Working Title

Mixed Method of Research; a rationale for the use of MAXQDA Data

Analysis Software
Key words

Data analysis, Mixed-method of research, Quantitative research method, Qualitative

research method

INTRODUCTION

Rene Descartes (1596–1650) believed that we can only be sure of our own existence because we
perceive through the activity of the mind. „„Cogito, ergo sum‟‟ („„I think; therefore I am.‟‟). By
this logic knowledge generation is central to man‟s existence. However, the act of gathering
information and checking data alone is not considered research. It is at best data collection that is
crucial for our everyday existence. The many ways of acquiring knowledge include tradition,
authority, logical reasoning, experience, intuition, borrowing and the scientific method. Of these
ways, the scientific method is the most sophisticated and reliable. In science, research is the
diligent systematic enquiry into nature and society to validate and refine existing knowledge and
to generate new knowledge. Leedy and Omrod, (2012) further elucidated that it is the process by
which a person attempts to find an answer to a question or a solution to a problem through a
systematic methodology with the aid of an evident fact. It is not only a set of skills but a way of
thinking, by looking at a situation inquisitively, analytically and critically so as to gain an in-
depth understanding of its relevance, rational, efficiency and effectiveness (Kumar, 2014). The
systematic way of achieving the aim of a research is through a research methodology (Rugg and
Petre, (2007). Grix (2010) defined research methodology as a systematic approach that describes
how a piece of research should be undertaken to achieve certain objectives through the use of
research methods.

RESEARCH METHODS

Research methods are techniques or processes employed by researchers to conduct, collect and
analyze data (Kinash, 2006). The three major methods within which social research is carried-out
are qualitative, quantitative and mixed methods approach (Cresswell, 2013). These methods are
sometimes referred to as paradigms or mode of enquiry, because they provide a pattern of
principles and practices that regulate inquiry within a discipline by providing lenses, frames and
processes through which investigation is accomplished (Weaver and Olson 2006, p. 460).

1
Quantitative Research

Quantitative research method shares its philosophical foundation with the positivist paradigm
(Weaver and Olson, 2008). Synonymous to positivism, quantitative research shows that reality is
independent of the researcher and must be considered from an objective point of view through
measurement and prediction (Biggam, 2015). This statement confirms that of Creswell (2014)
who defined quantitative research method as one in which objective theories are tested by
examining the relationship among variables. These variables are measured on instruments to
obtain numerical data which are then analyzed using mathematical based methods (in particular
statistics). In addition, to these definitions Kumar (2014) stated that quantitative method is
embedded in the idea of rationalism. It emphasizes on greater sample size, classification or
measurement of variables and finally it takes into account the quantifiable degree of variation in
situations, phenomena and issues.

Qualitative Research

Qualitative research method shares its philosophical foundation with interpretive paradigm
(Weaver and Olson, 2008). Synonymous to interpretivism, qualitative research supports the
views that both reality and truth are manifold. In order word, reality is subjective and constructed
by social actors (Biggam, 2015). Hence, through qualitative method, the researcher has the
chance to interpret and understand how people feel and think in the circumstances they find
themselves (Weaver and Olson, 2008). Kumar (2014) stated that, unlike quantitative method,
qualitative method is embedded in the idea of empiricism. It follows an open, flexible and
unstructured mode of enquiry that aims to explore diversity rather than quantity. Kumar further
explains that qualitative research emphasizes on description and narration of experiences,
perceptions and feelings rather than their measurement. It also communicate results in a
descriptive and narrative manner as opposed to quantitative method which reports findings
analytically.

Mixed Research

Many social issues require mixed methods approaches to fully understand the problems and help
to find solutions. The suggestion to integrate qualitative and quantitative data has become more
and more popular in the last years. The debate between the qualitative and the quantitative
paradigm that dominated since the 1970ies debate seems to be more balanced today. A lot of
researchers nowadays take a pragmatist approach to research by employing different methods
depending on the research question they aim to answer (Sukamolson, 2007). Mixed Method can
be referred to as an approach to social enquiry whereby, two methods or procedures are used to
undertake a research study. This is based on the belief that different methods or paradigms have
different strength. Therefore, in most cases their combined strength will lead to improve in
accuracy and depth of the findings. The collection or analysis of quantitative and/or qualitative

2
data in a single or multi-phased study is gaining rapid recognition in social researches. The
benefit of this approach is that it exploits the flexibility of the different approach to complement
the other, by providing a better understanding, convincing and accurate findings. This form of
research can be employed concurrently or sequentially to investigate the same phenomenon at
the same time (Grix, 2010). In essence one can use a qualitative in-depth interview and follow it
up with a questionnaire which could be used for statistical analysis (Neuman, 2000 in Grix,
2010).

PROBLEM STATEMENT

The decades of debate and confrontations between the proponents of both quantitative and
qualitative method have had significant impact on software developers. Consequently, many
software applications in the market are either used for qualitative or quantitative analyses. With
the current increase in the adoption of mixed-methods of research, software developers are
racing to meet up with the global research demands. However, problems such as interoperability
issues, marketing competition, complexity of technology and resistance to change have been
negatively affecting these software companies. However, those software companies who have
anticipated this wagon of change are now leading the race for integration of both quantitative and
qualitative data in one software package to enable ease of analysis. The only few programs on
the market that pay special attention to the mixed methods approach are MAXQDA, QDA Miner
and NVivo. This study will focus on MAXQDA which is currently among the world-leading
package for mixed method research analysis.

METHODOLOGY

This study undertook a review of documents related to social research in general and the
different philosophical foundations that underpins the two major research methods (i.e.
quantitative and qualitative method). The review also explored the general features of
MAXQDA software application and its mixed method capabilities which stand out among other
data analysis software applications that are confined to only one research method per package.
The review also explored some of the challenges to the adoption of MAXQDA among
researchers. The data obtained from the review were reported in a descriptive way using visual
image of the MAXQDA software interface to enable readers understand it better. Lastly
recommendations were to facilitate and appreciate the use of an integrative software better
research analysis.

FEATURES AND PROCEDURES FOR DATA INTEGRATION IN MAXQDA

MAXQDA is one of the pioneers in the field of method integration. Functions for combining
qualitative and quantitative data are already found in the very first versions of the program in the
late 1980s and early 1990s. With the option to manage a data set of standardized, quantitative

3
data parallel to the texts, the foundation was laid. Since version 10 there is a further tool in the
form of code variables, which allows the definition of variables below the case level of a
document and since version 12.2 there is a fully integrated statistics package “Stats” for carrying
out descriptive and inferential statistical calculations, the results of which can be used directly
for the integrative analysis of qualitative data.

Fig. 1 Mixed method features

In the Mixed Methods Tab (see figure 1), functions are available which either link documents
and variables, e.g. the topics from qualitative interview material with the variables from
standardized interviews, or which carry out quantitative evaluations based on the coding carried
out. The first are so-called joint displays in which both qualitative and quantitative data, results
and conclusions are presented together. Guetterman, Creswell and Kuckartz (2015) present
various mixed method designs and suitable joint displays in an overview article. MAXQDA
offers several joint displays suitable for common mixed method designs. These designs include
in particular

 Convergent Designs (qual. and quan. study parallel)

 Explanatory Designs (qual. study after quan.)
 Exploratory Designs (qual. study before quan.)

OVERVIEW OF MIXED METHODS FUNCTIONS IN MAXQDA

 Activate Documents by Variables – this enables activate documents to be included in

the Coding Query based on document variable values. For example, one can use this
function to identify what men between the ages of 40 and 50 said about migration
issues.>
 Interactive Quote Matrix – creates a Word file showing what different groups said
about a theme based on certain variable values that are specify. Each group‟s coded
segments for the specified codes are in a different column.For example, one can choose
to see how those with various levels of education differ on their approach to combating
homelessness.

4
  Crosstab – works parallel to the Code Matrix Browser, except that this function doesn‟t
work on the document level. Instead, one can create groups based on variable values and
compare how often each of these groups talks about each theme. For example, comparing
how often men talk about relationships in their life satisfaction interviews in comparison
to women.
 Quantitizing – This is the transformation of qualitative coding information into
quantitative variables. Quantitizing it allows the researcher to store the code frequencies
as document variables, such that for each document have information about how often a
code appears in that document. This information can then be analyzed statistically or used
for the selection of cases.
 Typology Table – shows an overview of variable values for qualitative typologies that
have been created (e.g. for people with various views on combating their own
homelessness). For example, one can see what the mean age, gender breakdown, and
average time already homeless is for the “apathetic pessimists” in comparison to the
“proactive optimists.”
 Similarity Analysis for Documents – selected documents are analyzed on the basis of
existing coded segments and document variables for their similarity, and the results are
presented in a similarity or distance matrix.
 Side-by-side Display of Results – This joint display compares the results of a qualitative
study with those of a quantitative study.
 QUAL Themes by QUAN Groups – This joint display is used to compile coded
segments or summaries in a table for groups formed on the basis of variable values.
 Statistics by QUAL Groups – The result of this function corresponds to the typology
table and divides documents into groups according to codes assigned to them. This joint
display allows comparison of average values, standard deviations and absolute and
relative frequencies of selected variables for these groups.
In the “Code System” window there are functions available that allows the use of code
frequencies for each document as document variables:

 Transform code into document variable or categorical document variable – Codes

can be added as document variables that specify how often the code occurs in the
document (“Quantitizing” as described above) or which subcode occurs most frequently
in a document. The latter is particularly useful for evaluative qualitative content analyses.

QUALITATIVE AND QUANTITATIVE DATA TYPES: TEXT AND DOCUMENT

VARIABLES (ATTRIBUTES)

All qualitative data to be analyzed with MAXQDA are managed in MAXQDA‟s document
system. If different methods have been used, e.g. open interviews, observations, focus groups,
the corresponding data may be arranged in different folders (“document groups”) of the
5
document system. For managing quantitative data it is essential if the data belongs either to the
entire document or to parts of the document. According to this relation MAXQDA differentiates
between

a) document variables (also called “attributes”) and

b) code variables (also called “profile variables”)

A typical example for document variables is a study with open interviews and with a
standardized questionnaire too. For instance, the same persons have filled out a questionnaire
with socio-demographic data and were then being interviewed about status passages in their
biography. In this case the data for each person would consist of the transcription of the
interview and a vector of numbers according to the codebook of the standardized questionnaire.
Thus, document variables are on the level of the case(s), therefore one could also name them
“case variables”. The data of a focus group may be taken as a typical example for code variables.
Different persons are attending a focus group that is usually transcribed verbatim. The data itself
consists of only one text file with different speakers. In this case, the document variables are no
longer useful for storing characteristics of the participants since we have more than only one
person. The code variable offers a way to store data for multiple units of a text file provided that
the contributions of the different speakers have been coded with their name. Both types of
variables are alike statistical variables in software like SPSS, SAS or SYSTAT.

Document Variables (Attributes)

MAXQDA offers to associate a set of quantitative data with every document in MAXQDA‟s
document system. The link between the document and its variables (in the terminology of
MAXQDA “attributes”) in the quantitative data matrix is created by the document name (in
MAXQDA 2007 “text name”) and the name of the document group (in the older version
MAXQDA 2007 “text group”).

6
 Fig. 2 Linking documents (text) and quantitative data

The definition of the document variables and the entry of variable values can be done in a way
that is quite similar to statistics software. In MAXQDA‟s „Attibutes“ feature the user can switch
between “data view” and “list of attributes”. Fig. 2 shows the data view, a rectangular data
matrix that has as many rows as documents are listed in the document system. The columns are
setup by the variables, here „Text group“, „Text name“, „Gender“, “Age 19” etc. The other view
„List of Attributes“ offers the variable list in order to define new variables or delete existing
variables.

7
 Fig. 3 The List of Document Attributes in MAXQDA

Code Variables (Profile variables)

In some cases it may be useful to associate a set of variables not with the entire text but with
parts of it. For instance, a transcript of a focus group or group discussion may contain different
speakers and it may be useful to associate a set of variables with each one. This wouldn‟t be
possible with variables on the text level since only one set of variable values would be allowed
per text.
MAXQDA 10 offers a new type of variables named “code variables” or “profile variables” that
are associated with parts of a text that have been coded previously.

Statistical analysis of the quantitative data

Export and import of the data matrix Statistical procedures are not offered by MAXQDA but an
easy-to-use interface allows exporting of the entire data matrix in tab-delimited format. The
matrix contains the variable names in the first row. So after seconds the statistical analysis can
start with the software like SYSTAT or SPSS.

The exported data file can also be draged into Excel and statistics and diagrams can be made
with this standard software.

8
An existing data matrix can also be imported into MAXQDA in order to use it in combination
with the qualitative data. It is necessary that the file contains the variables textgroup and text
name and that these variables correspond to the appropriate text.

Connecting text and variables in selective retrievals

Selective retrievals enable the user to bring together text and variable values. The function
„Activation by variables“ allows to formulate selection criterias as „Gender = male“ and all cases
are checked and activated if they fulfill the logical condition.

It is possible to formulate quite complex criteria like „gender =male“ AND „ Age over 24“ and
„income > 2.500€. The selected cases can be saved as a text set in order to analyze this special
selection in detail.

Having explained the general features of MAXQDA and its capabilities in analyzing both
quantitative and qualitative data, it is timely to quote a review of MAXQDA software application
by Sam Ladner a PhD. Senior UX Researcher at Amazon:

“I spent several months researching the options, and ultimately decided to trial
MAXQDA. We brought in a MAXQDA certified trainer, and bought a network
license so that our large team at Microsoft could use the tool. We were not
disappointed[…]
I was so convinced in its efficacy in the applied qualitative field that I bought
MAXQDA for my team when I joined Amazon. I was especially delighted when
they recently added the Statistical package, which allows us to avoid the extra
expense of buying SPSS”.

CHALLENGES TO THE ADOPTION OF MAXQDA

Like any new innovations, MAXQDA has encountered some inherent challenges by researchers
and users. The major challenges were identified as resistance to change, lack of awareness,
perceived complexity, and interoperability issue.

Resistance to change

The introduction of new technology or innovation has historically posed resistance to adoption.
The integrated MAXQDA is not an exception. Being a novel approach, some researchers are
apprehensive of the unknown, while to some, it presents a threat to their expertise or status in
order data analysis software. Nonetheless, the software promoters are now beginning to
demonstrate the clear personal gain and advantages of the software which has now led to
progressive acceptance.

9
Lack of Awareness

Lack of awareness of the integrated MAXQDA has considerably impacted on its adoption. Being
aware gives an insight into the positive direction one need to go to make changes and
successfully adopt a new technology to achieve stated aims and objectives. Thus, the integrated
MAXQDA teams are now adamant on creating awareness through various avenues and market
strategies.

Managing resistance to change starts with understanding it. The introduction of new technology
in the workforce has historically posed a major threat to workers’ sense of security, stability and
purpose. Key reasons people resist change can include:

Perceived complexity

Perceived complexity or difficulty with using new technology is associated with lack of skills
and lack of available resources to access that skill. Tutorial videos, documents and training
sessions are now made available and accessible to many researchers and users who are interested
in exploring the integrative platform of the MAXQDA. This endeavor is perceived as the right
step to counteract perceived complexity and difficulty of the software.

Interoperability issue

Interoperability refers to the basic ability of different computerized products or systems to

readily connect and exchange information with one another, in either implementation or access,
without restriction. MAXQDA presents some issues of interoperability with other common data
analysis software applications and this greatly its rapid adoption by researchers. Presently, there
is ease of data exchange with SPSS package even though to some extent there are still
limitations.

Reference

Babbie, E. (2015) The practice of social research. Boston: Cengage Learning.

Biggam, J. (2015) Succeeding with your Master’s Dissertation. Berkshire: 3rd ed. Open
University Press.
Carnall, C.A. (1990) Managing Change in Organisations. London: Prentice Hall International
(UK) Ltd.
Creswell, J. (2014) Research Design: Qualitative, Quantitative and Mixed Method Approach. 4th
ed. Sage.

Creswell, J. W (2013) Research design: qualitative, quantitative, and mixed method approaches.
4th ed. California: SAGA Publications, Inc.

10
Denscombe, M. (2007) The Good Research Guide for Small Scale Social Research Projects. 3rd
ed. Glasgow: Bell and Bain Ltd.

Grix, J. (2010) The Foundation of Research. 2nd ed. London: Palgrave Macmillan

Hammersley, M. (2012) Methodological Paradigms in Educational Research, British

Educational Research Association [online]. [Accessed 11 April 2021]. Available at: <
https://www.bera.ac.uk/wp-content/uploads/2014/03/Methodological-Paradigms.pdf>.

Kinash, S. (2006) Paradigms, methodology & methods. Bond University.Australia [online].

[Accessed 16 Febrary 2021]. Available at: <
http://bond.edu.au/prod_ext/groups/public/@pub-tls-
gen/documents/genericwebdocument/bd3_012336.pdf>

Kumar, R. (2014) Research Methodology: A Step-By-Step Guide for Beginners. 4th ed. London:
Sage.

Myers, D. (2013) Construction Economics: A New Approach. 3rd ed. New York: Routledge

Myers, M. D. (2009) Qualitative Research in Business & Management. London: Sage

Publication.
Sukamolson, S. (2007) Fundamentals of quantitative research. Language Institute Chulalongkorn
University. pp. 1-20.

Weaver, k. and Olson J. K. (2008) Understanding paradigms used for nursing research. Journal
of Advance Nursing [online]. 53(4), pp. 459-469 [Accessed 11 April 2015]. Available
at: < http://onlinelibrary.wiley.com.ezproxy.wlv.ac.uk/doi/10.1111/j.1365-
2648.2006.03740.x/pdf>.

11