See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332599843

Inductive Coding

Chapter · April 2019

DOI: 10.1007/978-981-13-3170-1_8

CITATIONS READS

91 10,801

2 authors:

Yanto Chandra Liang Shang

City University of Hong Kong Lingnan University
162 PUBLICATIONS 3,267 CITATIONS 42 PUBLICATIONS 505 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Yanto Chandra on 01 February 2021.

The user has requested enhancement of the downloaded file.

 Qualitative Market Research: An International Journal
An RQDA-based constructivist methodology for qualitative research
Yanto Chandra Liang Shang
Article information:
To cite this document:
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Yanto Chandra Liang Shang , (2017),"An RQDA-based constructivist methodology for qualitative
research ", Qualitative Market Research: An International Journal, Vol. 20 Iss 1 pp. 90 - 112
Permanent link to this document:
http://dx.doi.org/10.1108/QMR-02-2016-0014
Downloaded on: 09 January 2017, At: 17:42 (PT)
References: this document contains references to 132 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 2 times since 2017*

Access to this document was granted through an Emerald subscription provided by emerald-
srm:364597 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald
for Authors service information about how to choose which publication to write for and submission
guidelines are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as
well as providing an extensive range of online products and additional customer resources and
services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the
Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for
digital archive preservation.

*Related content and download information correct at time of download.

 The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1352-2752.htm

QMR
20,1
An RQDA-based constructivist
methodology for
qualitative research
90 Yanto Chandra and Liang Shang
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Department of Public Policy, City University of Hong Kong,

Received 11 February 2016 Hong Kong, Hong Kong
Revised 8 September 2016
21 November 2016
Accepted 21 November 2016

Abstract
Purpose – Qualitative research suffers from “contestation” and a lack of “boilerplate” problems to assessing
and presenting qualitative data, which have hampered its development and the broader acceptance of
qualitative research. This paper aims to address this gap by marrying the constructivist methodology and
RQDA, a relatively new open-source computer-assisted qualitative data analysis (CAQDAS)-based R
extension and demonstrate how the software can increase the rigor, transparency and validity of qualitative
research.
Design/methodology/approach – This paper highlights the constructivist approach as an important
paradigm in qualitative research and demonstrates how it can be operationalized and enhanced using RQDA.
It provides a technical and methodological review of RQDA, along with its main strengths and weaknesses, in
relation with two popular CAQDAS tools, ATLAS.ti and NVivo. Using samples of customer-generated
e-complaints and e-praises in the electronics/computer sector, this paper demonstrates the development of a
process model of customer e-complaint rhetoric.
Findings – This study offers step-by-step instructions for installing and using RQDA for data coding,
aggregation, plotting and theory building. It emphasizes the importance of techniques for sharing coding
outputs among researchers and journal gatekeepers to better disseminate and share research findings. It also
describes the authors’ use of RQDA in classrooms of undergraduates and graduate students.
Research limitations/implications – This paper addresses the “contestation” and “boilerplate” gaps,
offering practical, step-by-step instructions to operationalize and enhance the constructivist approach using
the RQDA-based approach. This opens new opportunities for existing R users to “cross over” to analyzing
textual data as well as for computer-savvy scholars, analysts and research students in academia and industry
who wish to transition to CAQDAS-based qualitative research because RQDA is free and can leverage the
strengths of the R computing platform.
Originality/value – This study offers the first published review and demonstration of the RQDA-based
constructivist methodology that provide the processes needed to enhance the rigor, transparency and validity
of qualitative research. It demonstrates the systematic development of a data structure and a process model of
customer e-complaint rhetoric using RQDA.
Keywords Qualitative research, CAQDAS, Constructivist, RQDA
Paper type Technical paper

Introduction
Qualitative research is one of the most popular methods used in the business and social
sciences journals. It is a diverse collection of tools and approaches that deal with
predominantly “non-numerical” data and span from interpretive textual analysis
Qualitative Market Research:
(Thompson, 1997), grounded theory (Eisenhardt, 1989a, 1989b), case study (Yin, 2003),
An International Journal discourse analysis (Paulus and Lester, 2016), ethnography (Arnould and Wallendorf, 1994),
Vol. 20 No. 1, 2017
pp. 90-112 to netnography (Kozinets, 2002). The styles of theorizing (reasoning and writing) in the
© Emerald Publishing Limited
1352-2752
qualitative research enterprise are also diverse, from the so-called thick description, pattern
DOI 10.1108/QMR-02-2016-0014 description, narration (Cornelissen, 2016), process theorizing (Langley, 1999; Giesler and
 Thompson, 2016) and complex theorizing (Tsoukas, 2016). Scholars also agree that there are Methodology
different logics and model of writing qualitative papers (Coviello, 2014; Hunt, 1991; for qualitative
Goulding, 2005; Langley and Abdallah, 2011).
The methodological diversity of the qualitative research enterprise is a two-edge sword
research
(Hunt, 1991; Cornelissen, 2016); it leads scholars to the paths of discovery that push the
frontiers of scholarship, while at the same time, it often creates confusion as to what qualifies
or characterizes a high-quality qualitative research. Given the diversity of the nature of
91
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

qualitative research, we argue that qualitative research enterprise may suffer from a
“contested concept” problem (Choi and Majumdar, 2014; Gallie, 1955) – just like a work of art
where there is no agreement on what constitutes a “work of art” (Gallie, 1956, p. 167) – which
makes it difficult for stakeholders to judge qualitative research. We argue that this “work of
art” problem is hampering the legitimacy of the qualitative research enterprise as a whole
and may explain why qualitative research remains the minority in top-tier journals in the
business fields (Hanson and Grimmer, 2007; Pratt, 2008; Short et al., 2009). This echoes
scholars’ critiques that qualitative research is lacking in rigor and validity. As Pratt argued,
qualitative research is often challenging because of “no accepted boilerplate” (Pratt, 2009,
p. 856) in helping readers or gatekeepers understand how a researcher assesses and presents
qualitative data. Assume that a qualitative scholar found the evidence of a “talking dog” that
constitutes a groundbreaking concept and that this makes a major contribution worthy of a
Noble Prize. The challenge is how the author can demonstrate and persuade readers of the
existence of a “talking dog”. Qualitative scholars trying to submit their work to the best
journals in their field are often confronted by this type of problem.
Our extensive reviews[1] of top-tier and important journals in marketing reveal that
although qualitative scholars have used both the constructivist (theory building and
exploration; Belk et al., 2003; Gebhardt et al., 2006) and positivist (theory testing and
refinement; Grayson and Shulman, 2000) qualitative research approaches, theory-building
remains the more popular approach in qualitative research, as it helps scholars discover new
concepts, mechanisms, processes and turning them into models or theories to describe and
explain social reality. Although theory-building characterizes the bulk of qualitative papers
in marketing, there is a lack of review papers that provide “how to guide” in doing rigorous,
transparent and credible steps in building theory from scratch and in a step-by-step manner.
Importantly, the constructivist approach was used “almost like an art”, without a
step-by-step demonstration of how the qualitative data were turned into a model, mechanism
or theory (Malshe, 2009; Malshe et al., 2013; Reficco and Jaén, 2015). The majority of
qualitative scholars preferred “manual analysis” approach (Basit, 2003; Commuri, 2009;
Coviello, 2014), and only a minority used computing technologies (i.e. software) to facilitate a
rigorous and transparent data analysis (i.e. NVivo software, see Coviello and Joseph, 2012;
Homburg et al., 2014).
Arguably, the major differences in the conduct of research activities in the twenty-first
century as opposed to those in the previous century are the advancement of computing
technologies, particularly the widespread diffusion of open source software (Daniel and
Stewart, 2016; Phipps, 2015), and the prevalence of technology savvy scholars and students
(Weller, 2011) and the greater demand for transparency in publishing due to rising cases of
scientific misconduct (Altman and Moher, 2013; Corbyn, 2012). These trends are irreversible
and present new opportunities for methods-oriented scholars to experiment with new (and
hopefully better) ways of doing qualitative research. Technology, particularly software, is a
powerful tool to facilitate qualitative research and can enhance (and impede) the
transparency in doing qualitative research (i.e. researcher agency; Woods et al., 2015a)
through data sharing, articulation of decisions made during data analysis and presentation.
 QMR The increasing retraction of papers in top tier journals in the field of sciences (Fang et al.,
20,1 2012; Gewin, 2014) is a serious concern and calls for more transparency and ethically driven
qualitative research scholarship including those in the social sciences (Karabag and
Berggren, 2016).
In this paper, we seek to enhance the legitimacy of and address the “contestation” and
“transparency” problems in qualitative research enterprise by proposing generally accepted
92 qualitative research techniques as a way of building a common language and understanding
about the nature and process of doing qualitative research. We do not advocate a
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

standardized approach to doing qualitative research but rather propose a set of processes
that can help qualitative (and quantitative) scholars appreciate and take into account the
issues of rigor, transparency and credibility as important traits of qualitative research
(Sinkovics and Alfoldi, 2012; Woods et al., 2015a). Specifically, we seek to make
methodological contributions to the qualitative research enterprise by fusing the widely
accepted constructivist paradigm and a new open source computer-assisted qualitative data
analysis (CAQDAS) software to enhance the rigor, transparency and credibility of
qualitative research.
Our methodological demonstration by combining constructivism and a new open source
CAQDAS to offer “rigor and systematization” may resonate with a critical realist approach to
qualitative research (Easton, 2010; Sobh and Perry, 2006; Tsang, 2014). In doing so, we first
conducted a review of the major paradigms in qualitative research and their key ontological
and epistemological differences. We then discussed CAQDAS literature and focused on two
main (proprietary) CAQDAS software and how they fared with RQDA, the new open source
CAQDAS of interest. Next, we reviewed the strengths and limitations of RQDA and
articulated a step-by-step constructivist research approach and how to operationalize it
using RQDA. We demonstrated how we conducted data collection and analysis and, finally,
how we translated the findings into a process model. We concluded this article with our
reflection on the benefits and potential benefits of our RQDA-enabled constructivist
qualitative research, its limitations and future research opportunities. To the best of our
knowledge, this is the first paper that provides a “how to” guide in conducting an
RQDA-enabled constructivist approach in doing qualitative research.

Major paradigms in qualitative research methodology

Qualitative research is a challenging research enterprise, which is particularly true with
respect to scholars aiming to publish in top-tier journals, as there is often “no accepted
boilerplate” (Pratt, 2009, p. 856) in terms of the rules (investigation processes), formats and
norms that are involved. Although some qualitative scholars prefer “no boilerplate” in the
conduct of qualitative research (Friese, 2011; Cornelissen, 2016), there are several paradigms
that have gained legitimacy in recent years. Perhaps the most identifiable ones are
constructivism and positivism (Table I).
Constructivism – which views social reality as subjective and co-constructed through
human experience (Peters et al., 2013; Ramoglou and Tsang, 2015) – emphasizes the role of
qualitative methods as a means to interpret, explore and discover new concepts, constructs,
theories and frameworks or models – and tends to promote a “flexible” way in how
qualitative research is conducted (Coviello, 2014). Positivism – which views social reality as
objective and “exist out there” for verification (Hunt, 1994; Peters et al., 2013) – emphasizes
qualitative methods as a way of validating constructs, propositions or hypotheses using
qualitative and quantitative data (Dubé and Paré, 2003; Langley and Abdallah, 2011) – and
therefore promotes “rigor and systemization” in how qualitative research is performed; it is
inspired by the dominant hypothetico-deductive paradigm in the social sciences. The former
 Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Aspects of comparison The positivist approach The constructivist approach

Philosophical underpinnings of Empiricism (reality exists “out there” waiting to be captured, described and Constructivism (reality is socially constructed
social reality explained) and interpreted)
Methodological foundation Yin’s (2003) case-study-as-experiment logic; one case is (dis)confirmed by Strauss and Corbin’s (1990) grounded theory
other cases to reach theoretical saturation by open and axial coding
Purpose of use Theory testing, refinement, replication, elimination of alternative explanation Theory building, exploration, discovery
The role of research question; a Both are needed at the outset of the study Research question is needed at the outset of
priori constructs the study; consult but not dictated by a priori
constructs or theory
Sampling rationale Theoretical sampling that seeks maximum variation (e.g. polar types Revelatory and richness driven sampling (e.g.
sampling of high vs low performing firms); representativeness is important critical case to test a theory; extreme or
revelatory case that offers new insights);
representativeness is not crucial
Number of cases Preference for multiple cases (6 to 10 cases) to increase variations on Single case and multiple cases (up to 4 cases);
dependent variable (e.g. firm performance) not seeking variations but explaining rich,
interesting cases
Types of data collected Qualitative (e.g. interviews, observations, secondary data) and quantitative Primarily qualitative (e.g. interviews collected
(e.g. questionnaire) data at multiple rounds) data
Flexibility of interview protocol Interview protocol is akin to a measurement device to capture reality Preserve the flexibility in adjusting interview
systematically protocol; backtrack to informants based on
insights in subsequent interviews
Analytical process and final Start with within-case and cross-case analysis; develop tabulation of evidence Start by developing data structure
product to refine constructs; end with testable propositions (aggregating first-order concept, second-order
categories, aggregate dimensions); leading to
a process model and new concept development
Style of the presentation of Start with propositions, then tables of evidence (with short quotes, some No fixed style. But tend to start with the data
research findings numerical data, other constructs observed), finally a “box and arrows” model structure and process model (consisting
concepts and their relationships), then
present the power quotes in “data tables”
Approach in data analysis Primarily use manual data analysis (i.e. using pens/pencils, paper, Mostly use manual data analysis, but a few
highlighter, folders, etc.) uses CAQDAS (Nag et al., 2007)
Examples of important findings A model of fast decision-making (Eisenhardt, 1989); possessions as The discovery of “sense giving” concept;
irreplaceable or replaceable (Grayson and Shulman, 2000) “identity ambiguity”, “transitional identity”
concepts (Gioia et al. 2013)
Key papers Eisenhardt (1989a, 1989b), Eisenhardt and Graebner (2007) Gioia et al. (2013), Nag et al. (2007); Vaccaro
and Palazzo (2015)
93

qualitative research
approach in
constructivist
positivist and
A comparison of
Table I.
research
for qualitative
Methodology
 QMR is often used for theory-building and discovery research, while the latter for theory-testing
20,1 and refinement research. Some scholars argue that the two might be complementary (Hunt,
1991; Lin, 1998; Sobh and Perry, 2006) because the social reality is partly real, partly
co-constructed and interpreted and hence structured in various levels; thus, both camps can
be combined to better study the social reality. This gives rise to what some called critical
realism (Easton, 2010; Sobh and Perry, 2006; Tsang, 2014).
94 As shown in Table I, where we compared the two paradigms, the positivism and
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

constructivism approaches differ in 11 dimensions (for the sake of parsimony, we only

discussed the two polar types and did not discuss critical realism). The positivist approach
primarily focuses on theory/construct validation, refinement or disconfirmation, typically
involving one or more dependent variables (e.g. firm performance) and using multiple cases
(usually the more the better). For example, in their study of the speed of decision-making in
high velocity environment, Eisenhardt (1989a) used qualitative and quantitative data to
provide numerical estimate of the speed of decision-making (i.e. using decision durations as
data, number of meetings and classification of active versus passive conflict resolution to see
variations among cases) and used very limited informants quotes in the data presentation. In
another example, Grayson and Shulman (2000) used interview data by transforming the data
into categories (the presence or absence) of constructs which were then tested quantitatively
using chi-square analysis. In both examples, qualitative data and method were used in a
quantitative sense.
In contrast, the constructivist approach focuses on exploring and discovering new
concepts, mechanisms or processes and ultimately building a process model or theory by
using open and axial coding (Strauss and Corbin, 1990) of a small number of cases (e.g. from
1 to 4). As an example, Gebhardt et al. (2006) studied how two organizations changed to
become more market oriented and developed a four-stage process of creating market
orientation: initiation, reconstruction, institutionalization and maintenance. What these
scholars did was to enrich our understanding of the market orientation by focusing on its
adoption process by organizations (Jaworski and Kohli, 1993) that had been primarily used in
a quantitative sense in the past two decades of research on the topic. In a similar fashion,
Gioia et al. (2013) used qualitative research to unravel new concepts about organizational
identity such as “sense giving”, “identity ambiguity” and “transitional identity”. As such,
under the positivist approach, researchers conduct extensive literature reviews and derive
a priori constructs or theory prior to data collection; under the constructivist approach,
researchers are not dictated by a priori constructs or theory. The positivist approach relies on
theoretical sampling that seeks maximum variation of the data, while the constructivist
approach focuses on revelatory- and richness-driven sampling.
Another key difference is the approach to wrap up a qualitative research or “reach
closure”. The positivist approach focuses on tabulating evidence to refine constructs, paying
attention to variables that influence a dependent variable (e.g. decision durations, type of
decisions, types of stakeholders that influence high vs low firm performance (Eisenhardt,
1989a), using a few short informants’ quotes, and developing testable propositions and
representing the findings in a box-and-arrow model to demonstrate causality. For instance,
Eisenhardt (1989a) above offered “closure” by developing a model of strategic decision in
high velocity environments along with propositions as the summary. The constructivist
approach has no precise style of presenting the findings but tends to demonstrate the coded
qualitative data (or sometimes called “data structure”) that contains first-order, second-order
and aggregate theoretical dimensions; and a process model that demonstrates the
mechanisms or processes of a phenomenon under investigation. An example includes the
study of young activists who successfully challenged mafia organizations from the practice
 of paying protection money (Vaccaro and Palazzo, 2015). These scholars gradually coded Methodology
various first-order interview data into six second-order themes (i.e. values, hooking, for qualitative
anchoring, activating, securing and uniting) and finally aggregating them into two major
dimensions (i.e. moralizing and integrating). Researchers using the constructivist approach
research
present “power quotes” (i.e. the most revelatory examples of informants’ speeches or
statements) in the text and data tables. For example, Vaccaro and Palazzo (2015) creatively
inserted short informants’ quotes in the text and used tables to summarize various important
themes and the corresponding informants’ quotes to portray the data and their abstraction. 95
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

An overview of computer-assisted qualitative data analysis literature

CAQDAS has been used for about 30 years with NVivo and ATLAS.ti being its earliest
software applications. CAQDAS-based qualitative research has been used widely across
different fields, from arts and humanities, business, to nursing and medicine, and has been
highly popular in the health sciences (Woods et al., 2015a, 2015b). Scholars have published
numerous articles that discuss the advantages and disadvantages of various CAQDAS
software that supports qualitative research such as NVivo, Atlas.ti to MaxQDA (Franzosi
et al., 2013; Friese, 2011; Saillard, 2011; Woods et al., 2015a, 2015b), as well as the
methodological foundations behind CAQDAS-based qualitative research (Leech and
Onwuegbuzie, 2011; Muhr, 1991; Sinkovics and Alfoldi, 2012; Sinkovics et al., 2005).
Our review of the CAQDAS literature reveals a consensus among researchers who view
CAQDAS positively as a resource to enhance: the transparency, validity, rigor and
trustworthiness of qualitative research projects (Sinkovics and Alfoldi, 2012; Woods et al.,
2015a); the efficiency in coding and in code comparisons (Bazeley, 2007; Saillard, 2011); the
efficient handling and management of relatively large data sets (Morse and Richards, 2002);
the management of the “messiness” of data collection during fieldwork (Sinkovics and
Alfoldi, 2012, p. 817). Qualitative scholars are also becoming aware of the pitfalls (potential
and actual) of CAQDAS including “reflexive blindness” (i.e. lack of awareness of the
software’s influence on how researchers conduct research (Woods et al., 2015b), rigid and
inflexible analysis (MacMillan, 2005), using a technique/tool simply because the software
allows it (García-Horta and Guerra-Ramos, 2009), incorrect code label naming (Davis and
Meyer, 2009), contextually distancing the researcher from his/her the data (Sotiriadou et al.,
2014) and “coding trap” (being too close to the data; Gilbert, 2002), among others. These
pitfalls provide opportunities to improve the use of CAQDAS in qualitative research. We
argue that CAQDAS is like a sword (i.e. researcher agency; Woods et al., 2015b): its
usefulness depends on the user’s goals, how it is used and, like in the quantitative fields, what
its community of users believe about what it stands for and how it should be used (or not). It
is also subject to the legitimization process as qualitative scholars writing in leading
scholarly journals have legitimized CAQDAS as a preferred research tool.
With rapid advancement in computing technologies, there are new avenues to further
advance the constructivist qualitative research methodology by marrying it with
CAQDAS-based tool that increases the rigor, transparency, validity and replicability of
qualitative research. In the past decade, we see the rapid growth of the open source (i.e.
“free”) computing tools (e.g. from R, Perl to Python) in the industry and academia as
highly promising resources that will support the growth of qualitative research in the
future. Despite its popularity as a programming platform, R has mostly been applied to
quantitative research (e.g. multiple linear regression, Nimon and Oswald, 2013; topic
modeling analysis, Chandra et al., 2016; analysis of co-authorship networks, Wang et al.,
2016). To date, very little research has examined ways of using R to perform CAQDAS.
Given the large number of researchers who are familiar with R worldwide (i.e. 2 million
 QMR users worldwide, Vance, 2009a, 2009b), there are plenty of opportunities for these R users
20,1 to “cross over” to analyzing textual, qualitative data as a primary or secondary research
tool. RQDA, an R extension for CAQDAS-based research, provides a new methodological
arsenal for the current computer-savvy young scholars and students who use CAQDAS
or have an interest in using it.
Based on our experience, we are aware that some scholars prefer to use “manual
96 approach” (not using any technology or software to facilitate data storage, analysis, retrieval
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

and visualization, but rather using colored pens, highlighters, stickers, different folders and
containers to classify the data; Basit, 2003; Coviello, 2014) in doing qualitative research. This
practice has its own value and works well when the research data are small and research
team members work in adjacent locations. However, problems arise when the research data
are large (e.g. involving hundreds of interviews or more, thousands of news articles as data)
and researchers work collaboratively across the globe. In other situations, authors may like
to openly share the qualitative data and the processes of transforming them into the final
product (i.e. a process model or theory) to better persuade journal editors and reviewers to
demonstrate the rigor and comprehensiveness of how they carry out research – which can be
done using online platforms (e.g. github.com, stackoverflow.com). These, in our experience,
are important arguments for why CAQDAS is beneficial for qualitative research.

RQDA as a new open source computer-assisted qualitative data analysis

capabilities and resources
RQDA is a relatively unknown yet powerful CAQDAS tool that emerged alongside the
development and popularity of the open source R computing platform (Ihaka and Gentleman,
1996; Jackson, 2014). Open source software is an important intellectual movement (Daniel
and Stewart, 2016; Phipps, 2015) in the recent decades, as programmers and scholars
understood the value of sharing and making their computing work available for upgrade and
further extension by others, which will ultimately benefit the society as a whole. RQDA
offers new capabilities compared to existing CAQDAS tools and is thus a new way of seeing
and doing qualitative research. One way to better understand RQDA is to contrast it with
two of the most widely used CAQDAS tools: NVivo and ATLAS.ti (Woods et al., 2015a,
2015b).
As shown in Table II, the three CAQDAS tools can be contrasted along their technical
specifications, computing capabilities and methodological orientation. We highlight several
key differences, and the rest are shown in Table II. Unlike NVivo or ATLAS.ti, RQDA is
open-source software and therefore is free, but, like most open software, it does not come with
technical support (e.g. “how to” on the software or a hotline center) from its developer. RQDA
supports only text (.txt) files and does not support multimedia files as the other two do. But
it supports text data in various non-English languages, including Mandarin, Japanese,
Korean, Arabic, German, French, Spanish, Indonesian and other major languages.
In terms of its main CAQDAS capabilities, RQDA offers similar capabilities as the other
two software programs do, such as data coding and aggregation, but it offers only a simple
(two-level) code aggregation to support theory-building research. This was designed to
enable researchers to be the main driver of the theory-building process by designing simple
yet efficient data aggregation function (Denzin and Lincoln, 2005). RQDA users need to know
basic programming language in R to launch RQDA, which is not a requirement for NVivo
and ATLAS.ti users. However, the actual use of RQDA is similar to the other two software
programs via “point and click” in a graphical user interface (GUI). Therefore, existing R users
and computer-savvy researchers will likely find RQDA easy to launch and use.
 Aspects of comparison RQDA ATLAS.ti NVivo
Methodology
for qualitative
Technical specification
Year of first release 2008 1993 (prototype in 1989) 1999 (preceded by
research
NUD*IST, 1989)
License Free Proprietary Proprietary
Technical support Free and open source, but Paid version and Paid version and
no technical support technical support is technical support is 97
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

available available available

OS supported Win, OSX, Linux Win, OSX, iPad, Android Win, OSX
File compatibility Text (.txt) only Audio, graphic, text, Audio, graphic, text,
video video
Languages Allow import data in Allow import data in Allow import data in
several foreign languages several foreign languages several foreign languages

Capabilities
Programming language Need some knowledge of No programming No programming
skills R programming language skills needed, it language skills needed, it
language, also have point is a point and click is a point and click
and click function program program
Statistical functions Enables users to write R Enable data attributes to Enable data attributes to
commands for statistical be transformed into be transformed into
analysis and apply tabulations or cross- tabulations or cross-
various R packages for tabulations (matrix) for tabulations (matrix) for
statistical analysis “under further statistical further statistical
one platform” analysis in SPSS analysis in SPSS
Mode of use No simultaneous use of Lack of simultaneous use Allows simultaneous use
software by multiple of software by multiple of software by multiple
users users users
Main CAQDAS functions Coding, Aggregation, Coding, Aggregation, Coding, Aggregation,
Query, Memo, Query, Memo, Query, Memo,
Visualization Visualization Visualization
Code aggregation Up to two levels of Limited to no function for Good function for
hierarchical structure of hierarchical structure of hierarchical structure of
coding coding coding
Output sharing Export to a HTML file Export to SPSS and XML Export to RTF, Excel and
HTML table
Boolean search Allow for complex Allow for complex Allow for complex
Boolean search Boolean search Boolean search

Methodological
Treatment of all materials Yes, only .txt files Yes, all types of files Yes, all types of files
as one hermeneutic unit
Large data set recording, Yes Yes Yes
storage, indexing, sorting
Cross-border collaboration Yes, via file sharing Yes, real-time team Yes, real-time team
collaboration collaboration
Coding flexibility Yes Yes Yes
Codes comparison Yes Yes Yes
Exhaustive coding Yes Yes Yes
schemes
Researcher reflexivity Yes Yes Yes
needed
Closeness to data Yes Yes Yes
Data attributes addition Yes Yes Yes Table II.
Transparency and Yes Yes Yes A comparison of
credibility of data analysis RQDA, ATLAS.ti and
Efficient, accurate retrieval Yes Yes Yes NVivo CAQDAS
Visualization of codes Yes Yes Yes programs
 QMR Perhaps one of the key strengths of RQDA is that it enables users to write R syntax to
20,1 conduct statistical analysis of hundreds to tens of thousands of textual files, thus enabling
mixed method research in a single platform. One of the most useful features of RQDA is the
transformation of its coded outputs into HTML files, which enables data sharing among
researchers across geographical borders and eases keyword search on any codes or key
words of interest loaded into RQDA. RQDA shares very similar methodological abilities with
98 NVivo and ATLAS.ti, including the treatment of all loaded textual data (e.g. interview
transcripts, news articles, company prospectuses) as a single entity for analysis or “one
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

hermeneutic unit” (Muhr, 1991); the ability to record, store, index, sort and retrieve large data
sets; flexible coding; audit trails so that other researchers or journal reviewers can check how
data were analyzed; data attributes (e.g. source of data, year of data, demographic variables
of interviewees) and visible codes (see Table II for more details).

An RQDA-based constructivist approach in qualitative research

To extend and enrich the constructivist approach using the open source RQDA as a
CAQDAS tool, we create the so-called “RQDA-Based Constructivist Approach”. Before we
demonstrate the application of a constructivist approach using RQDA, we summarize the
constructivist approach step by step (Table III).
We posed a clear research question:
RQ1. What are the characteristics of customers’ e-complaints in an electronics/computer
products setting, and how can they inform organizations to better serve the
customers?
These questions reflect the rise of online consumer review platforms (Chevalier and Mayzlin,
2006; Chen et al., 2011) (e.g. Amazon, Yelp, TripAdvisor), which makes it easy for consumers
to publicize complaints about products they purchase, and the ubiquity of “smart” devices
(e.g. smartphones, tablets, laptops), which makes it easy for consumers to share their
dissatisfaction with products. Second, we conducted an initial literature review on
e-complaints of electronics and computer products. We found a dearth of empirical studies on
e-complaints in this product category and those that exist primarily focus on the hospitality
industry (e.g. hotels, cruises and tour businesses; Buhalis and Law, 2008; Ekiz et al., 2012;
Mattila and Mount, 2003; Robertson, 2012; Ye et al., 2011). This dearth of empirical studies on
electronics and computer products e-complaints provides an opportunity to demonstrate
how to develop a process model of consumer e-complaints in a new context and to offer
practical advice to managers/CEOs in the sector on how to avoid and manage such
complaints.
We selected www.customerservicescoreboard.com/index.php as our sampling site
because this platform offers a myriad of user-generated (UG) reviews on various types of
electronics and computer products. The site classifies all UG reviews as either “negative”
(e-complaints) or “positive” (e-praises). We chose an electronics and computer company
based in Silicon Valley from this site for the demonstration and carefully read various
e-complaints and e-praises written by users before sampling the e-complaints and including
them in our study. We did not interview customers, as we could have if following the usual
constructivist approach, because our study of e-complaints naturally takes place inside an
electronic medium and therefore the best way to study e-complaints is to study UG-generated
complaints found in the electronic medium. This is similar to the netnography method that
primarily uses naturalistic and unobstrusive approach to collecting data on online
communities (Kozinets, 2002). We relied on revelatory sampling by choosing five
e-complaints and five e-praises (n ⫽ 10) of 845 e-complaints and 127 e-praises on the
 Relevance to the current RQDA
Methodology
Process Meaning demonstration for qualitative
(1) Research question Articulate a clear research question at the outset “What are the characteristics of
research
of the study customers’ e-complaint in a electronics
and computer products setting?”, “How
can they inform organizations to better
serve the customers”
99
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

(2) Initial literature review Consult the literature lightly at the outset of the Literature review on e-complaints
study, not to be dictated by it. Refine the conducted. Found a dearth of studies on
research question after initial literature review e-complaint in the electronics and
as necessary computer products category; most focus
on online hotel bookings, and a dearth of a
theory building research on e-complaints
(3) Case selection Choose a case or cases that offer revelatory and Picked an online user-generated (UG)
rich qualities (i.e. critical, extreme, revelatory review platform that contains e-
case(s)) complaints and e-praises on a well-known
smart gadget
(4) Interview protocol Develop an interview protocol; adjust the Determined by the research question. No
interview questions as the research evolves to interview protocol was developed here as
capture new, important themes or constructs the data were collected from online UG
reviews
(5) Data collection Conduct interviews, collect additional data Sampled 5 negative and 5 positive UG
through documents and observations; data reviews (to create a contrast of customer
collection and analysis overlaps in time views) from a UG review platform on Step
3 above. Transformed each UG review
into separate ASCII or .txt files for entry
into RQDA
(6) Data analysis Code the first-order terms, aggregate them to the Loaded the five e-complaints .txt files into
second-order categories, and finally the RQDA, then the other five e-praises .txt
aggregate dimensions; assemble all these into a files
“data structure” Conducted constructivist coding process
on all e-complaints and then e-praises;
created first-level codes using RQDA and
aggregated them to second-level code
categories using Plots function, and finally
created a data structure using Microsoft Table III.
Visio (for high-quality visuals) A step-by-step process
(7) Theory articulation Construct dynamic relationships among second- Developed a process model of customer of the constructivist
order or aggregate dimensions; develop a e-complaint rhetoric in the electronics and qualitative research
process model by consulting the literature computer product sector approach

company. We copied, pasted and saved all e-complaints and e-praises as individual “.txt”
files. Next, we loaded them into RQDA, conducted open and axial coding and created a
two-order code aggregation. Based on this, we developed a data structure and a process
model. In the next section, we describe the steps to install RQDA through to the creation of a
process model (Table IV).

Step 1: preparation
RQDA, a package that is installable from and runs within the R system (Ihaka and
Gentleman, 1996), has a separate window running on the GUI (using RGtk2) that was
developed by Huang Ronggui in 2008 as a free CAQDAS application (Huang, 2014). RQDA
relies primarily on GUI – a user interface that enables users to communicate with electronic
devices via graphical icons and visual indicator, rather than syntax-based user interfaces.
Users must first install R or RStudio and then RQDA. Instructions and resources for
 Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

20,1

100
QMR

Table IV.

constructivist
methodology for
An RQDA-based

qualitative research
Steps RQDA-based data analysis Advantages and disadvantages Relevant methodological concepts

Preparation (cf. process 1, #1. Install R, RSQLite, GTK⫹, RQDA All data files are transportable online via emails, Allow “logistics and cognitive efficiency” by
2, 3, 4 and 5 in Table II) #2. Transform textual data into individual Dropbox, or using a device (e.g. USB stick) placing all research materials in one
ASCII or .txt files Ease of logistics handling when moving offices, “hermeneutic unit” (Muhr, 1991) inside
#3. Launch RQDA involving multiple researchers RQDA
#4. Create a new project Not all researchers are familiar with R Ease of “recording, storing, indexing, and
#5. Upload all .txt files into RQDA sorting” large sets of qualitative data
(Morse and Richards, 2002)
Allow “cross border collaboration” (Friese,
2011)
Coding (cf. process 6 in #6. One or more coder performs inductive Increased reliability and accuracy of data Enable “coding flexibility” (Friese, 2011)
Table II) data analysis together, by highlighting the interpretation Promote “researcher reflexivity”, but
text and clicking “Mark” (“Unmark”) to Ease of capturing new, unexpected insights beware of “reflexive blindness” (Woods
do (undo) the coding through re-coding of data et al., 2015a)
#7. Other analyst(s) critiques the coding Possible mechanical errors for inexperienced Ease “codes comparison” (Bazeley, 2007)
results; reconcile differences in users A “labor intensive” process (Sotiriadou
interpretation Possible loss of reflexivity or de-contextualizing et al., 2014) but enable “closeness to the
data data” (Saillard, 2011)
Data attributes (cf. process #8. Click “Attributes” and record key data Allow easy capture of meta information of the Attributes and memos might be written
6 in Table II) attributes (e.g. data sources, informants’ data and attributes of informants “all over the places” and not analyzed
demographics, dependent variables) Memos can be copied-and-pasted to a Word (Friese, 2011)
#9. Click “Memo” to write memos for each processor to aid data analysis
case
Codes abstraction (cf. #10. Use “Add To” button to assign all first- A highly systematic way of aggregating codes Improve “transparency and credibility”
process 6 in Table II) order codes to second-order code to a higher level of meaning (that results can be corroborated by others)
categories Challenging to aggregate too many first-order (Friese, 2011)
To create third-order code categories, copy codes (thousands of codes) to second-order codes Enable “complex, exhaustive coding
and paste all codes and re-arrange them schemes” and “efficient and accurate
using color font with Excel retrieval” of coded data (Woods et al.,
2015b)
Code plotting and sharing #11. Select all second-order using cursor, click Produce artistic, high-quality images of Enhance “systematization, trustworthiness,
(cf. process 6 in Table II) on “right mouse” and press “Plot Selected networks of codes operational effectiveness” in data analysis
Code Categories” Can re-arrange network codes using various (Sinkovics et al., 2008)
network plot models (e.g. Facilitate the “visualization” of findings
Fruchterman–Reingold, Kamada–Kawai, etc. (Friese, 2011)
Theory building (cf. #12. Identify the relationships among themes/ This is a manual process that can only be done Enable “progressive focusing” via complex
process 7 in Table II) concepts/variables that emerged to create using human interpreter iteration between theory and data
a process model (Sinkovics and Alfoldi, 2012)
 installing R under different systems (i.e. Unix, Windows and OS X) are available online Methodology
(https://cran.r-project.org/doc/manuals/R-admin.html). Users must first install R, which is for qualitative
available at https://cran.r-project.org/bin/windows/base/, or RStudio, which is available at
www.rstudio.com/products/rstudio/download/, and choose the supported platforms (e.g.
research
Mac OS, Windows, Ubuntu). Once R or RStudio is successfully installed, the user needs to
launch the R or RStudio console and type the command (or syntax) for Windows or Linux to
launch RQDA (Huang, 2014):
install.packages [“RQDA”, dependencies⫽c(“Depends”, “Imports”)] install.packages 101
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

(“RQDA”, repos⫽“http://R-Forge.R-project.org”) require(RQDA) or for Mac OSX users:

install.packages(“RQDA”, type⫽“source”)
The installation of RQDA can be quite complex, especially for users with no experience
with R because it involves the installation of GTK⫹ and R and RQDA packages within R. To
address this problem, Huang (2014) offers an alternative for non-R users who use Windows
OS by providing an RQDA.bat file for download at (https://onedrive.live.com/?id⫽41C1EE
655922D297%21278&cid⫽41C1EE655922D297) or (http://pan.baidu.com/share/home?uk⫽
3691332575#category/type⫽0) to launch RQDA directly. For more details on installation,
see http://rqda.r-forge.r-project.org/.
Once RQDA is launched, users can click the “New Project” button to create a new project
and save the project file (.rqda file) in the right-hand directory. Next, all the files for analysis
(i.e. the five e-complaints and five e-praises files in .txt format) are imported to RQDA. Go to
the “Files” menu and click the “Import” button. Each case (e.g. in this study the “case” is an
electronics and computer company) may consist of multiple data files, and once all relevant
files are labeled (e.g. e-complaint file #1, #2, #3; e-praises file #1, #2, #3; or positive
interview #1, #2, #3; negative interview #1, #2, #3; news article #1, #2, #3) and imported
in RQDA, they form a single “hermeneutic unit” (Muhr, 1991) for a complete analysis. Once
all files are imported, users are ready to start coding the data. For cross-border collaborations
(Friese, 2011), users should use the same imported data files and share the coded outputs file
in a Dropbox or share the file via email.

Step 2: coding
One of the most important functions of RQDA is coding. RQDA allows inductive and
deductive coding; its operation is intuitive. Inductive coding or open coding, which is the
preferred option under the constructivist approach (Strauss and Corbin, 1990; Gioia et al.,
2013), allows a user to code a textual unit (e.g. paragraphs, sentences, words) that is close to
(i.e. the basic or lowest level) the data without being predicated on any theory, construct or
concept. For example, “they used my debit card to charge amounts I never authorized” is
coded as “monetary loss”. In deductive coding (positivistic approach; Fereday and
Muir-Cochrane, 2006; Bazeley and Jackson, 2013), categories are coded based on established
concepts, variables or theories. For example, if a researcher has pre-conceived theory that
“market orientation” (Jaworski and Kohli, 1993) will influence the presence or absence of
customer e-complaints, then he/she can code any UG reviews that contain e-complaints (e.g.
“the customer service department doesn’t work well with the technical support and that’s
why they took six months to solve my technical problems”) or e-praises (e.g. “this company
asked for my feedback and the services keeps getting better”) as evidence of the “market
orientation” construct.
When an imported .txt file is double-clicked, a new window appears on the upper
right-hand corner; under the Codes menu, the user then clicks the mark button to apply a new
code to a text passage (or unmark to cancel the code). Codes can be merged by right clicking
in the Codes menu and clicking Merge Selected with […]. In addition, users can also add
 QMR conditions (e.g. “with memo” or “with code category”) within the same menu. A sample of
20,1 coded text containing themes of an e-complaint of an electronics and computer product is
shown in Figure 1.
In the constructivist approach, one or more coders perform the coding task, while another
research team member(s) critiques the coded results and reconcile any differences in the
interpretation of the data (Gioia et al., 2013). This is a good practice to ensure reliability and
102 accuracy of data analysis. Coders need to pay attention to the process of “zooming in and
zooming out” in the coding process to avoid de-contextualizing the data (Sotiriadou et al.,
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

2014) also called “coding traps” (Gilbert, 2002) and remain reflexive of the influence of the
RQDA on data analysis (Woods et al., 2015a). Users need to constantly compare code labels
(Bazeley, 2007) and refine the code labels iteratively so the codes best represent the themes
that emerged [i.e. coding flexibility (Friese, 2011)]. These processes may be quite
time-consuming for novice users but can be efficient and effective with experience.

Step 3: data attributes

RQDA users assign attributes to each file. In fact, in most CAQDAS literature, scholars
recommend that coders add data attributes or memos to enrich understanding (Friese, 2011;
Paulus and Lester, 2016; Saillard, 2011). The “attribute” function basically enables users to
add variables to individual files to show a subset of files and to perform statistical analysis
on attributes. Attributes can be the type of files (e.g. interviews or news articles) or more
detailed information (e.g. informant #1 vs #10). To review the attributes of an individual file,
researchers can simply click the Attribute button in the File menu. In addition, researchers
can record their thoughts during the coding process by using the “memo” and “journal” tools,
which help them organize data and recall important information (e.g. “informants #2 and #6
expressed very negative feelings about their personal life, which may influence their online
reviews”; or “informants #1, #3, #5, and #6 are all engineers by training and therefore they
are more critical of the technical aspects of their online reviews”), in addition to helping them
aggregate data and build theory. It is important for RQDA users to always remind
themselves to ensure that the attributes and memos that are “all over the place” (Friese, 2011)
are captured and analyzed to avoid information loss.

Step 4: code abstraction

After the coding process is completed, first-level codes can then be aggregated into higher-level
code categories (called abstraction to second-level categories). To do this, a user clicks the Code
Categories menu, then clicks the Add button and types a second-level category label (e.g. “they
used my debit card to charge amounts that I never authorize” and “this faulty laptop means I
couldn’t work for four days” can be aggregated to a higher-level concept called “monetary loss”).
The user then places the cursor on a new second-level code category (e.g. “monetary loss”) and
moves the first-level codes into it by clicking all first-level codes that share similar meaning using
the Add To button. The user assigns all first-level codes to the second-level code categories until
no first-level code remains. This process is consistent with the constructivist theory-building
process (Strauss and Corbin, 1990; Gioia et al., 2013; Vaccaro and Palazzo, 2015). This process can
be iterative, and a user can amend first-level codes and their assignment to second-level
categories, if necessary.
The coding (Step 2) and code abstraction (Step 4) processes enables other researchers to
corroborate and/or check the results to see whether they arrive at the same coding and code
aggregation (i.e. analytical triangulation, Yin, 2009); it also facilitates communication and
sharing of the coding processes and coded outputs among researchers, and between
researchers and gatekeepers (e.g. journal reviewers, editors), among others. In this way, these
processes increase the transparency and credibility (Friese, 2011) of qualitative research.
 Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

of a coded text
A sample screenshot
Figure 1.
103
research
for qualitative
Methodology
 QMR Importantly, it facilitates complex and exhaustive coding schemes and efficient and accurate
20,1 retrieval of coded data (Woods et al., 2015a).
Figure 2 shows the 24 first-level categories or themes that were coded from the ten raw
texts imported to RQDA (see Step 2 above). Each of the coded categories can be amended by
clicking the Rename button or deleted by clicking the Delete button in the Codes menu.
Figure 3 illustrates how we reduced and abstracted the 24 first-level codes to a higher
104 level meaning by creating 11 second-order categories. As a research team, we interpreted,
discussed and then combined the first-level codes with similar meaning (i.e. “feeling
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

stressful”, “feeling sad”, “feeling powerless”, “feeling angry”, “feeling disappointed”) into a
new second-order category (i.e. “customer dissatisfaction”) using the Add To button until all
first-level codes were assigned to a second-level category. One of us acted as a “devil’s
advocate” to challenge/review the aggregated codes (Gioia et al., 2013).

Step 5: code plotting

When the code abstraction or aggregation is completed (as shown in Figure 3), the user plots
the coding results in two-dimensional outputs. This process allows users to visualize the
coding results and thus enhance the data interpretation and analysis process (i.e. visual
thinking; Arnheim, 1969). Another advantage of doing code plotting is that it produces
aesthetic and publication ready visuals. To do this, the user points the cursor to all (or
selected) second-level code categories shown in the console and right-clicking the mouse and
simultaneously pressing the Plot Selected Code Categories button (Figure 3). This process
automatically produces nodes and arrows in the form of networks, where each node
represents first-level codes connected to second-level codes at the network’s center.
The output can be modified using various network types (e.g. the Fruchterman-Reingold
or Kamada Kawai layout) and is visually appealing. This code visualization process (Friese,
2011) enhances understanding and communication between researchers. In the language of
the constructivist approach, during this code abstraction process, users can transition from
an inductive to abductive theorizing process by moving back and forth between the emergent
findings and the literature/theories to find new concepts within the data (Alvesson and
Kärreman, 2007). A sample of first- and second-level category plots of the coded texts are
shown in Figure 4.
RQDA coding outputs can be exported as HTML files, rendering the analyses more
accessible and enhancing the trustworthiness, transparency and credibility (Sinkovics et al.,
2008; Woods et al., 2015a, 2015b) of the qualitative data analysis. In contrast to a
platform-specific file format, which can be cumbersome to access and read, HTML files that
show all first- and second-level codes can be shared with readers, particularly reviewers and
editors, as well as other users and the open source community. To do this, the user applies the
very helpful Export function. RQDA project data, which consists of codes, categories,
attributes, memos and other information, are stored in .rqda files. Users can export the .rqda
files to other data analysis software such as SPSS or the R program for further statistical
analysis. Figure 5 offers a small part of first- and second-level codes in an HTML file.
RQDA uses colors to highlight different codes; researchers can retrieve all coding under
individual codes at any time. This function allows researchers to organize and keep track of
many “messy” texts using certain color as the common denominator. Users simply double
click the code, and a new window showing all coding appears. This allows the user to see all
codes that were marked with a particular color or meaning. RQDA also enables conditional
retrieval. For example, users can opt to retrieve codes with or without a coding category. In
Figure 6, we demonstrate the selective retrieval of two second-level categories (i.e. customer
dissatisfaction and communication problems) to zoom in on specific second-level codes.
 Methodology
for qualitative
research

105
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Figure 2.
The first-level
categories or themes
coded from all coded
data (24 codes)
 Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

20,1

106
QMR

categories
Figure 3.

second-level code
code categories to
Abstracting first-level
 Methodology
for qualitative
research

107
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Figure 4.
A sample plot of
first-level and
second-level code
categories

Another powerful feature of RQDA is that users can perform complex queries (e.g. Boolean
searches) on coded data by typing syntax or commands in the R window. (For users who are
familiar with and have SQL on their desktop, they can also use SQL to perform complex
queries. SQL is designed for data management and is powerful in structured data computing,
and can therefore supplement RQDA. For users who are well-versed in SQL and “think” in
SQL but want to use R for additional statistical analysis or data visualization, there are two
 QMR
20,1

108
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Figure 5.
An HTML file
containing a few coded
categories for sharing
among researchers

useful R packages that optimized for convenience, which are sqldf and plyr (Anderson, 2012;
Grothendieck, 2014). While this function appears trivial in the case of a few texts, it is
effective and efficient with hundreds or thousands of texts. For example, if a user wants to
see the coding for code 1 or code 2, he/she can enter the command RQDAQuery (“select seltext
from coding where cid⫽1 or cid⫽2” ) in the R window (Huang, 2014). This function helps
qualitative researchers further manage data and complex operations and meets the needs of
various users (e.g. doing small N or large N studies) and project types (e.g. purely qualitative,
mixed method research).

Step 6: theory building

Finally, we iteratively re-categorized all first- and second-level code categories in a “data
structure” format (Gioia et al., 2013; Vaccaro and Palazzo, 2015) to refine codes to the best
possible higher-level categories that describe and explain the empirical reality and the
literature on e-complaints. After three iterations, we developed a data structure as shown in
Figure 7. As shown in Figure 7, we provided two of the most representative quotes from each
of the 24 first-level code categories created (i.e. company arrogance, booking problems,
complicated diagnosis, helpful service, patient service, friendly service, customer’s lack of
technical knowledge, feeling angry, feeling disappointed, feeling powerless, feeling sad, feeling
stressful, great bargain, great service recovery, hardware malfunction, loss of productivity,
miscommunication, misinformed, not caring about customers, service rep spoke too fast, slow
response, sue the company, unauthorized charging and vendor mistakes). We then abstracted
them to 11 second-level codes (i.e. communication problems, customer dissatisfaction,
excellent service recovery, information asymmetry, love the bargains, malfunctions,
monetary loss, negative image for company, poor customer service, quality customer service
and vendor problems).
Next, we developed five aggregate dimensions that help us explain the why’s and how’s
of e-complaints in the electronics and computer sector: service quality, coordination, financial,
technical and responsibility orientations. “Service quality” orientation refers to customers’
attention to the service quality provided by companies. While service quality and its
 Methodology
for qualitative
research

109
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Figure 6.
A selected plot view of
two second-level code
categories

dimensions (e.g. reliability, responsiveness, empathy) have been studied for many years in
marketing and psychology (Bolton and Drew, 1991; Parasuraman et al., 1985), the service
quality orientation refers to the service quality perceived by customers, or a customer-centric
side of the attention-based view theory (Chandra, 2016; Ocasio, 1997), which predicts the
number of e-complaints and e-praises.
“Coordination” orientation refers to customers’ attention to firm’s ability to
coordinate its supply chain and service delivery units (external to internal and
vice versa). This orientation extends the buyer–vendor coordination literature in
marketing and operations management (Huq et al., 2016; Sarmah et al., 2006; Schmid
et al., 2016) by highlighting the importance of customer perception of a firm’s service
value chain as key to their e-complaints or e-praises.
 QMR First-level code categories Second-level code categories Aggregate theoretical dimensions

20,1 G1. This company is the worst they take money but have NO
interest at all in their customers.
Poor customer service
G2. I have been trying to talk to them about my problem for 6
months, they just stop answering

B1. I am retired and 70 years old and they have increased my

stress level immeasurably!
Customer dissatisfaction
B2. I too plan on writing the CEO but it seems that will be a waste
110 of time as well. But what can we do?
Service quality orientation
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

I1. They replaced it even though we had gone past the 14 day
dateline. They refunded us in full – they just want us to be happy
Excellent service recovery
I2. A simple apology would have worked but they went above and
beyond to ensure customer satisfaction.

K1. She was patient to wait on the phone for my phone to restore
which took 25 mins, checking every 3 minutes how its going. Quality customer service
K2. the service was very painless and friendly.

C1. You as the ‘experts’ should have advised that photos are not
automatically saved, must be backed up before we change over
Information asymmetry
C2. They provide this service and assured me that if I moved up
to the new phone, my call dropping would disappear. It doesn’t

A1. This isn't booking. To book means to have a date to bring it in

for attention. I was left without a computer for 4 days.
Communication problems Coordination orientation
A2. You have to get through like 15 computer questions and stuff
to get to Costumer Support

H1. I sent a letter of distress to (their reseller) and received no

response.
Vendor problems
H2. Your connection to (a telco) will certainly continue to erode
the confidence of the your product

E1. They used my debit card to charge amounts that I never

authorized Monetary loss
E2. This means surrendering it for up to 4 days. What an outrage!
Financial orientation

J1. They seem to have set up a good bargain nonetheless, $310

flat rate and they will return the computer to original working Love the bargains
condition. Great deal, it was done in one day

D1. Guess what!! No ring tone or alert when receiving a text.

Malfunctions Technical orientation
D2. Graphics processing unit (video card) died on my device.

F1. They get away with treating customers like dirt because they
thinks they are immune from any danger from irate customers.
Negative Image Responsibility orientation
F2. I am about to file a fraud charge against them with my bank.
Figure 7.
A data structure of
customer e-complaints Note: This ‘data structure’ was drawn using Microsoft Visio for optimal visualization. The
in the electronics and non-shaded boxes are quotes and codes from e-complaint samples; the shaded boxes are
computer sector quotes and codes from e-praises samples

“Financial” orientation refers to customers’ attention to financial gains or losses in their

interactions with a technology provider, or what the services marketing literature calls
“distributive justice” (Liao, 2007; Mattila, 2001; Roschk and Gelbrich, 2014), which
emphasizes the importance of customers’ perceived fairness of the complaint outcomes
(e.g. that a firm repairs or replaces a faulty product or helps a customer use a product).
This demonstration also suggests that customers attach great importance to financial
incentives (e.g. money back guarantee, product replacement, discounts), following a
service failure.
“Technical” orientation refers to customers’ attention to technical matters including
technical errors when making e-complaints and e-praises. Marketing and service
 management scholars have studied various service recovery procedures (Dabholkar and Methodology
Spaid, 2012; Oliveira and Roth, 2012) and highlight the negative influence of technology error for qualitative
(e.g. faulty products) on customer satisfaction. Our demonstration reveals the importance of
the firm’s technical capabilities as a predictor of e-complaints or e-praises.
research
Finally, “responsibility” orientation refers to customers’ attention to the firm’s sense of
business responsibility (Garriga and Melé, 2004), particularly delivering its promises and
treating customers with respect and fairness. It may also reveal and reflect customers’
expectation about service recovery as a part of firm’s corporate social responsibility, which 111
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

has gained popularity in the past decade (Choi and La, 2013; Joireman et al., 2015).
In “reaching closure” as per the constructivist approach, we went back and forth between
our coded data and the emergent literature on e-complaints (Alvesson and Kärreman, 2007;
Gioia et al., 2013) and saw an opportunity to extend the e-complaint literature by positioning
our process model as a customer e-complaint rhetoric model (Figure 8).
As shown in Figure 8, we theorize that e-complaints and e-praises are influenced and can
be predicted by five customer rhetoric orientations: service quality, coordination, financial,
technical and responsibility. We found that customers used e-complaints as a means of
action to right a wrong; an e-complaint is a verbal act against the firm (e.g. publicly
condemning the firm), and a means of coordination with others (e.g. persuading others to
report the complaints to local news stations), and a means of thoughtful action to reflect
about the world (e.g. What went wrong with this experience with the firm?). These three
actions are parallel to the use of language as a social action, coordinated action and thoughtful
action, respectively, according to the speech act theory (Austin, 1975; Chandra, 2016;
Holtgraves, 2013). The verbalized thoughts in the form of e-comments or e-praises are aimed
to change the e-commenter’s (i.e. the customer’s) as well as potential customers’
psychological and behavioral state (e.g. dissatisfaction, switch to other firms and/or spread
negative reviews of the firm/product, punish the offending firm by drawing customers
away). Ultimately, they can affect the firm’s performance (i.e. bottom line). The e-complaints,
which could extend to customers launching negative campaigns or taking legal action
against a firm, can also influence the firm’s performance, i.e. its bottom line. In this
demonstration, we simplified the theorizing process leading to the process model (Figure 8);
the actual theorizing process for a complete qualitative paper is usually much more elaborate
and explains how the process model enriches and extends the literature.

Service quality orientation

Coordination orientation
Customers’ psychological and
behavioral change
E-complaints or e -praises
as social action
Financial orientation
as coordinated action
as thoughtful action
Firm performance
Figure 8.
Technical orientation
A process model of
customer e-complaint
rhetoric in the
electronics and
Responsibility orientation
computer sector
 QMR Discussion
20,1 In the following, we reflect a few key insights gained from using the RQDA-based
constructivist qualitative research in this study’s demonstration and our accumulated
experience using RQDA since 2013 in research and teaching university-level research
methodology classes. We are also experienced users of NVivo and ATLAS.ti, which qualify
us to evaluate the relative merits and shortfalls of RQDA and to compare it with these two
112 CAQDAS tools.
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Advantages of RQDA-based constructivist qualitative research

Two of RQDA’s main advantages are that it is free and compatible with various operating
systems (OS) including Windows, Mac and Linux. Based on our demonstration (Table II),
RQDA seems to fare well in terms of OS compatibility compared to NVivo and ATLAS.ti. In
the age of budget-cutting, universities, corporations and nonprofits can benefit from free
CAQDAS software such as RQDA. Existing R users can easily “cross over” to RQDA to
analyze textual data using RQDA as a primary or secondary analytical tool, and, in our view,
they will be the early adopters of RQDA. Computer-savvy faculty, graduate students and
independent researchers doing CAQDAS-based qualitative research will find RQDA a
valuable new methodological tool. Although RQDA does not come with any official technical
support from its developer, we have never found any bugs or faced any technical failures
during our four years of experience with RQDA. However, we have maintained good
communication with RQDA developers to ensure that our knowledge of RQDA is current.
Additionally, we have found that RQDA’s capabilities meet most qualitative researchers’
needs and that its capabilities are comparable to fee-based CAQDAS software such as
ATLAS.ti and NVivo for coding, attributes addition and code abstraction (Table I). RQDA
use is simple and intuitive and therefore suitable for non-programmers and novice
researchers, as the package uses a very straightforward “point, touch and click” (GUI-based)
window. Extensive knowledge of R is not necessary to operate RQDA. However, basic
programming skills using syntax in R can be helpful in launching RQDA. Based on our
experience, one of RQDA’s most powerful features is its code plotting capability, which offers
elegant and publication-ready images of node networks that no other CAQDAS tools offer.
Moreover, RQDA also enables users to analyze data in different languages (nine languages,
total), thus facilitating cross-cultural research, which can enhance various disciplines
including linguistics, cross-cultural studies, psychology, marketing and organizational and
managerial sciences.
Furthermore, RQDA users can use R functions to import batches of files (e.g. hundreds or
thousands) using a simple syntax. RQDA is supported by SQLite (a relational database
management system; James and Falcon, 2011), and thus its functionality can be enhanced by
writing R commands. Because RQDA seamlessly integrates with R, it enables statistical
analysis of the coding and additional functions for data manipulation and analysis. NVivo
and ATLAS.ti also can record attribute data in the form of categorical or ordinal variables
(e.g. male/female, high/medium/low performance), which can be manually inserted into in
the CAQDAS software or imported from a spreadsheet, which allows further statistical
analysis [e.g. cluster or correspondence analysis (Bazeley, 2002, 2007; Friese, 2011)].
However, in our literature review, we found that very few researchers actually used the
quantitative applications of NVivo or ATLAS.ti, and that qualitative data analysis was the
primary application used.
The uniqueness of RQDA is that it does not require the support of external applications
(e.g. Excel spreadsheets or SPSS) to perform statistical analysis, which are needed for NVivo
or ATLAS.ti (Bazeley, 2007; Friese, 2011). Rather, RQDA relies on more than 5,000
 application packages available in the R platform to conduct statistical analysis on textual Methodology
data or advanced statistical analysis. For example, Ye (2010) combined RQDA with tm, an R for qualitative
text-mining package that uses statistical probability techniques to categorize large bodies of
words. It may also be possible to combine RQDA and topic modeling, a new computational
research
technique that seeks to extract themes from large bodies of textual data using statistical
techniques (Blei, 2012).
Finally, we have used the above demonstration in our methodology courses at the
undergraduate, master’s and PhD levels for the past three years. We have found that 113
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

students with no experience in CAQDAS or qualitative research became competent users of

the RQDA-based constructivist methodology in just nine-hour lessons. However, most of
them required additional reading of specific substantive domain literature (e.g. services
marketing, e-marketing or technology management literature) to proficiently transform the
data structure into a “process model”. But no students experienced any technical challenges
in operating and using RQDA. We argue that the current computer-savvy young people are
well-equipped to use a light-weight programming platform such as RQDA with proper
instruction, and no prior experience is necessary to learn RQDA.

Constraints and opportunities for RQDA-based constructivist qualitative research

We found that RQDA has certain shortcomings. First, RQDA supports only plain text (.txt)
files. It neither does support multimedia (audio, videos) files, spreadsheets, tables, photos or
other non-plain text formats nor does it support transcription of audio and video files, which
are supported by NVivo and ATLAS.ti. Therefore, NVivo may be a better choice for
ethnographers and visual methodologists who work with a large number of photo and video
data. Future refinements of RQDA could accommodate these types of files.
Another shortfall of RQDA is that the code-category relationships can only be established
for one hierarchy – from the first- to second-level code category. That is, it does not provide
a complex tree-like category structure to organize concepts as do NVivo and ATLAS.ti
(Bazeley, 2007; Friese, 2011). The original purpose of RQDA was to simplify data
aggregation structures and to help researchers avoid creating too many layers of data
abstraction. As Huang (2014) argues, the tree structure may be too cumbersome to describe
relationships between abstract concepts or to identify theories based on text patterns.
Moreover, if a project has too many codes and categories, the plot map can become quite
messy and indecipherable. However, based on our own experimentation and experience
using RQDA in large-scale projects (e.g. developing a process model of entrepreneurial
emancipation using over 50 interviews as data; studying the narratives of the emergence of
social enterprises using 200 biographical profiles) in the past four years, users can overcome
this limitation by exporting codes and categories using RQDA to other applications, such as
spreadsheets (e.g. Excel) to create further “aggregate” coding hierarchies and develop “data
structures” as per the constructivist approach. Therefore, there are rooms for improvisation
and bricolage in using RQDA.
Moreover, because RQDA is a package within R, users must have some basic knowledge
of R language. Based on our experience in using and teaching RQDA (and R) in
undergraduate and graduate methods courses, users can use some “shortcuts” to simplify
the syntax use in RQDA. For instance, the R syntax required for a RQDA launch can be saved
in a “notepad” (non-programmable) file, which can be simply copied and pasted directly to
the R console. In this way, a user can launch RQDA very efficiently and without worrying
about making a typo when typing in R syntax (which often occurs among novice users). For
more complex statistical analysis of texts, a substantial amount of time is needed to learn the
R language and other relevant R packages. This is both a constraint and an opportunity.
 QMR Implications
20,1 Our methodological demonstration offers at least two important implications to the
discourse and literature in qualitative methodology. First, we unpacked the ontological and
epistemological roots of two main paradigms in qualitative research – positivism and
constructivism – and proposed the RQDA-based constructivist qualitative research
methodology and the step-by-step procedures to transform a research question and
114 qualitative data into a process model. Our proposed approach leverages the strengths of
constructivism as a lens to study social reality and the relatively new open source CAQDAS
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

software called RQDA. It will enable future qualitative researchers to conduct research with
rigor, transparency and credibility. This fills in the “boilerplate”, operationalization and
transparency gaps in the empirical and review papers on qualitative research methodology.
Second, we offered an in-depth and critical review of RQDA, an R extension for qualitative
data analysis, and grounded it within the broader CAQDAS literature and how it fared with
two popular CAQDAS software, NVivo and Atlas.ti. By doing so, we provided the first
official review of RQDA as a new methodological arsenal for qualitative scholars. Our review
of the RQDA also provides a theoretically informed debate about the role of CAQDAS and
how it can advance qualitative research scholarship. This fills in the theoretical gap about
RQDA as a new CAQDAS tool and answers some of the methodological and practical
questions behind it.

Conclusion and implications for future research

Qualitative research faces legitimacy, contestation and transparency problem. It “lacks a
boilerplate” (Pratt, 2009) in qualitative data assessment and presentation. Thus, it often fails
to achieve the status of other forms of research, such as quantitative (variance-based,
experiment-based, simulation-based and mathematics-based) research (Van de Ven and
Engleman, 2004; Yang and Chandra, 2013); moreover, qualitative research is not as well
published as the quantitative research in top-tier marketing, psychology and management to
political sciences journals. The purpose of this review and demonstration is to offer a means
to enhance the transparency, validity and replicability of qualitative research by integrating
a new CAQDAS tool called RQDA and the constructivist approach in qualitative research,
which we called RQDA-based constructivist qualitative research. Despite the promise of
RQDA to enhance the constructivist approach, there is a dearth of literature and “how to”
guide on RQDA (ours is among the first). The very few published journal articles that cite
RQDA tend to use it to perform only basic data analysis (Schwarzkopf et al., 2013;
Vanwindekens et al., 2013; Lamprinakis, 2012).
In this article, we propose and demonstrate how the RQDA-based constructivist approach
can be implemented to develop a process model (as a means to make new discoveries) and
how it adds a valuable tool to qualitative scholars’ methodological toolbox. We propose that
the constructivist approach be used as a methodological prototype for future
CAQDAS-based research. As we demonstrated, the power of the constructivist approach lies
in how it derives “new concepts” and provides a “data structure” and a “process model”,
which offers a systematic and transparent means to presenting qualitative research. In this
way, the constructivist approach increases the legitimacy of CAQDAS-based qualitative
research. Therefore, our study provides a new avenue for qualitative scholars to further
integrate other CAQDAS software (e.g. MaxQDA, NVivo, ATLAS.ti) with the constructivist
approach.
Our RQDA-based constructivist methodology can be applied to various types of
qualitative research, from grounded theory, ethnography, discourse analysis, to the verbal
autopsy method, among others (Leech and Onwuegbuzie, 2011; Woods et al., 2015a). Going
 forward, RQDA developers can consider extending its capabilities to handle multimedia files Methodology
so it can analyze a broader set of social-economic-political phenomena. This is important for qualitative
given the rise of social media (e.g. Facebook, Twitter, YouTube, Instagram and cameras in
smartphones and other handheld electronics) globally. One potential but under-used
research
capability of RQDA is its support of netnography research, participant-observational
research that uses computer-mediated communications to study the behavior of members of
an online group (Kozinets et al., 2010) and visual sociology research, where observations of
115
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

human and their social behavior relies on photos and videos capture (Harper, 2012). With a
little inventiveness or “bricolage” [making use of what one has – existing tools, materials,
skills (Denzin and Lincoln, 1994)], one can conduct netnography using RQDA and other
supporting applications such as spreadsheets or drawing tools (e.g. Microsoft Visio or
draw.io).
Future research can also explore the integration of RQDA and the positivist approach to
qualitative research (i.e. the deductive-coding approach), an approach that supports theory
testing, refinement, replication and elimination of alternative explanations. RQDA enables
users to perform simple and advanced statistical analysis on coded materials, and this offers
a new avenue to extend and enrich the current Eisenhardt approach. In particular, when
users are analyzing case studies with “embedded designs” [i.e. multiple observations per
each unit of analysis (Yin, 2003)], and if there are hundreds or thousands of observations to
analyze in a study, then an RQDA-based Eisenhardt approach is a good option.
Another opportunity is to combine the inductive RQDA-based constructivist approach
and quantitative content analysis (Krippendorff, 2013). This requires the researcher to
inductively code qualitative data and derive new concepts to create a coding scheme with
quantitative measures (e.g. metric, ordinal or categorical); independent coders then code the
same (and or additional) data sets to quantitatively confirm (or disconfirm) the tentative
qualitative findings. In a rather similar fashion, a recent trend among content analysts
involves using computer-aided text analysis (CATA) tools that transform words into
numbers for statistical analysis [e.g. multiple regression or analysis of variance using
Linguistic Inquiry and Word Count, Chandra, 2016; or DICTION software (Short and Palmer,
2007)]. This presents a new opportunity to combine CATA with CAQDAS, particularly
RQDA, to offer a more holistic analysis of large corpuses of textual data.
RQDA offers researchers with some knowledge of statistical programming in R an
effective and innovative means to conduct mixed method and qualitative research under a
single R platform. It offers a new avenue for existing R users to “cross over” to analyzing
textual data as a primary or secondary research tool. Likewise, it offers computer-savvy
scholars, students and independent researchers an opportunity to join the “CAQDAS
movement” and to carry out and publish high-quality qualitative research. We believe that
the time is ripe for a wider adoption of RQDA either as a stand-alone tool or in combination
with other analytical tool(s).

Note
1. We conducted a systematic literature search using a combination of key words “qualitative”,
“qualitative method*”, “positivis*” and “interprevis*” via Google Scholar, JSTOR, Web of Science
and direct search on 10 marketing journals which are Journal of Marketing, Journal of Marketing
Research, Journal of Consumer Research, Journal of Academy of Marketing Science, Marketing
Science, Marketing Letters, Journal of Business Research, Journal of Marketing Management,
Journal of Strategic Marketing and Industrial Marketing Management. Only 90 articles from these
sources are directly relevant to the analysis, and we used the insights derived from these journal
articles to form our understanding of the nature of qualitative research in the field.
 QMR References
20,1 Altman, D.G. and Moher, D. (2013), “Declaration of transparency for each research article”, British
Medical Journal, Vol. 347, p. f4796, available at: www.bmj.com/content/347/bmj.f4796.short
Alvesson, M. and Kärreman, D. (2007), “Constructing mystery: empirical matters in theory
development”, Academy of Management Review, Vol. 32 No. 4, pp. 1265-1281.
Anderson, S. (2012), “A quick introduction to plyr”, available at: http://seananderson.ca/courses/12-
116 plyr/plyr_2012.pdf (accessed 7 November 2016).
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Arnheim, R. (1969), Visual Thinking, University of California Press, London.

Arnould, E.J. and Wallendorf, M. (1994), “Market-oriented ethnography: interpretation building and
marketing strategy formulation”, Journal of Marketing Research, Vol. 31 No. 4, pp. 484-504.
Austin, J.L. (1975), How to Do Things with Words, Oxford University Press, Oxford.
Basit, T. (2003), “Manual or electronic? The role of coding in qualitative data analysis”, Educational
Research, Vol. 45 No. 2, pp. 143-154.
Bazeley, P. (2002), “The evolution of a project involving an integrated analysis of structured qualitative
and quantitative data: from N3 to Nvivo”, International Journal of Social Research Methodology,
Vol. 5 No. 3, pp. 229-243.
Bazeley, P. (2007), Qualitative Data Analysis with NVivo, Sage, London.
Bazeley, P. and Jackson, K. (Eds) ( 2013), Qualitative Data Analysis with NVivo, Sage Publications,
London.
Belk, R.W. (Ed.) ( 2007), Handbook of Qualitative Research Methods in Marketing, Edward Elgar
Publishing, Cheltenham, Gloucestershire.
Belk, R.W., Ger, G. and Askegaard, S. (2003), “The fire of desire: a multisited inquiry into consumer
passion”, Journal of Consumer Research, Vol. 30 No. 3, pp. 326-351.
Blei, D.M. (2012), “Probabilistic topic models”, Communications of the ACM, Vol. 55 No. 4, pp. 55-65.
Bolton, R.N. and Drew, J.H. (1991), “A multistage model of customers’ assessments of service quality
and value”, Journal of Consumer Research, Vol. 17 No. 4, pp. 375-384.
Buhalis, D. and Law, R. (2008), “Progress in information technology and tourism management: 20 years
on and 10 years after the internet – the state of eTourism research”, Tourism Management,
Vol. 29 No. 4, pp. 609-623.
Chandra, Y. (2016), “A rhetoric-orientation view of social entrepreneurship”, Social Enterprise Journal,
Vol. 12 No. 2, pp. 161-200.
Chandra, Y., Jiang, L.C. and Wang, C.J. (2016), “Mining social entrepreneurship strategies using topic
modeling”, PLoS One, Vol. 11 No. 3, pp. e0151342.
Chen, Y., Fay, S. and Wang, Q. (2011), “The role of marketing in social media: how online consumer
reviews evolve”, Journal of Interactive Marketing, Vol. 25 No. 2, pp. 85-94.
Chevalier, J.A. and Mayzlin, D. (2006), “The effect of word of mouth on sales: online book reviews”,
Journal of Marketing Research, Vol. 43 No. 3, pp. 345-354.
Choi, B. and La, S. (2013), “The impact of corporate social responsibility (CSR) and customer trust on the
restoration of loyalty after service failure and recovery”, Journal of Services Marketing, Vol. 27
No. 3, pp. 223-233.
Choi, N. and Majumdar, S. (2014), “Social entrepreneurship as an essentially contested concept: opening
a new avenue for systematic future research”, Journal of Business Venturing, Vol. 29 No. 3,
pp. 363-376.
Commuri, S. (2009), “The impact of counterfeiting on genuine-item consumers’ brand relationships”,
Journal of Marketing, Vol. 73 No. 3, pp. 86-98.
Corbyn, Z. (2012), “Misconduct is the main cause of life-sciences retractions”, Nature, Vol. 490 No. 7418,
p. 21.
 Cornelissen, J.P. (2016), “Preserving theoretical divergence in management research: why the Methodology
explanatory potential of qualitative research should be harnessed rather than suppressed”,
Journal of Management Studies, forthcoming.
for qualitative
Coviello, N. (2014), “How to publish qualitative entrepreneurship research in top journals”, in Fayolle, A.
research
and Wrightm, M. (Eds), How to Get Published in the Best Entrepreneurship Journals: A Guide to
Steer Your Academic Career, Edward Elgar, Cheltenham, pp. 167-182.
Coviello, N.E. and Joseph, R.M. (2012), “Creating major innovations with customers: insights from small
117
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

and young technology firms”, Journal of Marketing, Vol. 76 No. 6, pp. 87-104.
Dabholkar, P.A. and Spaid, B.I. (2012), “Service failure and recovery in using technology-based
self-service: effects on user attributions and satisfaction”, The Service Industries Journal, Vol. 32
No. 9, pp. 1415-1432.
Daniel, S. and Stewart, K. (2016), “Open source project success: resource access, flow, and integration”,
The Journal of Strategic Information Systems, Vol. 25 No. 3, pp. 159-176.
Davis, N.W. and Meyer, B.B. (2009), “Qualitative data analysis: a procedural comparison”, Journal of
Applied Sport Psychology, Vol. 21 No. 1, pp. 116-124.
Denzin, N.K. and Lincoln, Y.S. (Eds) (1994), Handbook of Qualitative Research, Sage Publications,
Thousand Oaks, CA, pp. 1-32.
Denzin, N.K. and Lincoln, Y.S. (2005), “Introduction: the discipline and practice of qualitative research”,
in Denzin, N.K. and Lincoln, Y.S. (Eds), The Sage Handbook of Qualitative Research, 3rd ed.,
Sage, Thousand Oaks, CA, pp. 1-32.
Dubé, L. and Paré, G. (2003), “Rigor in information systems positivist case research: current practices,
trends, and recommendations”, MIS Quarterly, Vol. 27 No. 4, pp. 597-636.
Easton, G. (2010), “Critical realism in case study research”, Industrial Marketing Management, Vol. 39
No. 1, pp. 118-128.
Eisenhardt, K.M. (1989a), “Making fast strategic decisions in high-velocity environments”, Academy of
Management Journal, Vol. 32 No. 3, pp. 543-576.
Eisenhardt, K.M. (1989b), “Building theories from case study research”, Academy of Management
Review, Vol. 14 No. 4, pp. 532-550.
Eisenhardt, K.M. and Graebner, M.E. (2007), “Theory building from cases: opportunities and
challenges”, Academy of Management Journal, Vol. 50 No. 1, pp. 25-32.
Ekiz, E., Khoo-Lattimore, C. and Memarzadeh, F. (2012), “Air the anger: investigating online complaints
on luxury hotels”, Journal of Hospitality and Tourism Technology, Vol. 3 No. 2, pp. 96-106.
Fang, F.C., Steen, R.G. and Casadevall, A. (2012), “Misconduct accounts for the majority of retracted
scientific publications”, Proceedings of the National Academy of Sciences, Vol. 109 No. 42,
pp. 17028-17033.
Fereday, J. and Muir-Cochrane, E. (2006), “Demonstrating rigor using thematic analysis: a hybrid
approach of inductive and deductive coding and theme development”, International Journal of
Qualitative Methods, Vol. 5 No. 1, pp. 80-92.
Franzosi, R., Doyle, S., McClelland, L.E., Putnam Rankin, C. and Vicari, S. (2013), “Quantitative narrative
analysis software options compared: PC-ACE and CAQDAS (ATLAS. Ti, MAXqda, and
NVivo)”, Quality & Quantity, Vol. 47 No. 6, pp. 3219-3247.
Friese, S. (2011), “Using ATLAS.ti for analyzing the financial crisis data”, Forum Qualitative
Sozialforschung/Forum: Qualitative Social Research, Vol. 12 No. 1.
Gallie, W.B. (1955), “Essentially contested concepts”, Proceedings of the Aristotelian Society,
Aristotelian Society, Wiley, London, Vol. 56, pp. 167-198.
Gallie, W.B. (1956), “Art as an essentially contested concept”, The Philosophical Quarterly (1950), Vol. 6
No. 23, pp. 97-114.
 QMR García-Horta, J.B. and Guerra-Ramos, M.T. (2009), “The use of CAQDAS in educational research: some
advantages, limitations and potential risks”, International Journal of Research & Method in
20,1 Education, Vol. 32 No. 2, pp. 151-165.
Garriga, E. and Melé, D. (2004), “Corporate social responsibility theories: mapping the territory”, Journal
of Business Ethics, Vol. 53 Nos 1/2, pp. 51-71.
Gebhardt, G.F., Carpenter, G.S. and Sherry, J.F., Jr. (2006), “Creating a market orientation: a longitudinal,
118 multifirm, grounded analysis of cultural transformation”, Journal of Marketing, Vol. 70 No. 4,
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

pp. 37-55.
Gewin, V. (2014), “Retractions: a clean slate”, Nature, Vol. 507 No. 7492, pp. 389-391.
Giesler, M. and Thompson, C.J. (2016), “Process theorization in cultural consumer research”, Journal of
Consumer Research, forthcoming.
Gilbert, L.S. (2002), “Going the distance: ‘closeness’ in qualitative data analysis software”, International
Journal of Social Research Methodology, Vol. 5 No. 3, pp. 215-228.
Gioia, D.A. and Chittipeddi, K. (1991), “Sensemaking and sensegiving in strategic change initiation”,
Strategic Management Journal, Vol. 12 No. 6, pp. 433-448.
Gioia, D.A., Corley, K.G. and Hamilton, A.L. (2013), “Seeking qualitative rigor in inductive research:
notes on the Gioia methodology”, Organizational Research Methods, Vol. 16 No. 1, pp. 15-31.
Goulding, C. (2005), “Grounded theory, ethnography and phenomenology”, European Journal of
Marketing, Vol. 39 Nos 3/4, pp. 294-308.
Grayson, K. and Shulman, D. (2000), “Indexicality and the verification function of irreplaceable
possessions: a semiotic analysis”, Journal of Consumer Research, Vol. 27 No. 1, pp. 17-30.
Grothendieck, G. (2014), “sqldf: perform SQL selects on R data frames”, available at: http://CRAN.R-
project.org/package⫽sqldf (accessed 7 November 2016).
Hanson, D. and Grimmer, M. (2007), “The mix of qualitative and quantitative research in major
marketing journals, 1993-2002”, European Journal of Marketing, Vol. 41 Nos 1/2, pp. 58-70.
Harper, D. (2012), Visual Sociology, Routledge, Oxon.
Holtgraves, T.M. (2013), Language as Social Action: Social Psychology and Language Use, Psychology
Press, Mahwah, NJ.
Homburg, C., Wilczek, H. and Hahn, A. (2014), “Looking beyond the horizon: how to approach the
customers’ customers in business-to-business markets”, Journal of Marketing, Vol. 78 No. 5,
pp. 58-77.
Huang, R.G. (2014), “RQDA: r-based qualitative data analysis: R package version 0.2-7”, available at:
http://rqda.r-forge.r-project.org (accessed: 18 November 2015).
Hunt, S.D. (1991), “Positivism and paradigm dominance in consumer research: toward critical pluralism
and rapprochement”, Journal of Consumer Research, Vol. 18 No. 1, pp. 32-44.
Hunt, S.D. (1994), “On rethinking marketing: our discipline, our practice, our methods”, European
Journal of Marketing, Vol. 28 No. 3, pp. 13-25.
Huq, F.A., Chowdhury, I.N. and Klassen, R.D. (2016), “Social management capabilities of multinational
buying firms and their emerging market suppliers: an exploratory study of the clothing
industry”, Journal of Operations Management, Vol. 49 (September), pp. 19-37.
Ihaka, R. and Gentleman, R. (1996), “R: a language for data analysis and graphics”, Journal of
Computational and Graphical Statistics, Vol. 5 No. 3, pp. 299-314.
Jackson, J. (2014), “R programming language gaining ground on traditional statistics packages”,
available at: www.pcworld.com/article/2480920/r-programming-language-gaining-ground-on-
traditional-statistics-packages.html (accessed 30 October 2015).
James, D.A. and Falcon, S. (2011), RSQLite: SQLite interface for R. R package version 0.11, 1, available
at: http://cran.r-project.org/web/packages/RSQLite
 Jaworski, B.J. and Kohli, A.K. (1993), “Market orientation: antecedents and consequences”, The Journal Methodology
of Marketing, Vol. 57 No. 3, pp. 53-70.
for qualitative
Joireman, J., Smith, D., Liu, R.L. and Arthurs, J. (2015), “It’s all good: corporate social responsibility
reduces negative and promotes positive responses to service failures among value-aligned
research
customers”, Journal of Public Policy & Marketing, Vol. 34 No. 1, pp. 32-49.
Karabag, S.F. and Berggren, C. (2016), “Misconduct, marginality and editorial practices in management,
business and economics journals”, PloS One, Vol. 11 No. 7, e0159492.
119
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Kozinets, R.V. (2002), “The field behind the screen: using netnography for marketing research in online
communities”, Journal of Marketing Research, Vol. 39 No. 1, pp. 61-72.
Kozinets, R.V., Wilner, S., Wojnicki, A. and de Valck, K. (2010), “Networks of narrativity: understanding
word-of-mouth marketing in online communities”, Journal of Marketing, Vol. 74 No. 2, pp. 71-89.
Krippendorff, K. (2013), Content Analysis: An Introduction to its Methodology, Sage, California.
Lamprinakis, L. (2012), “Organizational innovation and institutional change: the case of valio in
Finland”, International Journal on Food System Dynamics, Vol. 3 No. 2, pp. 95-105.
Langley, A. (1999), “Strategies for theorizing from process data”, Academy of Management Review,
Vol. 24 No. 4, pp. 691-710.
Langley, A. and Abdallah, C. (2011), “Templates and turns in qualitative studies of strategy and
management”, in Berg, D.D. and Ketchen, D.J. (Eds), Building Methodological Bridges, Research
Methodology in Strategy and Management, Emerald Group Publishing, Bingley, pp. 201-235.
Leech, N.L. and Onwuegbuzie, A.J. (2011), “Beyond constant comparison qualitative data analysis:
using NVivo”, School Psychology Quarterly, Vol. 26 No. 1, pp. 70-84.
Liao, H. (2007), “Do it right this time: the role of employee service recovery performance in
customer-perceived justice and customer loyalty after service failures”, Journal of Applied
Psychology, Vol. 92 No. 2, p. 475.
Lin, A.C. (1998), “Bridging positivist and interpretivist approaches to qualitative methods”, Policy
Studies Journal, Vol. 26 No. 1, pp. 162-180.
MacMillan, K. (2005), “More than just coding? Evaluating CAQDAS in a discourse analysis of news
texts”, FORUM: Qualitative Social Research, Vol. 6 No. 3, available at: www.qualitative-research.
net/fqs/fqs-eng.htm
Malshe, A. (2009), “Strategic sales organizations: transformation challenges and facilitators within the
sales–marketing interface”, Journal of Strategic Marketing, Vol. 17 Nos 3/4, pp. 271-289.
Malshe, A., Al-Habib, M., Al-Torkistani, H. and Al-Khatib, J.A. (2013), “Challenges to sales force
transformation in emerging markets”, Journal of Strategic Marketing, Vol. 21 No. 4, pp. 347-367.
Mattila, A.S. (2001), “The effectiveness of service recovery in a multi-industry setting”, Journal of
Services Marketing, Vol. 15 No. 7, pp. 583-596.
Mattila, A.S. and Mount, D.J. (2003), “The impact of selected customer characteristics and response time
on e-complaint satisfaction and return intent”, International Journal of Hospitality Management,
Vol. 22 No. 2, pp. 135-145.
Morse, J.M. and Richards, L. (2002), Read Me First for a User’s Guide to Qualitative Methods, Sage,
Thousand Oaks, CA.
Muhr, T. (1991), “Atlas/ti – a prototype for the support of text interpretation”, Qualitative Sociology,
Vol. 14 No. 4, pp. 349-371.
Nag, R., Corley, K.G. and Gioia, D.A. (2007), “The intersection of organizational identity, knowledge, and
practice: attempting strategic change via knowledge grafting”, Academy of Management
Journal, Vol. 50 No. 4, pp. 821-847.
Nimon, K.F. and Oswald, F.L. (2013), “Understanding the results of multiple linear regression beyond
standardized regression coefficients”, Organizational Research Methods, Vol. 16 No. 4,
pp. 650-674.
 QMR Ocasio, W. (1997), “Towards an attention-based view of the firm”, Strategic Management Journal,
Vol. 18 (Summer), pp. 187-206.
20,1
Oliveira, P. and Roth, A.V. (2012), “The influence of service orientation on B2B e-Service capabilities: an
empirical investigation”, Production and Operations Management, Vol. 21 No. 3, pp. 423-443.
Parasuraman, A., Zeithaml, V.A. and Berry, L.L. (1985), “A conceptual model of service quality and its
implications for future research”, Journal of Marketing, Vol. 49 No. 4, pp. 41-50.
120 Paulus, T.M. and Lester, J.N. (2016), “ATLAS.ti for conversation and discourse analysis studies”,
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

International Journal of Social Research Methodology, Vol. 19 No. 4, pp. 405-428.

Peters, L.D., Pressey, A.D., Vanharanta, M. and Johnston, W.J. (2013), “Constructivism and critical
realism as alternative approaches to the study of business networks: convergences and
divergences in theory and in research practice”, Industrial Marketing Management, Vol. 42 No. 3,
pp. 336-346.
Phipps, S. (2015), “The rise and rise of open source”, available at: www.infoworld.com/article/2914643/
open-source-software/rise-and-rise-of-open-source.html (accessed 21 November 2016).
Pratt, M.G. (2008), “Fitting oval pegs into round holes: tensions in evaluating and publishing qualitative
research in top North American journals”, Organizational Research Methods, Vol. 11,
pp. 481-509.
Pratt, M.G. (2009), “From the editors: for the lack of a boilerplate: tips on writing up (and reviewing)
qualitative research”, Academy of Management Journal, Vol. 52 No. 5, pp. 856-862.
Ramoglou, S. and Tsang, E. (2015), “A realist perspective of entrepreneurship: opportunities as
propensities”, Academy of Management Review, Vol. 41 No. 3, pp. 410-434.
Reficco, E. and Jaén, M.H. (2015), “Case method use in shaping well-rounded Latin American MBAs”,
Journal of Business Research, Vol. 68 No. 12, pp. 2540-2551.
Robertson, N. (2012), “Self-service technology complaint channel choice: exploring consumers’
motives,” Managing Service Quality: An International Journal, Vol. 22 No. 2, pp. 145-164.
Roschk, H. and Gelbrich, K. (2014), “Identifying appropriate compensation types for service failures a
meta-analytic and experimental analysis”, Journal of Service Research, Vol. 17 No. 2, pp. 195-211.
Saillard, E.K. (2011), “Systematic versus interpretive analysis with two CAQDAS packages: NVivo and
MAXQDA”, Forum Qualitative Sozialforschung/Forum: Qualitative Social Research, Vol. 12
No. 1.
Sarmah, S.P., Acharya, D. and Goyal, S.K. (2006), “Buyer vendor coordination models in supply chain
management,” European Journal of Operational Research, Vol. 175 No. 1, pp. 1-15.
Schmid, S., Grosche, P. and Mayrhofer, U. (2016), “Configuration and coordination of international
marketing activities”, International Business Review, Vol. 25 No. 2, pp. 535-547.
Schwarzkopf, D., Behrend, S., Skupin, H., Westermann, I., Riedemann, N.C., Pfeifer, R., Günther, A.,
Witte, O.W., Reinhart, K. and Hartog, C.S. (2013), “Family satisfaction in the intensive care unit:
a quantitative and qualitative analysis”, Intensive Care Medicine, Vol. 39 No. 6, pp. 1071-1079.
Short, J.C., and Palmer, T.B. (2007), “The application of DICTION to content analysis research in
strategic management,” Organizational Research Methods, Vol. 11 No. 4, pp. 727-752.
Short, J.C., Moss, T.W. and Lumpkin, G.T. (2009), “Research in social entrepreneurship: past
contributions and future opportunities”, Strategic entrepreneurship journal, Vol. 3 No. 2,
pp. 161-194.
Sinkovics, R.R. and Alfoldi, E.A. (2012), “Progressive focusing and trustworthiness in qualitative
research”, Management International Review, Vol. 52 No. 6, pp. 817-845.
Sinkovics, R.R., Penz, E. and Ghauri, P.N. (2005), “Analysing textual data in international marketing
research”, Qualitative Market Research: an International Journal, Vol. 8 No. 1, pp. 9-38.
Sinkovics, R.R., Penz, E. and Ghauri, P.N. (2008), “Enhancing the trustworthiness of qualitative research
in international business”, Management International Review, Vol. 48 No. 6, pp. 689-714.
 Sobh, R. and Perry, C. (2006), “Research design and data analysis in realism research”, European Journal Methodology
of Marketing, Vol. 40 Nos 11/12, pp. 1194-1209.
for qualitative
Sotiriadou, P., Brouwers, J. and Le, T.A. (2014), “Choosing a qualitative data analysis tool: a comparison
of NVivo and Leximancer”, Annals of Leisure Research, Vol. 17 No. 2, pp. 218-234.
research
Strauss, A. and Corbin, J. (1990), Basics of Qualitative Research, Vol. 15, Sage, Newbury Park, CA.
Thompson, C.J. (1997), “Interpreting consumers: a hermeneutical framework for deriving marketing
insights from the texts of consumers’ consumption stories”, Journal of Marketing Research, 121
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Vol. 34 No. 4, pp. 438-455.

Tsang, E.W. (2014), “Case studies and generalization in information systems research: a critical realist
perspective”, The Journal of Strategic Information Systems, Vol. 23 No. 2, pp. 174-186.
Tsoukas, H. (2016), “Don’t simplify, complexify: from disjunctive to conjunctive theorizing in
organization and management studies”, Journal of Management Studies, forthcoming.
Vaccaro, A. and Palazzo, G. (2015), “Values against violence: institutional change in societies dominated
by organized crime”, Academy of Management Journal, Vol. 58 No. 4, pp. 1075-1101.
Van de Ven, A.H. and Engleman, R.M. (2004), “Event and outcome-driven explanations of
entrepreneurship”, Journal of Business Venturing, Vol. 19 No. 3, pp. 343-358.
Vance, A. (2009a), “R you ready for R?”, NY Times Blog, available at: http://bits.blogs.nytimes.com/
2009/01/08/r-you-ready-for-r/ (accessed: 8 September 2016).
Vance, A. (2009b), “Data analysts captivated by R’s power”, New York Times, available at: www.
nytimes.com/2009/01/07/technology/business-computing/07program.html?pagewanted⫽all&_
r⫽0 (accessed: 21 November 2015).
Vanwindekens, F.M., Stilmant, D. and Baret, P.V. (2013), “Development of a broadened cognitive
mapping approach for analysing systems of practices in social – ecological systems”, Ecological
Modelling, Vol. 250 (February), pp. 352-362.
Wang, D., Yan, K.K., Rozowsky, J., Pan, E. and Gerstein, M. (2016), “Temporal dynamics of collaborative
networks in large scientific consortia”, Trends in Genetics, Vol. 32 No. 5, pp. 251-253.
Weller, M. (2011), The Digital Scholar: How Technology is Transforming Scholarly Practice,
Bloomsbury Academic, London.
Woods, M., Paulus, T., Atkins, D.P. and Macklin, R. (2015a), “Advancing qualitative research using
qualitative data analysis software (QDAS)? Reviewing potential versus practice in published
studies using ATLAS.ti and NVivo, 1994-2013”, Social Science Computer Review, Vol. 34 No. 5,
doi: 0894439315596311.
Woods, M., Macklin, R. and Lewis, G.K. (2015b), “Researcher reflexivity: exploring the impacts of
CAQDAS use”, International Journal of Social Research Methodology, Vol. 19 No. 4, pp. 1-19.
Yang, S.S.J. and Chandra, Y. (2013), “Growing artificial entrepreneurs: advancing entrepreneurship
research using agent-based simulation approach”, International Journal of Entrepreneurial
Behavior & Research, Vol. 19 No. 2, pp. 210-237.
Ye, B. (2010), “Using RQDAtm and tm to do text-mining”, available at: http://rqda.r-forge.r-project.org/
Using_RQDA2tm.pdf (accessed 19 November 2015).
Ye, Q., Law, R., Gu, B. and Chen, W. (2011), “The influence of user-generated content on traveler
behavior: an empirical investigation on the effects of e-word-of-mouth to hotel online bookings”,
Computers in Human Behavior, Vol. 27 No. 2, pp. 634-639.
Yin, R. (2003), Case Study Research: Design and Methods, Sage Publications, California.

Further reading
Friese, S. (2013), ATLAS.ti User Guide and Reference, Atlas GmbH, Berlin.
Friese, S. (2014), Qualitative Data Analysis with ATLAS.ti, Sage, London.
 QMR Graebner, M.E. and Eisenhardt, K.M. (2004), “The seller’s side of the story: acquisition as courtship and
governance as syndicate in entrepreneurial firms”, Administrative Science Quarterly, Vol. 49
20,1 No. 3, pp. 366-403.
Nag, R. and Gioia, D.A. (2012), “From common to uncommon knowledge: foundations of firm-specific
use of knowledge as a resource”, Academy of Management Journal, Vol. 55 No. 2, pp. 421-457.
Ozcan, P. and Eisenhardt, K.M. (2009), “Origin of alliance portfolios: entrepreneurs, network strategies,
122 and firm performance”, Academy of Management Journal, Vol. 52 No. 2, pp. 246-279.
Downloaded by CITY UNIVERSITY OF HONG KONG, Doctor Yanto Chandra At 17:42 09 January 2017 (PT)

Tracey, P., Phillips, N. and Jarvis, O. (2011), “Bridging institutional entrepreneurship and the creation of
new organizational forms: a multilevel model”, Organization Science, Vol. 22 No. 1, pp. 60-80.

About the authors

Yanto Chandra, PhD, is an Assistant Professor at the Department of Public Policy at the City University
of Hong Kong. His research interests include social, sustainable, developmental entrepreneurship,
social investment and entrepreneurship, in general. Methods of interest: computer-aided text analysis,
corpus linguistics, text mining, scientometrics; qualitative and visual methods. He serves on the
editorial board of the Journal of Social Entrepreneurship, Social Enterprise Journal, Journal of
International Marketing and Journal of Data Mining & Digital Humanities. He has published in Journal
of Social Entrepreneurship, Social Enterprise Journal, PLos ONE, Asia Pacific Journal of Management,
Journal of World Business, Journal of International Marketing, Technovation, among others.
Previously, he was an Assistant Professor at the University of Amsterdam and University of Leeds, and
has 7 years of work experience in large multinational companies. Yanto Chandra is the corresponding
author and can be contacted at: [email protected]
Liang Shang is a PhD candidate in Public Policy and Management at the Department of Public
Policy, at the City University of Hong Kong, after completing a degree in Social Policy and
Administration at the Hong Kong Polytechnic University. Her research focuses on social
entrepreneurship, non-profit organizations and civil society, in general. She is an experienced RQDA
user.

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]

View publication stats