‘dialogue’ consideration. Is dialogue between the researcher, the participants and the
audience facilitated?
2 The ‘scaling’ consideration. Do individual stories promote a broader view of society for the
audience?
3 The ‘recognition’ consideration. Is the representation sufficiently ‘realistic’ so that audiences will
recognise themselves in it?
4 The ‘appeal’ consideration. Are the audiences attracted to the style of presentation, narratives
and issues?
You can see from this that the assessment of qualitative research is a complex process
but general guidelines can be applied to both traditional and to newer approaches. These
guidelines need to be flexible in the way they are interpreted so they don’t become set in
concrete – they need to retain the capacity to encompass and reflect the ongoing changes
that are an intrinsic part of qualitative research.
Summary
The paradigms of positivism, postpositivism and realism, critical theory, constructivism/
interpretivism, postmodernism and poststructuralism, and mixed/multiple methods
provide a complex field for you to navigate. Although each epistemology has achieved
dominance at particular times, all are available for you to consider for your own studies. You need to choose the one that best reflects your research question and preferred
orientation or you can choose to blend different traditions. In both cases, you will need
to be familiar with the traditions on offer so you can adequately justify and evaluate the
choices you have made.
Student exercise
You have been asked to evaluate a community phone-in support programme that provides
counselling for people facing various crises. It has been funded by the government which is
interested to know how useful the programme has been and whether to continue financing it.
Which paradigm would you choose within which to undertake this research and why?
Please visit the companion website www.sagepub.co.uk/grbich2 for possible answers.
Further reading
Paradigms
Denzin, N. and Lincoln, Y. (eds) (2008) The Landscape of Qualitative Research: Theories and Issues (3rd edn).
Thousand Oaks, CA: Sage. Part II looks at competing epistemologies (positivist, postpositivist, constructivist, critical theory) as well as specific interpretive perspectives, feminisms, racial discourses, cultural studies,
sexualities, and queer theory.
INTRODUCTION
01-Grbich-Ch-01.indd 13
13
07/09/2012 3:39:36 PM
Crotty, M. (1998) The Foundations of Social Research: Meaning and Perspective in the Research Process. Sydney:
Allen and Unwin. A detailed and accessible discussion of positivism, constructivism, interpretivism, hermeneutics, feminism, critical inquiry and postmodernism is provided here.
Guba, E. and Lincoln, Y. (1994). Competing paradigms in qualitative research. In N.K. Denzin and Y.S. Lincoln
(eds), Handbook of Qualitative Research. Thousand Oaks, CA: Sage. The chapter by Guba and Lincoln examines various knowledge traditions that are relevant to qualitative research.
Mixed methods
Piano Clark, V. and Creswell, J. (2010) Designing and Conducting Mixed Methods Research. Thousand Oaks, CA:
Sage. A very accessible how-to text, which also discusses many of the complexities involved in combining
methods.
Tashakori, A. and Teddlie, C. (2010) Sage Handbook of Mixed Methods in Social and Behavioural Research.
Thousand Oaks, CA: Sage. The authors discuss paradigmatic issues, the strengths and weaknesses of mixed
methods designs, and provide specific examples as well as demonstrate how to teach and perform collaborative research using a mixed methods research design.
Evaluation
Goffman, E. (1974) Frame Analysis: An Essay on the Organisation of Experience. New York: Harper and Row.
Goffman’s book explores the concept of the framing of experience by individuals.
Kitto, S., Chesters, J. and Grbich, C. (2008) Quality in qualitative research: criteria for authors and assessors in
the submission and assessment of qualitative research articles for the Medical Journal of Australia. Medical
Journal of Australia, January 188 (4): 243–6.
MacLaren, G. and Reid, I. (1994) Framing and Interpretation. Melbourne: Melbourne University Press. McLaren
and Reid show how the different levels of framing influences can impact on researcher interpretation of
incidents and events.
References
Agar, M. (2004) We have met the other and we’re all nonlinear: ethnography as a nonlinear dynamic system.
Complexity, 10 (2): 16–24.
Freud, S. ([1900]1913) The Interpretation of Dreams. Trans. A. Brill. New York: Macmillan.
Grbich, C. (2004) New Approaches in Social Research. London: Sage.
Kitto, S., Chesters, J. and Grbich, C. (2008) Quality in qualitative research: criteria for authors and assessors in the
submission and assessment of qualitative research articles for the Medical Journal of Australia. Medical Journal
of Australia, January 188 (4): 243–6.
Marx, K. and Engels, F. ([1867]1999) Das Kapital. Washington, DC: Regnery Publishing.
14
01-Grbich-Ch-01.indd 14
GENERAL APPROACHES
07/09/2012 3:39:37 PM
TWO
Design methodologies, data management
and analytical approaches
In this chapter, the major analytic traditions of qualitative inquiry: iterative; subjective; investigative
and enumerative will be discussed and linked to current design methods. How to manage your data
and how to undertake preliminary data analysis will also be addressed in some detail.
KEY POINTS
• What constitutes data in qualitative research
• Design methodologies
• Traditions of inquiry and analytic approaches
° iterative
° subjective
° investigative
° enumerative
• How to match the analytical approach to the data type
• How to prepare data for analysis
° transcription
preliminary data analysis
°
What can constitute data in qualitative research?
The major data types lie in:
•• Face-to-face/telephone interview
•• Email and internet interviews
•• Focus groups (audio and video)
•• Nominal groups
•• Delphi groups
•• Observation notes, video, webcam, film, photos
•• Document collation (existing textual, aural
and visual)
Document collation can include information from newspapers, radios, TV, DVD, films,
videos, internet chat rooms, policy documents, clinical case histories, photos, drawings,
02-Grbich-Ch-02.indd 15
07/09/2012 3:40:01 PM
paintings, clothing, graffiti, books, emails and diaries: in short, any information that can
shed light on your research question. Be creative and do not be afraid to introduce new
data sources. The field is very flexible and finding the best answers for your research question should be your major priority.
Design methodologies
One of your tasks after you have determined your research question is to choose the
most useful methodology – ways of collecting and treating your data. There are many
methodological options to choose from or to combine in qualitative research designs.
Table 2.1 provides a listing of the most commonly used approaches. Each methodology in this table generally comprises guidelines attached to a theoretical underpinning. Your role is to select one or several methodologies that will best help answer
your research question. The first option, hermeneutic inquiry, provides the general
underpinning for all qualitative research, but it can also be used on its own. It comes
from Hans Gadamer’s Hermeneutic Circle (1975), which relates to the curious researcher
who goes out into the field with a question, gathers data in the best ways they can,
returns to make sense of his/her findings then returns to the field. This process is
repeated in a continuing spiral (iterative) until the research questions are satisfactorily
answered and the part answers gathered in the field have produced a creative and
meaningful whole.
The major methodological options (listed in Table 2.1) are not static, they provide a
feast of continual change.
Table 2.1   Common qualitative methodologies within specific paradigms
Realist/critical paradigm
Interpretivist/Constructivist
paradigm
Postmodern (pm)/
poststructuralist (ps) paradigm
Hermeneutic circle
(Gadamer)
Grounded theory (GT)
(Strauss)
Grounded theory
(Glaser)
Grounded theory
pm
Ethnography:
Classical/Critical
Critical
Ethnodrama/auto/cyber
Phenomenology
(classical/realist/
transcendental)
Transcendental/heuristic/
hermeneutic
pm/ps
Feminist research
(GT, Phenomenology)
Feminist research
(Memory work)
pm/ps
Evaluation
Summative
Formative
pm/ps
Action research
Participatory action
pm/ps
16
02-Grbich-Ch-02.indd 16
GENERAL APPROACHES
07/09/2012 3:40:01 PM
Traditions of inquiry and analytic approaches
Although certain methodological approaches favour particular analytic approaches, others call on different approaches depending on the research question and purpose. There
are four major traditions of inquiry in terms of analytic process: iterative, subjective, investigative and enumerative. Within each of these fall general design types, which can be
used flexibly; some design types occur in more than one tradition while combinations of
design approaches and traditions of inquiry can also occur in the same study.
The following is a brief introduction to these four broad types of qualitative analytic
approaches, the design of which will be described in greater detail in Chapters 4–20.
1. Iterative inquiry
Iterative/hermeneutic approaches involve seeking meaning and developing interpretive
explanations through processes of feedback. This involves a series of actions: defining the
question, going out to the field, examining the data collected, adjusting the question/sampling approach/design aspect/data collection tools in light of emerging issues and current
literature, subjecting this data to a critically reflective process of preliminary data analysis (see
later this chapter) to determine ‘what is going on’ in order to build up a picture of the data
and going back to the field to find out more. These processes are repeated until the accumulated findings indicate that nothing new is likely to emerge and that the research question has
been answered. There is recognition within this process that both you and those whom you
are researching construct meaning together, and that you will attempt to minimise both your
impact on the setting and possible over-interpretation of the situation, in favour of highlighting the views of those researched. Post data collection, thematic analysis often occurs (see
Chapter 5). This is a process whereby data is segregated, grouped, regrouped and re-linked in
order to consolidate meaning and explanation prior to display. Iterative approaches include
the basic hermeneutic approach as well as more defined approaches such as grounded theory,
phenomenology, ethnography, oral history, action, evaluation, sociocultural narratives, feminist versions of all of the above, and memory work.
2. Subjective inquiry
Subjective approaches are defined as those where there is a focus on you the researcher
and on what takes place within your own thoughts and actions in a specified context. The
focus is the collation of your own emotions and experiences. Here you will need to maintain a detailed and critically reflective diary record and be prepared to subject yourself to
regular periods of debriefing with a colleague or supervisor. When your experiences are
the sole or partial target of the research, you occupy a dual role – that of researcher and
researched. Preliminary data analysis is again a key analytic technique, with thematic
analysis being a further option depending on how much decontextualising and segmenting you regard as appropriate or desirable. Subjective approaches include: autoethnography, heuristic phenomenology and some postmodern versions of ethnography, grounded
theory, feminist, evaluation and action research, where the researcher has chosen to
include a significant segment of subjective data.
DESIGN METHODOLOGIES, DATA MANAGEMENT AND ANALYTICAL APPROACHES
02-Grbich-Ch-02.indd 17
17
07/09/2012 3:40:01 PM
3. Investigative (semiotic) inquiry
Investigative semiotic approaches involve the uncovering of information relating to languages within cultural contexts. The understanding of signs and symbols is central to
this approach, in particular their mythical strength and the embedded power of particular discourses which you will need to disentangle to reveal the original elements as well
as to identify arguments that have been marginalised. There is considerable variety
amongst the continuum of possibilities relating to analysis of documentation, visuals
and body language, varying from the looser ethnographic content analysis approach,
which attempts to contextualise the document and to identify and describe the values
and attitudes evident, to the precision of some forms of discourse and semiotic analysis,
which focus on grammatical structure. But again, the flexibility and the ever-changing
nature of qualitative approaches allow you considerable variation, especially when you
provide adequate justification. Investigative approaches include: structuralist, poststructuralist, content analysis, feminist research, as well as discourse analysis, conversational
analysis and narratives of the sociolinguistic type.
4. Enumerative inquiry
This involves the listing or classifying of items by percentages, frequencies, ranked order, or
whatever is useful to the research question. These approaches involve you in the production
of ‘objective’ accounts of the content of verbal, written, or visual texts, the development of
codes and categories often prior to analysis, and the definition and measurement of units of
analysis. Flow charts, logical reasoning processes, the seeking of links between antecedents
and outcomes through identification of ordered (ranked) word frequency, key words in
context and incidence counting. The development of previously decided codes can also be
seen in the imposition of ‘matrixes’ (conceptual frames of interlinking variables from which
propositions with causal implications have been derived) where you apply this to one case
and then further apply it to other cases to develop cross-case analysis. There is an underlying assumption in this process that fully predesigned instrumentation will enhance validity and generalisability. Enumerative approaches are often questioned in qualitative research
because of their tendency to atomise and decontextualise the data and the fact that connection does not necessarily equal causation. Approaches include transcendental realism,
quasi-statistical and matrix analyses.
These four traditions can be summarised as shown in Table 2.2.
Which analytical approach for which data type?
One of the more difficult areas to grasp is which type of analysis to use with which type
of data. The key here is to look at the type of data and be prepared to justify what you are
doing and why. However, there are guidelines. If you choose a particular method, such
as those in the iterative/subjective columns in Table 2.2 it is anticipated that you will be
collecting data mostly through interview/observation of others or the self and therefore
18
02-Grbich-Ch-02.indd 18
GENERAL APPROACHES
07/09/2012 3:40:01 PM
Table 2.2   Traditions of design and analytic inquiry
Iterative (hermeneutic)
Subjective
Investigative (semiotic)
Enumerative
Grounded theory
Autoethnographic
Structural
Quasi-statistical
Phenomenology
Classical/
transcendental
Heuristic
phenomenology
Poststructural
Transcendental
realism
Ethnography
Postmodern versions of
iterative approaches
Discourse analysis
Matrix analysis
Oral history
Content analysis
Action evaluation
Conversation analysis
Feminist research
and memory work
Narratives –
sociocultural
Feminist research
Narratives –
sociolinguistic
the dynamics of interaction (including your contribution and influence) will impact in
questions, sampling, attitude and the nuances of verbal and non-verbal communication.
In this situation, you are expected to recognise your contribution AND you are also
expected to avoid imposing a frame/matrix/set of codes (see Chapter 21) on your data
because you want the data to speak for itself initially – so, do not force the data into
predefined categories/codes/themes. The analytic approaches you would be expected to
consider here would be preliminary data analysis (see below) and thematic analysis (see
Chapter 5).
If you are dealing with existing documentation (see Part 4) as in the investigative and
enumerative approaches (above) then you can use any approach that helps you make
sense of the data, including preliminary data analysis and thematic analysis. So, I hear
you ask, if you can use any reasonable approach with already existing documentation,
why can’t you, for example, use content analysis or any other form of enumeration with
your self-collected interviews? Well, there is nothing to stop you running these interviews through a content analysis computer package in order to discover the frequency of
occurrence of particular words and even the use of these words within a limited defined
context (several words on either side of the key word). But, this will not give you properly contextualised themes, and if this is all you do, you will be missing out on achieving
the gold standard of qualitative research, which is the detailed analysis and presentation
of rich in-depth information via emerging rather than imposed themes.
Summary
From what we have looked at so far in this chapter, you can see that there are a number of
signposts along the research path which require your attention as a qualitative researcher
before you move into actually conducting data analysis:
DESIGN METHODOLOGIES, DATA MANAGEMENT AND ANALYTICAL APPROACHES
02-Grbich-Ch-02.indd 19
19
07/09/2012 3:40:02 PM
•• making decisions regarding your research paradigm/s
•• seeking the methodology/ies that will best answer your research question/s and the data sources
that will provide you with the best information
•• choosing the best analytic tradition to use for your completed data set.
Let us assume you have collected some interview/observational data. Now what?
How to prepare interview/observational
data for analysis
There are two stages needed here:
•• transcription of your data so you can undertake the next stage
•• preliminary data analysis.
Transcription
Transcription involves getting the dialogue or narrative off the devices on which you
have recorded it and into a document formatted so there is a clear researcher-defined
column for notes, as seen below.
Transcription of interview
Researcher’s notes
Q. Tony, now that you have been a male primary caregiver for a year
how do you think others view you in this role?
A. It’s the perception of the division between male and female roles
that I find a key to most of the injustice I’ve encountered during this
past year. When it comes to the crunch, most males prefer to be
breadwinners; they see this as the more important role. This has
also to do with money and title and status.
I remember when my wife graduated from Medical School,
amongst the group she was part of, what struck me as odd even
then was that the men all looked forward to their futures as doctors
and that’s natural enough, but there was never any question that
they wouldn’t work full time. Several were in stable relationships
and their partners had started careers of their own. But none of the
graduates male or female ever questioned who would stay home
should they start a family. There was no question of whose career
was expendable. As soon as they had children, both the wives of
doctors and the female medical graduates would all give up work
and retire to the house.
Source: Grbich, data set: Primary caregiver males
20
02-Grbich-Ch-02.indd 20
GENERAL APPROACHES
07/09/2012 3:40:02 PM
Preliminary data analysis
For many of the methodological qualitative designs (see Table 2.1) the initial analytic stage
involves preliminary data analysis. Preliminary data analysis is an ongoing process that is
undertaken every time data is collected. It involves a simple process of checking and tracking
the data to see what is coming out of it, identifying areas that require follow up and actively
questioning where the information collected is leading/should lead you, the researcher. It is
a process of engagement with the text, not so much to critique it or to summarise what is
emerging from it, but more to gain a deeper understanding of the values and meanings
which lie therein.
Regardless of whether the data collected comes from written observations, transcriptions of interviews or the perusal of existing documents, you should undertake this process in order to highlight emerging issues, to allow all relevant data to be identified and
to provide directions for seeking further data.
The process with regard to interview data is demonstrated below.
Interview segment of a transcript
analysed using preliminary data analysis
Interview segment
Q. Tony, now that you have been a male primary
caregiver for a year how do you think others view
you in this role?
A. Its the perception of the division between male
and female roles that I find a key to most of the injustice I’ve encountered during this past year. When it
comes to the crunch, most males prefer to be breadwinners; they see this as the more important role.
This has also to do with money and title and status.
I remember when my wife graduated from Medical
School, amongst the group she was part of what struck
me as odd even then was that the men all looked
forward to their futures as doctors and that’s natural
enough, but there was never any question that they
wouldn’t work full time. Several were in stable relationships and their partners had started careers of
their own. But none of the graduates male or female
ever questioned who would stay home should they
start a family. There was no question of whose career
was expendable. As soon as they had children, both
the wives of doctors and the female medical graduates would all give up work and retire to the house.
Preliminary data analysis
Injustice – what injustices has he
experienced? *I need to check
this aspect with him and with
other participants.
Do all men prefer to be breadwinners? *I need to check this
with the group.
This fits in with the societal view
of the nurturing role of women.
It also matches with Talcott
Parsons’s views of instrumental
(male) and nurturing (female)
societal roles.
Did Tony have different earlier
socialisation experiences than
the others in order to want to
be a father at home? *I need
to follow up with him and the
others about this
*Note to self
DESIGN METHODOLOGIES, DATA MANAGEMENT AND ANALYTICAL APPROACHES
02-Grbich-Ch-02.indd 21
21
07/09/2012 3:40:02 PM
There is some diversity within the literature as to how this process of preliminary data analysis might occur, but given that it is idiosyncratic, each researcher must decide what works
for him or her. Examples of what other researchers do can be found in the book by Ian Dey
(1993), who identified the techniques for early interactive reading of data segments as:
•• free association – writing freely regarding words, phrases and topics in order to avoid and
release fixed researcher assumptions;
•• comparing interviews with own experiences; identifying aspects of the research map – the self, the
situated activities, emergent meanings, understandings and definitions, identification of aspects
impacting on the contextual setting as well as documenting interactions, history, events, strategies, process and consequences;
•• shifting the focus among the levels of data to highlight other areas; reading the data in different
sequences; critiquing the data – Who? What? Why? When? So what?
•• transposing the data by asking ‘What if?’ in order to seek new perspectives.
Michele Bellavita (1997: 181) has a similar but looser approach. She allows herself to:
•• go over the data segment initially, noting ideas and then trying to create names for chunks of
data – listing topics, grouping them, noting exceptions and brainstorming;
•• play with metaphors, analysing specific words and employing the ‘flip flop’ technique (looking
at aspects from different perspectives, asking ‘Why?’ and ‘What if?’)
•• attempt to re-present some of the data in the form of a poem or vignette which may form part
of a later display of the overall database.
Face sheets
Each data set that has been transcribed and has undergone preliminary data analysis
then requires some form of identification and of summary of this process, such as a face
sheet. This is a cover sheet that is attached to the front of your data transcription and
identifies the study question, time and place of interview/observation and summarises
the main outcomes from your preliminary analysis. The face sheet from the above interview segment would then look like this:
Face sheet: interview with “Tony”
data identifiers
Research question: How do men experience the role of primary caregiver?
••
••
••
••
••
Participant profile: age, status: “Tony” 42 years
Interview/observation/document date: 24/04/07; Interview 2
Time: 10am – 12.45pm
Place of interview: Tony’s home
Comments: Daughter Chrissy, 18 months, was present for the first hour before being put to bed
22
02-Grbich-Ch-02.indd 22
GENERAL APPROACHES
07/09/2012 3:40:02 PM
Issues emerging from the interview that need to be followed up by the researcher:
••
••
••
••
••
••
Do all these men experience a sense of injustice?
Does working part-time make a difference to perceptions of ‘injustice’?
Do all men prefer to be breadwinners?
Do all women prefer the ‘at home’ role?
Do some/all these men prefer the home role to the breadwinner role?
Investigate the early socialisation of these men.
Summaries of issues emerging
You will find it helpful during data collection to start accumulating emerging issues into
potential themes. You do this by summarising supportive data for a particular aspect
every 3–5 sessions of interviewing or observation. The advantage of this is that by the
end of data collection you will not only have completed the twin processes of preliminary data analysis and judicial summaries but you will be in a position to start interpreting and conceptualising your data.
Following the chapter on mixed methods, the group of Chapters 4–8 will investigate the iterative orientation with regard to specific methodologies that can be used
within any of the realist/ critical/ interpretivist/constructivist/ postmodern/poststructural
paradigms: namely ethnography, grounded theory, phenomenology and feminist
approaches. The analytical tools most pertinent to the postmodern/poststructural traditions will be dealt with specifically in Chapters 9–13. The data dealt with in Chapters
4–13 will be predominantly gathered by the researcher through interview and observation, while Chapters 14–20 will deal with the analysis of documents and other data
already in existence.
Student exercise: Preliminary data analysis
Here is a segment of an interview response to the question: ‘What are your views on
homelessness?’
Well I think it’s a terrible thing that in a modern society we have a problem such as this. I don’t
know though what we can do about it. I want to do something but I don’t know what. You are
constantly reminded of the problem if you go to town. They are always there asking for money. You
don’t know though if they really need it or if they are in fact doing quite well from begging. But if
you have any caring qualities you can’t ignore requests for help, when you have so much yourself.
They say a lot of people choose to live on the streets – well I don’t believe that. I think that is an
easy thing for the authorities to say – it lets them off the hook. So really I guess the answer to your
question is that I am not sure what to think about homelessness. I feel that something should be
done but I feel powerless to do anything personally. I think that the time has come for the
government to do something. It makes me feel very uncomfortable; I don’t know what to do.
Please visit the companion website www.sagepub.co.uk/grbich2 for possible answers.
DESIGN METHODOLOGIES, DATA MANAGEMENT AND ANALYTICAL APPROACHES
02-Grbich-Ch-02.indd 23
23
07/09/2012 3:40:02 PM
Further reading
Qualitative methods
Hennink, M., Hutter, I. and Bailey, A (2011) Qualitative Research Methods. London: Sage. An accessible introduction, especially to the iterative approach.
Design and analysis
Denzin, N. and Lincoln, Y. (2011) The Sage Handbook of Qualitative Research. London: Sage. A good overall
introduction to qualitative methods.
Grbich, C. (2009) Qualitative Research in Health: An Introduction. London: Sage. http://books.google.com/bo
oks?id=MeMB9wp0p5sC&pg=PA26&source=gbs_toc_r&cad=4#v=onepage&q&f=false (accessed 1 August
2011). A simple introduction to qualitative research.
References
Bellavita, M. (1997). In M. Ely, R. Vinz, M. Downing and M. Anzul (eds), On Writing Qualitative Research: Living
by Words. London: FalmerRoutledge.
Dey, I. (1993) Qualitative Data Analysis: A User-friendly Guide for Social Scientists. London: Routledge.
Gadamer, H. (1975) Hermeneutic circle. Philosophy & Social Criticism/Cultural Hermeneutics, 2 (4): 307–16.
24
02-Grbich-Ch-02.indd 24
GENERAL APPROACHES
07/09/2012 3:40:02 PM
THREE
Incorporating data from
multiple sources: mixing methods
The mix of qualitative and quantitative data in terms of design, management and presentation will
be dealt with in this chapter. The advantage of utilising an innovative mix of data sources is that
apart from providing a broader view of the research question, the combination of data allows the
reader to view the phenomenon under study from different perspectives.
KEY POINTS
•• The advantages of combining quantitative and qualitative data are that you can maximise
the impact of both
•• For mixed methods to be successful, issues of sampling, design, data analysis and data
presentation need careful attention
•• Two approaches to mixing methods regarding design are concurrent and sequential, but
other new mixes are emerging
•• Does qualitative data miss out in such a mix? And what is the next move?
A brief history of qualitative and
quantitative approaches
If we go back in time there appear to be three stages of debate relating to quantitative and
qualitative approaches:
1. ‘Never the twain shall meet’
Up to the mid-1980s the two approaches were polarised. They were seen as deriving from
completely different theoretical underpinnings with very different methods and quite separate orientations; one toward objectivity and the other toward subjectivity. The outcome of
this was that people argued fiercely for the superior capacities of one camp or the other and
combinations were almost never to be seen.
03-Grbich-Ch-03.indd 25
07/09/2012 3:40:15 PM
2. Rapprochement
During the late 1970s philosophical and other differences gave way to a closer examination of possible complementarities between the two approaches and a realisation that
although different they could well enhance each other, particularly when run in parallel.
This led to a rash of studies within the quantitative tradition where qualitative methods
were used as an ‘end-on’ approach. That is, before developing survey questions a qualitative data set would be collected to identify the ‘right’ questions, or, after running a survey,
some qualitative interviews would be carried out to clarify and expand on ‘the reasons
behind’ some of the statistical results gained. The dangers here lay in oversimplifying the
complexities of data analysis and the rigors of theoretical interpretation in the qualitative
data set, in favour of a shallow identification of ‘emerging issues’ or ‘recurring themes’.
3. Cooperation and mixing
During the past 25 years as the limitations of both quantitative (in terms of detail) and
qualitative designs (in terms of numbers) were recognised and designs to incorporate both
equally or as part of a range of mixed method approaches emerged – these ranged from a
grab bag of supermarket-type selection of approaches to formal integration of quantitative
and qualitative methods. Paradigmatically, postpositivists challenged positivists, and constructivists and critical theorists leapt into the ring, and when the dust settled it was agreed
that different methods could be combined and in fact could even drive research under an
encompassing but loose paradigm such as pragmatism or transformative action. Researchers
from different fields seeking grant monies found it useful to develop a mixed methods
design to enhance the breadth of exploration in a particular project.
What are the major differences between
qualitative and quantitative?
Qualitative tends to be seen primarily as an inductive approach using a research question
to move from instances gained in the data collection to some form of conclusion, often
via comparison with existing concepts or theory. Questions tend to be exploratory and
open-ended and data is often in narrative form. Reality is seen as a shifting feast, subjectivity is usually viewed as important and power is shared with the participants who are
the experts on the matter under investigation. Analysis predominantly deals with meanings, descriptions, values and characteristics of people and things. The outcome sought
is the development of explanatory concepts and models: appropriately theoretically
underpinned, uniqueness is favoured and widespread generalisation (apart from logical
generalisation – that is from similar instance to similar instance) is avoided.
In contrast quantitative is generally viewed as deductive, where the conclusions drawn
follow logically from certain premises – usually rule based – which are themselves often
viewed as proven, valid or ‘true’. Reality is seen as static and measurable, objectivity (distance, neutrality) is important, linearity (cause–effect) may be sought, outcomes are the
major focus and pre-specified/developed hypotheses will dictate questions and approach.
Researcher control of the total process is paramount, precision and predictability are important and statistical approaches to identify numbers and to clarify relationships between
26
03-Grbich-Ch-03.indd 26
GENERAL APPROACHES
07/09/2012 3:40:15 PM
variables will dominate data analysis. Theory testing is the key and generalisation and
predictability are the desired outcomes, survey and experimental research being the
main design options.
Advantages of combining quantitative
and qualitative results
The advantages of a combined approach are:
•• clarifying and answering more questions from different perspectives
•• enhancing the validity of your findings
•• increasing the capacity to cross-check one data set against another.
Other positive outcomes include:
•• Providing the detail of individual experiences behind the statistics – the intensive/extensive
view. So if, for example, 99% of survey respondents indicate that they are extremely dissatisfied
with the resources offered to them by the government to support their role of carer then qualitative interviews can identify in detail what aspects of support the 1 per cent were satisfied with,
as well as documenting the experiences of dissatisfaction of the remainder.
•• Helping in the development of particular measures – to clarify in detail through the results of
qualitative data what the major issues are and then to use the analysis of these responses to
refine a survey questionnaire in order to gain more focused results from a larger sample. For
example, in the restructuring of a department it would be sensible to find out from all members
what they considered to be the positive aspects of the current organisation and what were
considered to be barriers to productivity. This detailed information could be used to structure a
questionnaire to suggest specific changes which might then be rated for feasibility.
•• Tracking stages over time – a multi-staged approach can clarify, from different perspectives,
how a situation is progressing. For example: an initial qualitative interview might pick up on
the range of stressors faced as couples with children separate, following breakdown of relationships. Three months later an intervention might provide clarification of the long-term roles and
responsibilities that each parent needs to take into account regarding their children’s care. As
well, the financial and emotional costs for each couple can then be calculated at a time which
is less stressful than the initial separation period. At six months post separation, a survey could
explore the current situation; for those who indicated that things were not working out as
planned, further in-depth interviews could tease out their problems and seek solutions.
Philosophical integration of qualitative
and quantitative approaches
There are two current options here:
•• Pragmatism (Rossman and Wilson, 1985), whose origins lie in the works of Charles Peirce, William
James and John Dewey. Pragmatism seeks ways through the polarised quantitative–qualitative
debate to find practical solutions to the problem of differing ideologies and methodologies. The
current movement in pragmatism rejects former polarised positions, focusing instead on the best
way to answer a particular research question but with recognition of culture, context, individual
INCORPORATING DATA FROM MULTIPLE SOURCES: MIXING METHODS
03-Grbich-Ch-03.indd 27
27
07/09/2012 3:40:15 PM
experience, the constructed nature of reality, uncertainty, eclecticism, pluralism and the need
for creative innovation of method. As well, pragmatism encompasses an instrumental view of
theory which endorses the values of praxis, democracy, freedom, equality and progress, but also
accepts that bifurcations, infinite loops and provisional truths can emerge within methodological
design, analysis and interpretation (Burke Johnson and Onwuegbuzie, 2004). The advantages
of pragmatism are that it places a major focus on the research question and that the methods
become simply the best tools for providing the most comprehensive answer to the question; thus
methods can become hybrid and creative to achieve this goal. The disadvantages are that some
topics do not benefit from ‘pragmatic’ solutions and a focus on applied ‘band aid’-style research
outcomes with short-term solutions may benefit only a few participants.
•• The transformative paradigm (Mertens, 2010), where the view is that:
° multiple realities are shaped by cultural, social, economic and political influences
° knowledge is historically and socially situated
° issues of power between researcher and researched need to be explicitly addressed.
The transformative ethical orientation comprises a strong human rights agenda within notions
of beneficence and social justice – with the implication that all research must be for the betterment of humankind. Methods include multiple approaches, multiple methods, techniques and
theories, which should provide the basis for action and social change.
Conducting mixed methods research: prior questions
Is your research question one for which mixed methods would be the best approach? If
so, which design would be the best?
•• A mutual research design (Armitage, 2007), involving acceptance that the two approaches come
from completely different paradigms, celebrating their differences and keeping them separate
within the design process – the ‘separate but together’ position? Or
•• A mixed methods design? And if so:
° At which points will mixing occur – design? analysis? interpretation?
° What sampling approaches will you utilise from the probability and non-probability suite?
° How are you going to manage data analysis? Here you can consider quantitising – converting
°
°
qualitative data into quantitative data or qualititising – converting quantitative data into
qualitative data.
To what degree will you qualitatively analyse quantitative data and vice versa?
How are you going to display your results – separately? integrated? consolidated?
Mixed method design
Various forms of labelling and terminology have been used for mixed method design: synergy,
integration, triangulation, concurrent, parallel, merging, sequential, exploratory and explanatory.
In terms of design, however, this has come down to two broad options: collecting quantitative
and qualitative data at the same time, and intermingling sampling and analytical approaches
to whatever degree is useful (concurrent); and collecting qualitative and quantitative data
at different times for knowledge expansion while maintaining paradigmatically separate
approaches to whatever degree is seen as useful (sequential). A third option, that of mixing
concurrent and sequential or mixing named approaches, is also emerging.
28
03-Grbich-Ch-03.indd 28
GENERAL APPROACHES
07/09/2012 3:40:15 PM
1. Concurrent or parallel methods
Here you would consider using multiple reference points, where separate data sets are collected at the same time with the ultimate aim of merging the two data sets, either in a
visual display such as a matrix by transforming the data (see quantitising and qualititising
data in crossover/mixed analyses below) or in the final discussion. Designs might involve
using dual sites with the same sampling approach but with different data (quantitative
and qualitative) then using the synthesised results to build up a complex picture.
Research example A concurrent approach can be seen in the study undertaken by
Vicki Piano Clark et al. (2010) that examined how the alumni of a Graduate Fellows
group, who worked in elementary/middle schools in their districts for 10 hours per
week helping to address teachers’ needs, perceived the impact of their participation.
Data collection was via an online questionnaire with open- and close-ended questions
(concurrent qualitative and quantitative data). Data was analysed separately: quantitative with SPSS for descriptive and inferential statistics and qualitative using open coding
in MAXQDA (see Chapters 21 and 22). Three merging strategies were trialled in order to
develop more precise strategies for researchers:
•• merging via data transformation (using qualitative findings to develop a quantitative variable and
comparing relationships with other variables and developing better variables from combining
qualitative and quantitative findings)
•• merging with a matrix (quantifying difference in qualitative findings, examining these, perusing
differences in the quantitative results using a qualitative typology)
•• final discussion merging (corroborating findings, developing a more complete picture, establishing
divergence).
The researchers discovered that different merging approaches suited different research
questions but that data transformation merged at data level, matrix at results level, and
discussion at the interpretive level.
2. Sequential: explanatory/exploratory
You could undertake a qualitative study to explore a particular issue or phenomenon and
using an iterative approach you could create hypotheses from these results which you could
test using a survey or experimental design. Or, you could develop a short questionnaire
survey to elicit key issues that can then be explained in depth using qualitative approaches
of interviewing and observation. Synthesis of the two sets of results is needed to clarify the
dual outcomes and to utilise the increased validity these two approaches provide.
You could also use the sequencing approach at a lower level within questionnaires to
provide base data to enable movement into other data sets. For example, using both
focused (limited response) quantitative and open-ended (qualitative) questions in a survey permits a more holistic view of the questions to be addressed and allows you to follow
up responses:
e.g. Have you received any services from the government?
Yes/No
If Yes, could you name each service and describe your experiences of this service.
INCORPORATING DATA FROM MULTIPLE SOURCES: MIXING METHODS
03-Grbich-Ch-03.indd 29
29
07/09/2012 3:40:15 PM
These descriptions can then be fleshed out further in a follow up face-to-face or telephone interview, the results of which can then be re-incorporated in a more focused
follow up survey.
A typical sequencing design could involve:
Stage 1: Representative survey of the population.
Stage 2: Exploratory qualitative interviews or focus groups to tease out the findings of the survey.
Stage 3: Hypotheses generated from stages 1 and 2 are tested in various interventions which are
then evaluated.
Stage 4: Participatory action research where the participants take control of the development,
implementation and evaluation of the most successful of these interventions.
Research example: quant–qual and qual–quant-qual Peter Davies and Bob Baulch
(2011) used a sequential approach to explore poverty dynamics in rural Bangladesh in
a longitudinal study. Already in existence was a rich set of historical data evaluating the
short-term impact of microfinance projects, new agricultural technologies and education
transfers. This data comprised a number of repeated quantitative surveys and qualitative
focus groups and semi-structured interview data. This data was then further extended:
•• the first phase involved collecting more qualitative focus group data to examine perceptions of
change
•• the second phase undertook a quantitative survey updating the historical surveys
•• the third phase collected the qualitative life histories of 293 individuals selected from and nested
within the quantitative sample to explore poverty transitions.
Comparing data analyses of the different data sets showed more transitions out of poverty
in the quantitative than the life history data. This led to a number of queries regarding:
•• per capita household expenditure not accurately reflecting economic wealth
•• failure of expenditure base transition matrices to indicate how close to the poverty line expenditure
actually was
•• the cost of illnesses not being reflected in expenditure-based assessment
•• changes in household size not being picked up as a factor in moving out of poverty
•• recall errors occurring in qualitative data as people tried to remember past events.
In integrating the data, these data mismatches between qualitative and quantitative data
were tabled, summarised and sequentially reduced via the application of various explanations. One set of mismatched data is shown in Figure 3.1.
Thus in using mixed data, both sets served to highlight the deficiencies of the other
and to clarify weaknesses in poverty assessments.
Research example: qual–quant Jeffrey Edmeades et al. (2010) used a sequential
approach to identify the circumstances surrounding abortion in Madhya Pradesh, India.
This study had two phases. The first involved the gathering of an initial qualitative data set
(10 focus groups plus 17 key informant interviews plus 41 in-depth interviews using a ‘storytelling’ technique) to identify key social and cultural factors in the advent, experiences and
resolution of unwanted pregnancies. The narrative patterns that emerged from analysis of
this data were adhered to and used to shape the questions of the second stage, a subsequent
30
03-Grbich-Ch-03.indd 30
GENERAL APPROACHES
07/09/2012 3:40:15 PM
bom
1937
father died
1945
brothers separated
married son
households after a
daughter
separated and
quarrel 1965
married
3 cows moved away
load of rice lost
1987
poisoned by
2003
when boat
nephew
sister married
sank in cyclone
own 2003
1964
tk. 70,000 loss
illness
1971
2002
3
3
son married
son working son married 1998
in flour mill
1994
trading rice
1979-1988
cultivating own land
1968-1971
wife took
1947-1967
loans from
own work
Grameen
cultivating crops
Bank, used to
and dealing in
cattle 1982-2007 buy land, from
2002
2
2
1930
1940
1950
1960
1970
1980
1990
Per capita expenditure
Poverty line (BBS)
Land owned (decimals)
Household members
1996
243
618
216
9
2007
1492
877
93
2
2000
2007
Figure 3.1   Mismatch caused by diseconomies of scale (70-year-old man) (from Davies and
Baulch, 2011, reproduced with permission)
survey which was then more fluid and conversational than is usual. This sequential
structure facilitated gathering information both on individual women and on individual
pregnancies, producing high-quality survey data that countered the problem of previous
research – that of the under-reporting of abortion providing a more in-depth understanding
of these women’s motivation for abortion and their problems with societal pressures and
access to services, than had previously been available to researchers.
3. Combining concurrent and sequential
approaches and other mixes
Concurrent plus sequential
Formally putting these two approaches together is a recent innovation, although many researchers will have previously collected concurrent and
sequential data within either the qualitative or quantitative traditions or will have moved
sequentially among various concurrent data sets in developing and merging analytical
procedures (Piano Clark et al., 2010). Donald Nicolson et al. (2011) examined the usability and readability of five patient-focused medication information websites that provided
information on the same pharmaceutical drug. They combined concurrent data, which
involved sitting with each of the 15 participants and observing their use and understanding
of the sites. Participants were encouraged to ‘think aloud’ and Macromedia captivate – a
computer program – recorded voice and monitored online activities. Sequential data
INCORPORATING DATA FROM MULTIPLE SOURCES: MIXING METHODS
03-Grbich-Ch-03.indd 31
31
07/09/2012 3:40:16 PM
(in the form of a post test semi-structured interview) gained participants’ views on site
accessibility immediately after usage. The researchers found analysing four sets of concurrent data to be demanding of time and energy, but the detailed data, although from
only a small group of participants, provided sufficient information to develop evidencebased guidelines for the design and content of the five sites.
Mixing named brands
Some researchers have mixed specific rather than generic
approaches. Elizabeth and Andrea Quinlan (2010) put together institutional ethnography and social network analysis to link one personal experience of rape with institutional
and textual representations . Data comprised a medicolegal text (sexual assault evidence
kit) and one woman’s narrative experience – a well-publicised rape case in Canada that led
to the writing of a book, The Story of Jane Doe (Doe, 2003). From thematic analysis of the
narrative and the kit, a map was created using the software program Visualyzer in order to
quantitise and simplify the qualitative data and display it via nodes to illustrate relationships between variables and between themes. This allowed the two views of rape to be
mapped for comparison exposing the internal structures of both: personal experience is
portrayed weblike above the line and the sexual assault evidence kit (hierarchical with no
connection between themes) is shown below the line of fault (see Figure 3.2).
Breathing
Physicality
Nightly Doings
Terror
Sense of Safety
Commit to Tell
Desire to Live
His Action
Desire to Die
Memorizing
Forensic Evidence
Assault
His Physical Body
Her Identity
Social Network
Clothing
Marital Status
Name
Address/Tele #
Date of Birth
Assault Date
Medical Professional Involved
History
Sexual
Medical
Attending Nurse
Injuries
Next of Kin
Social Worker
DNA (forensic)
Assault Location
Examining Physician
Assault Time
Figure 3.2   The line of fault (from Quinlan and Quinlan, 2010, reproduced with permission)
Issues to consider in attempting
to combine data sets
•• You need to be familiar with both quantitative and qualitative approaches.
•• Mixing of paradigms, data collection, analysis and interpretation takes time and skills to do well.
•• Combined designs are more expensive than single designs.
•• Are there benefits to converting qualitative to quantitative data?
The main danger is that the time-consuming traditional methods of qualitative analysis
may be bypassed or glossed over in favour of quicker but shallower approaches. This would
32
03-Grbich-Ch-03.indd 32
GENERAL APPROACHES
07/09/2012 3:40:16 PM
lead to oversimplified results and to researcher-directed rather than data-derived results.
Both the qualitative and quantitative data sets must be properly designed, collected and
analysed. The quantitative data needs to be systematic with clearly defined variables and
with a good sampling strategy to enable reliable predictions, while the qualitative data
needs clarification of process, context and culture as well as detailed and rich data properly
collected, analysed and interpreted using theory that serves to further illuminate results.
Questionnaires
The order of question type and the numbers of questionnaires may be important. Too
much or too many may sensitise participants. In an attempt to ascertain the essential
personal qualities of a computer educator, a three-phase sequential approach was used by
Pieterse and Sonnekus (2003).
•• First, a survey questionnaire was used to incorporate both quantitative and qualitative questions
to build up a picture from a sample taken from one location.
•• Then, during the second phase the qualitative aspects were integrated into the quantitative
items to present a generic picture gained from the first phase and this was then presented to a
national sample.
•• In the third phase, quantitative findings were integrated into the qualitative questions in an
online questionnaire for an international sample of IT Training Groups but with the additional
option of a two-way email contact.
These researchers found that the difficulty in providing both qualitative and quantitative
questions in the same survey, and in particular putting the closed questions earlier, was
that the qualitative information did not appear to add very much in terms of new information, suggesting that the participants had become sensitised to the content and orientation of the earlier questions.
Design
One issue is that of equality in the weighting of the data – is the study more qualitatively
or more quantitatively focused such that the second approach operates more as an addon, or minor data set? And is this important in the answering of the research question?
And what happens when one set of results contradicts the other? Can these differences
be resolved by ‘merging’ and ‘integrating’ and if they can does this create more ‘truth’ or
just more cleverly manipulated data?
The issues surrounding the definition and operationalisation of the term ‘triangulation’
have been debated for the past 50 years and are still completely unresolved, so the
re-imposition of this term in mixed methods does not paper the cracks nor does it give automatic validity to mixed methods (Denzin, 2010). Validity of findings has also been an issue
and the development of the Instrument Development and Construct Validation (IDCV)
tool (Onwuegbuzie et al., 2010) may help here. This is a 10-phase tool covering an interdisciplinary literature review, instrument development and evaluation, and construct validation process and product. Another tool, the Validity Framework (VF) (Leech et al., 2010),
provides a five-element approach covering foundational elements and construct validity in
qualitative, quantitative, mixed, inferential, historical and consequential data elements.
INCORPORATING DATA FROM MULTIPLE SOURCES: MIXING METHODS
03-Grbich-Ch-03.indd 33
33
07/09/2012 3:40:16 PM
Crossover/mixed analysis
Crossover analysis involves the use of different lens’ on qualitative and quantitative data
sets using intersubjectivity, constructivist and postpositivist views, mixing ‘objective’
with ‘multiple realities’ perspectives and switching tools – for example using factor analysis on themes derived from qualitative data or a thematic analysis on themes arising from
factor analysis, leading to sequential/concurrent mixed analytic approaches in order to
expand perspective (Onwuegbuzie et al., 2010) and shifting cases to variables within a
three-dimensional framework – case, variable and process/experience (Onwuegbuzie et al.,
2007, 2009, 2010).
These authors suggest the following techniques for crossover analysis:
•• reducing dimensionality of either data set (quantifying to basics)
•• integrating data display (visual presentation of both sets as one)
•• transforming data (qual to quant [numerical codes]) and quant to qual [themes]) for analysis
•• correlating data (correlating results from quantitising and qualititising)
•• consolidating data (merging multiple data sets to create new codes, variables etc.)
•• comparing data (comparing findings)
•• integrating data (integrating into one or two sets of data)
•• using warranted assertion analysis (seeking meta-inferences from both sets)
•• importing data (using follow up findings from qualitative to inform quantitative analysis and
vice versa). (adapted from Onwuegbuzie et al., 2010: 58-9)
The main problem that arises here is that sampling and data collection have been done
with a particular analytic approach in mind and therefore crossing over may be problematic in terms of usefulness. Thick data may (or may not) lend itself to variable analysis and
thin closed-ended data may not be productive in terms of thick contextual description –
the quantity and depth of information are simply not there.
Interpretation of results
Another issue is that of ‘transformation’, where the combining or homogenising of data
requires a theoretical interpretation that includes both data sets when only one (qualitative) is usually treated in this manner – the other being viewed as ‘fact’ reflecting ‘reality’
to prove/disprove a hypothesis. The tendency has been for a low level conceptual interpretation to be undertaken but moves to include multiple frameworks can be seen
(Nicolson et al., 2011).
Presentation of dual results
Separate data sets
Is it best to display each data set separately? The difficulties surrounding presentation of
qualitative and quantitative approaches lie in their integration, which leads to a very
large results section and requires regular summaries of data findings which will need to
culminate in a final drawing together of the findings so that the reader can make sense
of the diversity presented. An example of this is provided by Duncan and Edwards
34
03-Grbich-Ch-03.indd 34
GENERAL APPROACHES
07/09/2012 3:40:16 PM
(1997), who interviewed 95 lone mothers in three countries (UK, USA and Sweden) and
contextualised these interviews within census data. The results are largely displayed in
separate chapters (for example, a chapter on social negotiation of understandings followed by one on census data).
Combined data sets
In contrast to the separation of data sets, is it preferable to amalgamate the findings in
such a way that a neat display of graphical information occurs, followed by a few carefully
chosen qualitative quotes that serve to display the homogeneity (or diversity) of the data
gathered? The use of matrixes can serve to bring together variables, themes and cases, as
can lists, network diagrams and graphical displays. The totality of this approach may well
be neater and more powerful in capturing the reader’s attention through the different
perspectives presented but it may also result in the complexity of the findings being
‘dumbed down’ and those findings that are not matched by the other data set somehow
dropping off th…

Order an Essay Now & Get These Features For Free:

Turnitin Report

Formatting

Title Page

Citation

Outline

Place an Order