Topic Modeling in Management Research: Rendering New Theory from Textual Data

Abstract

Increasingly, management researchers are using topic modeling, a new method borrowed from computer science, to reveal phenomenon-based constructs and grounded conceptual relationships in textual data. By conceptualizing topic modeling as the process of rendering constructs and conceptual relationships from textual data, we demonstrate how this new method can advance management scholarship without turning topic modeling into a black box of complex computer-driven algorithms. We begin by comparing features of topic modeling to related techniques (content analysis, grounded theorizing, and natural language processing). We then walk through the steps of rendering with topic modeling and apply rendering to management articles that draw on topic modeling. Doing so enables us to identify and discuss how topic modeling has advanced management theory in five areas: detecting novelty and emergence, developing inductive classification systems, understanding online audiences and products, analyzing frames and social movements, and understanding cultural dynamics. We conclude with a review of new topic modeling trends and revisit the role of researcher interpretation in a world of computer-driven textual analysis.

Full text

TOPIC MODELING IN MANAGEMENT RESEARCH:
RENDERING NEW THEORY FROM TEXTUAL DATA
Submission to the Academy of Management Annals
Tim Hannigan (U. of Alberta: tim.hannigan@ualberta.ca), Richard F.J. Haans (Rotterdam
School of Management, Erasmus University: haans@rsm.nl), Keyvan Vakili (London Business
School: kvakili@london.edu), Hovig Tchalian (Claremont Graduate University:
hovig.tchalian@cgu.edu), Vern L. Glaser (U. of Alberta: vglaser@ualberta.ca), Milo Wang (U.
of Alberta: swang7@ualberta.ca), Sarah Kaplan (U. of Toronto:
sarah.kaplan@rotman.utoronto.ca), and P. Devereaux Jennings (U. of Alberta:
dev.jennings@ualberta.ca)
Corresponding Author:
Dev Jennings
Alberta School of Business
University of Alberta
dev.jennings@ualberta.ca
780-492-3998
Approx. Word Count = 27,500 + 3,000 for appendix (25,000 normal max.)
Key words: topic modeling, management theory, rendering, text analysis, big data, theory
building, qualitative analysis, mixed methods
____
We would like to thank the editors of the Academy of Management Annals for their support and helpful comments.
We also thank the participants in our various topic modeling presentations and reviewers and division organizers
(specifically Peer Fiss and Renate Meyer) at the Academy of Management meetings. In addition, we would like to
recognize Marc-David Seidel and Christopher Steele from the Interpretive Data Science (IDeaS) group for their role
in germinating these ideas, Mike Pfarrer for his comments on a later draft of the paper, and Kara Gehman for her
fine-grained edits on next-to-final drafts. Finally, we wish to express our appreciation to our life partners for not
only putting up with, but actively discussing this paper as it evolved.

1
ABSTRACT
Increasingly, management researchers are using topic modeling, a new method borrowed
from computer science, to reveal phenomenon-based constructs and grounded conceptual
relationships in textual data. By conceptualizing topic modeling as the process of rendering
constructs and conceptual relationships from textual data, we demonstrate how this new method
can advance management scholarship without turning topic modeling into a black box of
complex computer-driven algorithms. We begin by comparing features of topic modeling to
related techniques (content analysis, grounded theorizing, and natural language processing). We
then walk through the steps of rendering with topic modeling and apply rendering to
management articles that draw on topic modeling. Doing so enables us to identify and discuss
how topic modeling has advanced management theory in five areas: detecting novelty and
emergence, developing inductive classification systems, understanding online audiences and
products, analyzing frames and social movements, and understanding cultural dynamics. We
conclude with a review of new topic modeling trends and revisit the role of researcher
interpretation in a world of computer-driven textual analysis.
N = 168 words

2
TOPIC MODELING IN MANAGEMENT RESEARCH:
RENDERING NEW THEORY FROM TEXTUAL DATA
New methods can have profound impacts on management scholarship (Arora, Gittelman,
Kaplan, Lynch, Mitchell, & Siggelkow, 2016), as they enable scholars to take fresh approaches
to theory and re-examine previously intractable problems and old questions (Timmermans &
Tavory, 2012). For example, the introduction of event history analysis helped advance both
population ecology (Hannan & Carroll, 1992) and institutional analysis (Tolbert & Zucker, 1996)
research; the introduction of the case comparison method aided the development of strategy
process research (Eisenhardt, 1989); and the introduction of set theoretic methods and qualitative
comparative analysis (QCA) led to renewed investigations of configurations (Fiss, 2007; Ragin,
2008). Recently, the management field’s understandings of cognition, meaning, and
interpretation have been dramatically reshaped by the emergence of new computer-based
language processing techniques (DiMaggio, 2015), which have amplified and sharpened the
linguistic turn in management research (Alvesson & Kärreman, 2000). In our review, we focus
on one of the most commonly used new techniques: topic modeling.
During the last decade, social scientists have increasingly used topic modeling to analyze
textual data. Borrowed from computer science, this method involves using algorithms to analyze
a corpus (a set of textual documents) to generate a representation of the latent topics discussed
therein (Mohr & Bogdanov, 2013; Schmiedel, Müller, & vom Brocke, 2018). It has helped
scholars unpack conundrums in management theory, such as how critics’ framings of corporate
activities simultaneously affect and are affected by their audiences (Giorgi & Weber, 2015), and
how knowledge recombination is a double-edged sword with opposite impacts on an
innovation’s degree of novelty and its usefulness (Kaplan & Vakili, 2015). Similarly, topic

3
modeling has been used to generate new conceptual linkages, such as how a particular topic
appearing in media statements impacted departures of British parliament members (Hannigan,
Porac, Bundy, Wade, & Graffin, 2019), and to refine older constructs such as strategic
differentiation (Haans, 2019). Because of its features, topic modeling can serve as a bridge in the
social sciences, for it sits at the interfaces between case studies and big data, unstructured and
structured analysis, and induction and deduction (DiMaggio, Nag, & Blei, 2013; Grimmer &
Stewart, 2013; Mützel, 2015). Not surprisingly, its use in social science, and in management
theory more specifically, has increased greatly over the last decade.
As with all new methods, topic modeling techniques continue to be refined. In the current
emergent phase of its employment, scholars are still learning the best ways to reveal constructs
and develop theory (Evans & Aceves, 2016; Grimmer & Stewart, 2013)—which implies a need
for deeper insights into how topic modeling can inform new theories. There are also many
technical issues to resolve around topic modeling, such as how to collect and prepare data (Evans
& Aceves, 2016), how much supervision should be involved in topic creation (DiMaggio, 2015;
Schmiedel et al., 2018), which algorithms are most useful (Bail, 2014), and how new constructs
and conceptual linkages can be derived when developing theories from big data (Nelson, 2017,
Timmermans & Tavory, 2012). This review addresses these questions with the aim of expanding
its use and effectiveness.
We begin by comparing topic modeling’s technical and theory-building features to those
of close methodological cousins: content analysis, grounded theorizing, and general natural

4
language processing (NLP) of text.
1
Topic modeling’s attractive features and ease of use are
generating increased interest across the social sciences—raising the disconcerting possibility that
the method will become a technical “black box” without an appropriate appreciation of topic
modeling’s statistical and theoretical underpinnings and implications. In this review, we show
that topic modeling is best conceptualized as a “rendering process,” which can be understood as
a means to juxtapose data and theory (Charmaz, 2014) in order to generate new theoretical
artifacts such as constructs and the links between them (Whetten, 1989). This process involves
the rendering of corpora (preparing the sets of texts to be analyzed), the rendering of topics
(making analytical choices that determine how topics are identified within those texts), and the
rendering of theoretical artifacts (crafting topics into constructs, causal links or mechanisms). By
articulating this rendering process, we show that using the machine learning algorithms of topic
modeling do not reduce textual analysis to a mechanistic process, but actually foreground and
inform the analyst’s interpretive decisions and theory work.
Our own topic modeling analysis of topic modeling articles created or routinely used by
management researchers reveals five theoretical subject areas to which the technique has
contributed: detecting novelty and emergence, developing inductive classification systems,
understanding online audiences and products, analyzing frames and social movements, and
understanding cultural dynamics. For each subject area, we review key concepts and theoretical
relationships that have surfaced from the use of topic modeling and identify articles that
exemplify its application. We then turn to new trends in topic modeling in the rendering of
1
Topic modeling can be seen both as a specific NLP approach and as something distinct from NLP. Topic modeling
relies on interpretation and language-oriented rules, but is also unique in its emphasis on the role of human
researchers in generating and interpreting specific groups of topics based on the social contexts in which they are
embedded. Recent developments have also moved topic modeling further away from NLP, as researchers have
applied it to images (Cao & Fei-Fei, 2007) and music (Hu & Saul, 2009) rather than natural language.

5
corpora, topics, and theoretical artifacts. Our review demonstrates that topic modeling not only
appeals to diverse management audiences—those interested in topic, content, and category
models as well as mixed methods—but also can play a part in cultural structuralism (Lounsbury
& Ventresca, 2003), new archivalism (Ventresca & Mohr, 2002), and interpretative data science
(Breiger et al., 2018; Mattmann, 2013).
SITUATING TOPIC MODELING AS A TECHNIQUE
Thanks to widespread availability of digitized textual data from a variety of sources and
significant increases in computational power, it is now possible for social scientists to study large
collections of text (Alvesson & Kärreman, 2000; Langley & Abdallah, 2011; Vaara, 2010). Not
surprisingly, a variety of methods for textual analysis—often from neighboring disciplines—
have appeared as part of this “linguistic turn.” To distinguish the key characteristics of topic
modeling and situate it among this wider set of techniques, we first briefly examine three closely
related methods: content analysis (Duriau, Reger, & Pfarrer, 2007; Krippendorf, 1980, 2004;
Lasswell, 1948), grounded theorizing with textual data (Gioia, Corley, & Hamilton, 2013; Locke,
2001), and interpretive analysis using the broad class of NLP approaches. These three are
particularly useful for elucidating topic modeling’s features because they capture the extremes
from highly contextualized, careful assessment of smaller batches of selected texts to broader,
more algorithmic and systematic assessment of text from large corpora.
Content analysis. Social scientists have long been interested in using texts to understand
social phenomena (see Krippendorf, 1980 for a review). Content analysis, “a research technique
for the objective, systematic, and quantitative description of the manifest content of
communication” (Berelson, 1952, p. 18) represents arguably the most prominent and mainstream

6
approach in this domain (Nelson, 2017; Tirunillai & Tellis, 2014). It relies on the creation of
dictionaries or indices comprised of mutually exclusive lists of words that can then be applied to
texts to isolate meanings and systematically measure specific constructs of interest to the
researcher (Krippendorff, 2004). Since its introduction to management theory, scholars have
employed content analysis in flexible ways, using a range of data sources in areas as varied as
the study of management fads (Abrahamson & Fairchild, 1999), industry categories and CEO
compensation (Porac, Wade, & Pollock, 1999), corporate reputation (Pfarrer, Pollock, &
Rindova, 2010), and technology strategy (Kaplan, 2008a).
From its inception, content analysis scholars have been particularly concerned with the
reliability and validity of its various methods (Weber, 1990), advocating the use of protocols and
multiple coders to guide text selection and analysis. In recent years, those who employ content
analysis have increasingly relied on computer-aided text analysis using software and general
dictionaries such as General Inquirer and Linguistic Inquiry and Word Count (LIWC) to further
improve its scalability and systematic nature. At the same time, the mutually exclusive nature of
dictionaries precludes “polysemy” (DiMaggio et al., 2013, p. 578)—an important concept in
linguistics where the same word may have a different meaning based on the context in which it
appears. A common critique of content analysis has therefore been that it yields decontextualized
results by reducing complex theoretical constructs into overly general and simple indices (Dey,
1995; Prein & Kelle, 1995).
Grounded theorizing with textual data. To develop theory, scholars often use a highly
contextualized approach whereby they gather and engage intensively with texts and then use
comparative coding to identify higher-order constructs (Charmaz, 2014). By engaging in such
grounded theorizing with textual data, a researcher demonstrates a commitment to “‘discovery’

7
through direct contact with the social world studied coupled with a rejection of a priori
theorizing” (Locke, 2001, p. 34). Proponents of this approach urge researchers to start with a
loosely scoped research question and phenomenon of interest, with the researcher subsequently
identifying recurring patterns, ideas, or elements that emerge directly from the data. Doing so
often requires culling primary observations and key points and then using axial coding to identify
constructs or relationships (Denzin & Lincoln, 2011). Researchers then iteratively group codes
into higher-order categories to develop general theory. Rather than measurement, grounded
theorizing is thus fundamentally concerned with identifying deeper structures embedded in data
to attain a rich understanding of social processes.
During the last two decades, grounded theorizing has been used by many groups of
management scholars (Charmaz, 2014), including those interested in analyzing language in
organizations (Alvesson & Kareman, 2000), organizational processes and routines (Langley,
1999; Pentland & Feldman, 2005), and culture and identity (Hatch & Schultz, 2017; Nelsen &
Barley, 1997). Its theoretical flexibility also makes it the target of some critiques, because the
role and primacy of meaning, discourse, and understanding typically are not made explicit in
research studies (Locke, 2001). Practically speaking, the method also requires great knowledge
of context and expertise to apply; it can be not only time- and resource-intensive, but also
difficult to use with large scale textual data (Baumer, Mimno, Guha, Quan, & Gay, 2017;
Gehman, Glaser, Eisenhardt, Gioia, Langley, & Corley, 2018).
Interpretive analysis using NLP. Researchers in linguistics have long employed
computerization to enable systematized analysis of natural language informed by linguistic rules,
with NLP emerging in the 1980s as a way to combine dictionary-based data processing with
semantic use to map out likely interpretations of text (Manning & Schütze, 1999). Early versions

8
of NLP relied heavily on grammatical rules from language structure, but have given way to more
flexible, stochastic approaches to language use (especially as machine learning-based approaches
evolved with increased computing power). In management research, scholars often leverage NLP
tools to perform semantic parsing on big data and then interpret emerging patterns using
computer-aided recognition tools. Kennedy (2005, 2008) was one of the first to analyze media
data and sort through evaluations of firms using these tools. Recently, Mollick and others have
studied linguistic patterns in crowdfunding and other contexts involving pitches (Kaminski,
Jiang, Piller, & Hopp, 2017; Mollick, 2014).
Consistent with its roots in computer science, NLP has been developed to optimize
specific tasks or solve particular problems, such as part-of-speech tagging, word segmentation,
machine translation, and automatic text summarization. This has resulted in a rich and varied
toolkit that is deeply informed by linguistic rules and a firm appreciation for the complexities
underpinning human language. At the same time, a single unifying theory does not link the
various NLP tools, nor are there standard practices or rules about engaging in NLP-based work.
This has created certain challenges for management researchers in applying technical or
descriptive tools for theoretically informed purposes. Indeed, scholars have noted that
“cooperation between linguistics and the social sciences with regard to text analysis has always
been meager” (Pollach, 2012: 264); however, this does not imply that NLP approaches are, by
definition, unable to inform management theory.
Topic modeling. In the early 2000s, topic modeling was developed as a unique NLP-like
approach to information retrieval and the classification of large bodies of text (Blei, Ng, &
Jordan, 2003). Topic modeling uses statistical associations of words in a text to generate latent
topics—clusters of co-occurring words that jointly represent higher-order concepts—but without

9
the aid of pre-defined, explicit dictionaries or interpretive rules. In a pivotal article, Blei et al.
(2003) introduced a Bayesian probabilistic model using latent Dirichlet allocation (LDA) to
uncover latent structures in texts. LDA is a “statistical model of language” (DiMaggio et al.,
2013, p. 577) and is the simplest of several possible generative models available for topic
modeling (Blei, 2012). It focuses on words that co-occur in documents, viewing documents as
random mixtures of latent topics, where each topic is itself a distribution among words (Blei et
al., 2003). Importantly, an assumption of topic modeling is that documents are “bags of words”
without syntax, which defines meaning as relational (Saussure, 1959) and emerging from co-
occurrence patterns independent of syntax, narrative, or location within the documents (Mohr,
Wagner-Pacifici, Breiger, & Bogdanov, 2013).
Generating topics using statistical probabilities has three key benefits. First, researchers
do not have to impose dictionaries and interpretive rules on the data. Second, the method enables
the identification of important themes that human readers are unable to discern. Third, it allows
for polysemy because topics are not mutually exclusive; individual words appear across topics
with differing probabilities, and topics themselves may overlap or cluster (DiMaggio et al., 2013,
p. 578).
A comparison of text analysis techniques in management research. Figure 1
compares the use of topic modeling in social science and management research to the use of
grounded theory, content analysis, and general NLP approaches in articles listed in the Web of
Science and Scopus published between 2003 (the year Blei and colleagues’ foundational article
was published) and 2017. We included articles for topic modeling if “topic mod*” appears in
their titles, abstracts, keywords, or automated indexed keywords. We included articles for
grounded theorization, content analysis and NLP if they contain “ground theor*,” “content

10
analys*,” and “natural language process*,” respectively.
2
The bar charts in each panel represent
the cumulative number of articles in each year, with black bars showing the number of articles in
business and economics specifically, and white bars showing articles in the social sciences more
generally.
--- Insert Figure 1 about here ---
As a group, the four panels highlight the linguistic turn in social science, with increased
use of all of these approaches reflecting the increasing appetite in the field to study the structure
and meaning underpinning collections of text. By 2017, 1,000 topic modeling articles had been
published, with around 300 in the management domain specifically. Although this is just a
fraction of the literature relative to studies based on more established approaches, Figure 1 does
suggest that the use of topic modeling has been particularly high in the management domain.
Indeed, 29.8% of all articles based on topic modeling published between 2003 and 2017 fall
within the management domain, compared to 13.4%, 22.0%, and 22.9% for NLP, grounded
theorization, and content analysis, respectively. Figure 1 also reveals that topic modeling has
been adopted at an exceptionally rapid rate in recent years, with a compound annual growth rate
of 34.4% since 2010, versus 11.1% for NLP, 15.1% for grounded theory, and 16.5% for content
analysis. We suggest that topic modeling’s appeal primarily lies in its unique position at the
intersection of the other three approaches, a point that we elaborate in the conclusion.
RENDERING THEORY FROM DATA IN TOPIC MODELING
Given its increasing importance in the social sciences and its unique location between
2
Although these may under-count articles that do not mention the methodologies and over-count articles without
textual data, we suspect that these issues are equally salient for each approach. For illustration, adding “Linguistic
Inquiry and Word Count” and “LIWC” adds just 271 articles to the set of over 20,000 for content analysis.

11
human-based and machine-learned analysis of discourse, a more careful consideration of the
nature of topic modeling and the topic modeling process is useful for management researchers.
To date, much of the work on topic modeling has focused on issues of algorithm selection (e.g.,
Blei et al., 2003; Schmiedel et al., 2018) and its application to curated texts. We think it is
important to discuss the use of topic modeling from the pre-processing to theorization stages to
illustrate its possibilities for theory building.
We use the term “rendering” to describe the iterative creation of theory from corpora
through topic modeling. In the social sciences, Charmaz (2014, pp. 216, 369) employed the term
rendering to describe the process of “juxtaposing data and concept” and “categorizing data” for
interpretation, while computer scientists use rendering to create photorealistic or non-
photorealistic images in two or three dimensions via automated analysis and specific algorithms
(Strothotte & Schlechtweg, 2002). Drawing on these descriptions for inspiration, we define
rendering in topic modeling as a three-part process of generating provisional knowledge by
iterating between selecting and trimming raw textual data, applying algorithms and fitting
criteria to surface topics, and creating and building with theoretical artifacts, such as processes,
causal links, or measures. These three steps are displayed in Figure 2. To provide readers with
background information, we present definitions of common terms used in topic modeling in
Table 1.
--- Insert Figure 2 and Table 1 about here ---
Rendering corpora
In the first process—rendering corpora—an analyst, guided by theoretical and empirical
considerations, selects types of textual data. As with any form of empirical analysis, selection of
the sample (in our context, texts) is a crucial step that fundamentally shapes all subsequent steps.

12
For textual data in particular, selection needs to account for language, authoring, and document
sources—ensuring a logical fit with the research question being investigated while
simultaneously considering common issues such as representativeness, levels of analysis, and
temporal considerations (e.g., longitudinal vs. cross-sectional data). The analyst then compiles
such data for further pre-processing and cleaning. If the data are from one primary source, the
compiled text is considered a corpus; if from different sources, corpora.
On the whole, topic modeling tends to be applied more frequently to sampled corpora
than to a single, homogenous corpus (Borgman, 2015; Kitchin & McArdle, 2016). As a result,
topic modeling relies on a great deal of pre-processing with various techniques and rules of
practice to prepare texts for analysis (Nelson, 2017; Schmiedel et al., 2018). During pre-
processing, the texts are sorted, disassembled, and then trimmed according to broader content
analysis principles such as ignoring “stop words” (for example: “the” and “a”) and focusing on
nouns rather than verbs, adjectives, or adverbs. Topic modelers also often standardize word
forms, using stemming and lemmatizing (see Table 1) to transform words into their roots
(Kobayashi, Mol, Berkers, Kismihók, & Den Hartog, 2018). Recently, more refined techniques
such as WordNet have been developed to convert words to their singular forms or to use higher-
level synonyms (Miller, Beckwith, Fellbaum, Gross, & Miller, 1990). These considerations are
all crucial, as most topic modeling algorithms analyze words based on how they appear, letter-
by-letter (e.g., “firm” is not the same as “firms”). As such, these cleaning steps represent a form
of systematic, normatively-guided trimming to standardize words to allow the capture of
constellations of words that represent deeper socio-cultural structures (Mohr, 1998).
Rendering topics

13
During the second process—rendering topics—the analyst applies an algorithm to
identify appropriate topics. An algorithm provides an analyst with the ability to use a pre-
programmed set of rules to automatically reduce the dimensions of the corpora (e.g., Mohr,
1998). The most well-known algorithm, as discussed above, is LDA. According to Blei et al.
(2003, p. 994), the key assumption in LDA is that “each word in a document [is modeled] as a
sample from a mixture model, where the mixture components are multinomial random variables
that can be viewed as representations of ‘topics.’” The major theoretical and methodological
insight here is that documents are assumed to draw content from a latent set of topics with
probability-based parameters that can be adjusted to determine those topics. This implies that
words are generated from a topic, yet can also be used in different topics with different
probabilities. Because documents belong to the same corpus, the algorithm assumes that they
were generated from the same process, and thus each document constitutes a mixture of the same
set of “topics” in different proportions. Topics are a weighted vector of words and each topic
corresponds to a distinct concept (Grimmer & Stewart, 2013). However, unlike the dictionaries
used in content analysis, which are comprised of mutually exclusive lists of words
(Krippendorff, 2004, p. 132), in topic modeling, the same words can appear in different topics
(DiMaggio et al., 2013, p. 578), though likely in very different proportions and juxtaposed with
different words.
The inputs to the LDA algorithm include: (a) a set of documents that can be represented
as a document-word matrix—with rows representing each document in the corpora, columns
representing each unique word in the corpus, and cells indicating the number of times each word
occurs in each document—and (b) the number of topics to be estimated by the algorithm.
Importantly, most topic modeling algorithms (such as LDA) require probability draws for each

14
document, such that each document is considered “a bag of words” with no syntax. The outputs
from LDA include a topic-word matrix (vectors of the weights of words in each topic) and a
topic-document matrix (vectors of the weights of topics in each document). In subsequent
analyses, math (i.e., vector space calculations) can be applied to these outputs to classify texts
into categories, analyze themes, or compare corpora based on similarities.
Each successfully computed model is based on different parameters (e.g., number of
topics) and generates a distribution of topics over documents and/or words, which can be used by
the researcher to identify the eventual model that will be used in the study. The notion of fit is
typically invoked to decide how many topics are derived, how they are related, and what they
might mean. A researcher can focus on one of two notions of fit—rooted in a logic of either
accuracy or validity—and this focus has important implications for which topic model is judged
to provide the most appropriate fit given the research question.
One version of fit is based on a logic of accuracy, a central focus of computer scientists
who rely on metrics such as perplexity, log-likelihood and coherence (defined in Table 1) to
determine the number of topics and their salience (Azzopardi, Girolami, & van Risjbergen, 2003;
Chang, Boyd-Graber, Gerrish, Wang, & Blei, 2009; Mimno, Wallach, Talley, Leenders, &
McCallum, 2011). However, Chang et al. (2009) pointed to disparities between some
quantitative metrics and how people interpret topics: topic models that perform better on
quantitative metrics tend to infer topics that humans judge to be semantically less meaningful.
Indeed, DiMaggio et al. (2013, p. 582) suggested that “there is no statistical test for the optimal
number of topics or for the quality of a solution” and that “the point is not to estimate population
parameters correctly, but to identify the lens through which one can see the data most clearly.”
Therefore, social scientists tend to focus more on the logic of fit as validity (DiMaggio,

15
2015). DiMaggio et al. (2013) identified two key forms of validity: semantic or internal validity,
and predictive or external validity. To demonstrate internal validity, the researcher must confirm
that the model meaningfully discriminates between different senses of the same or similar terms.
To demonstrate external validity, the researcher must determine whether particular topics
correspond to information external to the topic model (e.g., by confirming that certain topics
became more salient when an external event relevant to those topics occurred) (DiMaggio et al.,
2013). For example, Kaplan and Vakili (2015) identified models with 50, 75 and 100 topics for a
corpora of nanotechnology patent abstracts and then used three expert evaluators to determine
that the 100-topic model was the most semantically meaningful. Jointly, these two forms of
validity are concerned with confirming that the topic model’s outputs are semantically
meaningful—a process that entails substantial interpretive uncertainty (DiMaggio, 2015). Due to
the uncertainty involved in the rendering of topics, most scholars in the social sciences attempt to
locate the optimal balance between the two logics of accuracy and validity to identify the “best”
topic model to be used in further theorizing.
In sum, topic modeling has advanced how we think about and interpret topics in textual
data by enabling researchers to uncover latent topics rather than imposing pre-established
categories on the data. It is superior to word-count techniques because it identifies ideas or
concepts based on constellations of words used across documents in a corpus. It is thus sensitive
to semiotic principles of polysemy (words with multiple meanings or uses), heteroglossia (uses
predicated on audiences and authors, as described by Bakhtin, 1982), and the relationality of
meaning (which is contextually dependent) (DiMaggio et al., 2013). As a result, topic model
outputs, after some interpretation and theoretical defense, are useful in generating theoretical
artifacts, especially in large and otherwise unmanageable data sets.

16
Rendering theoretical artifacts
In the third process—rendering theoretical artifacts—researchers iterate between theory
and the topics that emerge from the chosen model to create new theoretical artifacts or to build
theory with them (Whetten, 1989). The word- and topic-vectors offer a wide range of
opportunities for the researcher to build artifacts. The artifacts may be multi-dimensional
constructs, such as novelty (Kaplan & Vakili, 2015) or differentiation (Haans, 2019), captured by
a set of topics clustered or scaled around words or concepts. The artifacts may also be relational
(correlational, causal or process-based), thereby allowing researchers to uncover mechanisms.
For instance, Croidieu and Kim (2018, p. 11) used an “iterative, multi-step process” to
interpret the outputs of the topic model in order to discover concepts related to lay expertise
legitimation and the mechanisms underpinning it. They described their process for creating
theoretical artifacts from their algorithmic output in detail.
First, we started with the raw topics as descriptive codes. Second, we labeled these topics
as first-order concepts. We coded all labels separately and together as an author team,
extensively discussed the results, and recoded the topics when necessary. Third, we
grouped these topics into more abstract and general second-order themes. Fourth, we
analyzed the distribution of these second-order themes per year and iteratively developed
four aggregate dimensions, which we present in the following sections as the mechanisms
for expertise legitimation. Fifth, we refined the labeling and theorizing of these aggregate
dimensions by dividing our analysis into two periods…We chose these periods both for
their historical significance and because they are anchored by a central empirical puzzle
related to our theoretical framework…Last, we repeated this procedure multiple times to
ensure tight correspondence between our raw-topic data and our coding interpretations.
From this iterative coding work, we produced our findings and constructed our process
model. (Croidieu & Kim, 2018, p. 11)
The inherent flexibility of the rendering process has enabled topic modeling researchers
to develop better measures and clever extensions of existing theoretical constructs and
relationships, and to induce novel concepts, processes, and mechanisms. As such, topic modeling

17
can be used for either deductive or inductive theorizing. Indeed, during the rendering process,
different choices arise (e.g., around selection, fit, and the form of artifact) based on whether one
uses more deductive versus inductive theorizing. The many paths defined by these choices
provide further evidence of topic modeling’s flexibility and potential. Not surprisingly, topic
modeling is contributing to a wide array of management theory subjects, some arising from more
mature theory, some from emerging areas.
BUILDING MANAGEMENT KNOWLEDGE THROUGH TOPIC MODELING
During the 15 years since topic modeling was first employed in management research, its
use through rendering has enabled management scholars to explore subjects in new ways,
thereby building management knowledge. To systematically identify the subjects enhanced by
such rendering, we applied the topic modeling rendering process depicted in Figure 2 to topic
modeling articles in the literature (for similar meta-theorizing moves, see Mohr & Bogdanov,
2013, or Wang, Bendle, Mai, & Cotte, 2015). Although our rendering process was iterative and
recursive, we present our methodological approach as a series of sequential steps, as outlined in
Figure 1 (e.g., rendering our corpus, topics, and theoretical artifacts).
We began our analysis by curating a corpus consisting of all relevant topic modeling
articles from the Web of Science and Scopus. We winnowed those articles down by focusing on
management journals (e.g., ASQ, SMJ, etc.) and other journals that management scholars read.
We identified these journals based on both our first-hand experience and citations of articles that
have influenced management scholars. Following the procedure employed by Mohr and
Bogdanov (2013), we divided the articles into paragraphs to form 5,362 documents and used the
Stanford CoreNLP software (Manning et al., 2014) to lemmatize the words, yielding 351,786

18
distinct words for analysis. During our analysis, we sharpened our criteria for including and
excluding particular articles in our analysis as we interpreted the output of topic modeling
algorithms. Our final corpus contained 66 articles (for details, consult Table A1 in the
Appendix). We organized these procedures using the Jupyter Notebook software in Python,
which enabled us to track and visually annotate our process.
We continued our analysis by applying a collapsed Gibbs sampler with the LDA
algorithm to our corpus to render topics. Collapsed Gibbs sampling (Griffiths & Steyvers, 2004)
is an approach from the Markov Chain Monte Carlo framework that iteratively steps through
configurations to estimate optimal model fit. When combined with the LDA algorithm (Blei et
al., 2003), topics can be estimated with minimal configuration by the user. As is common
practice (e.g. Mohr & Bogdanov, 2013; Jha & Beckman, 2017), we used the MALLET software
tool (McCallum, 2002) to conduct this procedure. We approached the critical task of determining
the optimal number of topics by computing a variety of topic models. For each model, we
graphed the average coherence score across topics (Mimno et al., 2011), which revealed a
plateau value; we used this evidence as guidance and observed several models (i.e., those with
30, 35, 40, 45, and 50 topics) more closely from an interpretive perspective. Fligstein et al.
(2017) followed a similar procedure, moving from collapsed Gibbs sampling through various
models, using coherence and interpretability to narrow in on stable sets of topics. Finally,
following Mohr and Bogdanov (2013), we applied our 35-topic model (derived from separate
paragraphs) to each document to generate a distribution of topic weights (i.e., the topic-document
matrix where each row is a document and each column is a topic weight, with all weights adding
up to 1). We then sorted topics for salience based on average topic weights and word relevance
to identify 35 ordered topics.

19
Three co-authors then independently used the algorithmic output of the topic models to
render theoretical artifacts. Specifically, we each created a summary document for each topic
that contained three visualizations generated by the topic modeling algorithm: a weighted word
list, a weighted document list, and a multidimensional scaling visualization (Sievert & Shirley,
2014) that showed each topic in relation to other topics (see Appendix, Figure 2, for an example
of this theoretical artifact). The three authors then independently analyzed these documents to
generate first- and second-order codes (e.g., Bansal & Corley, 2014; Denzin & Lincoln, 2011;
Gioia et al., 2013; Pratt, 2009; Strauss & Corbin, 1998). Through a series of independent coding
exercises and interactive conversations, the authors then aggregated these first- and second-
order codes into broader management subject areas (e.g., Gioia et al., 2013). In other words, in
keeping with rendering practice, we tried not to impose too much meaning on the set of topics;
instead, we let the insights and themes for management theorizing emerge from them.
Our bottom-up, inductive analysis suggests that topic modeling has enhanced our
management theory knowledge in five subject areas: detecting novelty and emergence,
developing inductive classification systems, understanding online audiences and markets,
analyzing frames and social movements, and understanding cultural dynamics.
3
This specific
ordering of subjects is not determined by topic weights; moreover, the timing of their
identification in the model’s convergence does not reflect a strict ordering. In fact, our
preliminary analyses of the wider corpora in the field and understanding of the field’s evolution
reveal how analyses of novelty, classification and online audiences developed in parallel with
analyses of framing and cultural dynamics. In the sections that follow, we focus on how
3
In addition, some topics corresponded specifically to the method of performing topic modeling, and given our
interest in the rendering of management theory, we purposefully backgrounded these topics (see Appendix Table 2
for details).

20
theoretical knowledge in each subject area has been extended by rendering with topic modeling.
Subject areas, topic-based themes, exemplary articles, and theoretical contributions are
summarized in Table 2.
--- Insert Table 2 about here ---
Detecting novelty and emergence. Management researchers are interested in topics of
novelty and emergence because they apply to a variety of research streams, such as categories
(e.g., Durand & Khaire, 2017; Hannan et al., 2007; Kennedy & Fiss, 2013), cultural
entrepreneurship (e.g., Lounsbury & Glynn, 2001, 2019), innovation (e.g., Fleming, 2001;
Sørensen & Stuart, 2000), organizational forms (e.g., Rao et al., 2003), and changes in
managerial cognition and attention (e.g., Ocasio, 1997). Novelty is a key concern within
innovation studies (Kline & Rosenberg, 1986; Trajtenberg, 1990), but measures typically are
indirect. For instance, as noted by Kaplan and Vakili (2015), many studies identify emergence
based on the successful introduction of new innovations, thus raising concerns of endogeneity
and lack of causal identification.
Topic modeling offers a solution to fundamental challenges faced in these broad research
streams. Specifically, topic modeling can be applied to documents to generate theoretical insights
because: (a) the language used in documents represents their cognitive content (Whorf, 1956);
and (b) actors use vocabularies to describe similar ideas (Loewenstein, Ocasio, & Jones, 2012).
Thus, topic modeling can be used to discern the cognitive content of documents that describe
cases of novelty and emergence (i.e., innovation contexts) and assess the extent to which such
content is similar or different across documents. Topics rendered in our analysis include:
explaining shifts in patent citations (#25), understanding innovation (#24), managerial cognition
(#1), understanding knowledge dynamics (#14), and emerging organizational forms (#10).

21
The first topic in this subject area relates to the use of topic modeling to measure the
novelty of ideas in patents—an arena in which novelty has been heavily studied under the rubric
of recombination and innovation (Fleming, 2001). For instance, Kaplan and Vakili (2015)
applied topic modeling techniques to create representations of ideas in documents that can be
compared using mathematical distance to determine cognitive novelty. This measure of novelty
based on the actual cognitive content of documents provides several advantages over more
traditional measures of novelty based on citations in subsequent patents or publications
(Trajtenberg, 1990). In the popular citation-based approach, a patent is flagged as a breakthrough
if it has a substantial impact on subsequent technologies. However, citation-based measures of
technological novelty often confound novelty and impact (Momeni & Rost, 2016); consequently,
novel ideas may not be recognized as important precursors due to the processes by which
citations are produced (false negatives), and incremental ideas may be incorrectly identified as
novel when they generate substantial impact for reasons other than novelty (false positives).
In contrast to simple counts of citations or patent classes, a measure based on the
cognitive content of a document enables researchers to gauge the novelty of the idea(s)
presented, independent of their ex-post economic value. Kaplan and Vakili (2015) used topic
modeling to distinguish cognitive novelty from economic value. In their analysis of nanotube
patents, they reported a very small correlation between topics identified by LDA and patent
classes assigned by the U.S. Patent and Trademark Office (USPTO). Often, truly novel ideas are
assigned to classes that may not reflect their actual cognitive content. Their study has
implications for teasing out longstanding debates in management around contrasting theories of
creative processes surrounding the sources of innovative breakthroughs. In a related study,
Ruckman and McCarthy (2017) used topic modeling to analyze patents in an attempt to explain

22
why some patents are licensed over others. Their goal was to address conflicting findings in prior
research: some scholars have advocated a “status model” (Podolny, 1993), whereas others have
supported organizational learning explanations based on optimizing knowledge transfer in
licensing contracts (Arora, 1995). Ruckman and McCarthy used topic modeling to directly
measure cognitive content, enabling them to construct a set of “alternate patents” that could have
been licensed based on content, but were not. Thus, by controlling for cognitive content, they
were able to isolate other variables such as the licensor’s technological prestige and experience at
licensing, and characteristics of the patent itself such as combined technological breadth and
depth. Using better controls when comparing similar patents enabled them to produce a
contingent model of patent licensing likelihood based on licensor attributions and the
combination of technological breadth and depth as an attractive signal. Topic modeling has thus
enabled researchers who study patents and innovation to not only increase the precision of their
analyses, but also develop new theory about the role of knowledge dynamics on economic
outcomes.
A second topic in this subject area that is closely related to explaining shifts in patent
citations is the use of texts more generally as a means to measure innovation and creativity.
Toubia and Netzer (2016) proposed that creative and novel ideas should have some type of
structural signature that can be found in cognitive representations. Drawing on literature related
to cognitive creative processes in science (i.e., Rothenberg, 2014; Uzzi et al., 2013), they
explored this proposition as an optimal balance of familiarity and novelty. Toubia and Netzer
(2016) primarily adopted a semantic network analysis approach to explore the structural
argument of familiarity, showing how co-occurrences of word stems can constitute a common
substructure, what they called a “structural prototype.” In turn, they argued that creativity is a

23
function of a semantic network structure with a core substructure corresponding to a familiar
prototype, and novelty dimensions reflected as sufficient semantic distance in the overall
structure. They demonstrated this argument empirically across eight studies and 4,000 different
ideas in multiple domains that were coded by expert judges. They used LDA as a robustness
check to show that creativity was not simply a function of semantic distance. Interestingly, both
Toubia and Netzer (2016) and Kaplan and Vakili (2015) featured in this topic: in different
domains, the authors leveraged topic modeling techniques to theorize how to identify innovation
in documents through the direct measurement of cognitive representations.
The third and fourth topics—using topic models to understand managerial cognition and
knowledge dynamics—relate to actors detecting novelty within a body of knowledge. The core
idea of employing topic modeling to study knowledge dynamics is based on two related insights:
first, the language used in documents represents their cognitive content (Whorf, 1956); and
second, actors use similar vocabularies to describe similar ideas (Loewenstein, Ocasio, & Jones,
2012). In our analysis, the third topic reveals that topic models can be used to understand
changing cognition over time through varying managerial attention (Ocasio, 1997). When a
corpus covers the body of knowledge in a specific domain (e.g., scientific papers or patents in
the technology field), topic modeling can reveal an accurate depiction of the idea space in that
body of knowledge. However, topic modeling can also reveal how actors, as producers of
documents, can attend to ideas in the latent idea space. As Kaplan and Vakili (2015)
demonstrated, to the extent that describing a truly novel (or disruptive) idea requires using a new
vocabulary, one can identify the level of cognitive novelty in a document by measuring how
much it conforms to or deviates from previously established topics and their constitutive
vocabularies in the corresponding body of knowledge. Wilson and Joseph (2015, p. 417)

24
employed topic modeling to render the “patent background” as a “representation of a technical
problem” at a particular point in time. Because managerial attention is scarce, it is allocated
across a small set of technological problems, particularly at the level of a business unit (Argote
& Greve, 2007). Thus, the rise and fall of topics as technological problems reflect not only
managerial attention within a firm, but also novelty within the broader field or patent class.
Topic modeling has also been used to study knowledge dynamics in science by tracking
the novelty of ideas in journals over time. Conceptualizing scientific communities as “thought
collectives with distinct thought styles,” Antons, Joshi, and Salge (2018, p. 1) used topic
modeling to break down articles in terms of topical and rhetorical attributes. They demonstrated
that topical newness is not only associated with a paper “citation premium” in a scientific
community, but also significantly increases with a rhetorical stance of tentativeness rather than
certainty. Similarily, Wang et al. (2015) used topic modeling to discover emerging trends in
knowledge fields, noting that citation analyses and LDA together can be used to narrate a story
about novelty and progress against a broader backdrop of social structure, including niche topical
areas and author status dynamics. Both articles in this topic contextualize traditional citation
based measures of article impact against cognitive dynamics in topic analyses.
A final topic revealed by our analysis of this subject area reflects the use of topic
modeling to understand emerging organizational forms. This approach provides a method to
trace how meanings of organizational forms emerge longitudinally. Jha and Beckman (2017)
used topic modeling to show how field-level logics moderated actors’ attempts to carve out
organizational identities around charter schools. Topic modeling enabled the authors to connect
two traditionally distinct theoretical concepts—institutional logics and organizational
identities—and explain the relationships between them. Given how meaning has typically been

25
studied in organizational theory using concepts such as identity, institutional logics, and frames,
studying the emergence of meanings in spaces such as organizational fields and categories may
become an increasingly relevant application of topic modeling methods.
Topic modeling has increased precision and enabled deeper insights in studies of novelty
and knowledge dynamics, thereby facilitating the generation of new theory in a variety of
innovation-related contexts. Topic modeling provides considerable advantages over traditional
methods such as counts of patent filings or subsequent citations, which rely on existing
classification methods that were not designed to capture novel and emergent ideas. By directly
leveraging the cognitive content of texts (such as patents or papers), topic modeling augments
traditional measures of impact in knowledge fields. Furthermore, by separating measures of
impact from those of knowledge itself, topic modeling has advanced theory by empowering
researchers to invent more precise means to empirically test competing theoretical mechanisms.
In the bigger picture, these uses of topic modeling may help scholars address longstanding
questions in the management literature by conceptualizing the role of novelty with institutional
logics (Thornton et al., 2012), or delineating the roles of innovation and boundaries with
paradigms (Kuhn, 1996).
Developing inductive classification systems. Management researchers routinely use
topic modeling to develop inductive classification systems. Such systems are particularly
important in a variety of theoretical research streams, including studies of competitive dynamics
and optimal distinctiveness (Deephouse, 1999; Zhao et al., 2017), and the evaluation of risk
factors in corporate disclosures to investors (e.g., Fama & French, 1993). More generally, these
research streams are exploring classification as shared structures of meaning that are not
formally materialized. For example, studying institutional logics (Thornton et al., 2012) or

26
implicit understandings of early industry structure (Forbes & Kirsch, 2011) requires researchers
to develop inductive understandings of shared meanings that have categorical imperatives.
Researchers in each of these traditions who seek to identify categories of meaning in text face
challenges of analyzing large quantities of data without introducing researcher bias. Our analysis
reveals six topics in this subject area: understanding dynamics of meanings and networks in
knowledge fields (#34), understanding how categories affect competitive dynamics (#18),
understanding the relationships between risk and investment (#31), inducing underlying
meanings associated with cultural events (#32), and classifying sets of data and consumers (#4).
The first topic reveals how researchers use topic modeling to compare hidden meaning
structures in knowledge fields with networks of relationships among articles, journals, scholars
and citations. One approach has been to track the development of a journal or field by combining
historical topic modeling analyses with bibliometrics and authorship networks (Cho, Fu, & Wu,
2017; Wang et al., 2015) to confirm field-level insights using patterns of dominant topics while
rendering “hidden structures and development trajectories” (Antons et al., 2016, p. 726). This
approach has been applied in science to track the rise and fall of meanings within a journal
(Antons et al., 2016; Wang et al., 2015). For instance, Antons et al. (2016) used a semi-
automated topic model combining both inductive (machine) analysis and abductive (human)
labeling and generalization to add fine-grained detail to prior reviews of literature in the Journal
of Product Innovation Management. Their topic model revealed latent meaning structures not
identified in earlier reviews because the journal’s interdisciplinary character made it difficult to
identify and properly assess the breadth of papers published during its 30-year history.
A major benefit of Antons et al.’s (2016) approach is the ability to compare and contrast
content according to classification schemes in the field and then induce categories of topics.

27
They first applied the topic model analysis using LDA. After employing methodological best
practices and ensuring inter-rater reliability across 14 researchers, they clustered related topics
into six semantically-meaningful groups, including new ones the authors identified and labeled
(once again, inductively) in correspondence with the interpretation and theory-generation stages
depicted in Figure 3. The authors then made an abductive, conceptual link to disciplinary
trends—that is, they modeled “topic dynamics” by creating a weighting scheme. Finally, the
authors combined this human-centered approach with a final and more automated deductive
move, regressing topics that appeared more frequently than the median topics (those with a topic
loading greater than 10%) for each year of their analysis, tracing topic development by
comparing each of the topics against the mean, and in a final abductive iteration, classifying
them according to trajectory shape (“hot,” “cold,” “revival” and “evergreen”). The result is a
large-scale, many-to-many classification scheme across the entire study period that serves as a
comprehensive semi-automated literature review, balancing meaningful knowledge categories
with abductively rendered topics.
In another form of rendering in the classification of science, scholars have used topics as
intermediate artifacts to perform social network analyses of authorship behavior. Cho et al.
(2017) used topic modeling to augment co-authorship network data from 25 marketing journals
over a 25-year period. Building on the work of Wang et al. (2015), who used topic modeling to
map topic usage over time in the Journal of Consumer Research to predict promising research
topics for the future, Cho et al. (2017) showed that social network analysis revealed two major
communities of co-authors, whereas topic modeling analysis revealed three. They then used
these intermediate analyses to show that communities of highly-cited papers corresponded to
heterogeneous clusters of related topics, but that the communities identified by each method had

28
different features. In combining topic modeling with network analysis, Cho et al. (2017) showed
how journals comprise the ecology of a field, but the structures constituting it (communities) can
be seen at the levels of both citations and topics. Management scholars are not alone in
employing topic modeling analysis to advance field-level bibliometric studies, as it is being
adopted in psychology (Oh, Stewart, & Phelps, 2017) and the humanities (Mimno, 2012) as well.
Topic modeling has thus provided scholars with a way to both develop new understandings of
cultural meanings and to connect those understandings with network and other structural features
of fields.
A second topic relates to the role of categories in shaping competitive dynamics.
Questions around optimal distinctiveness have long been of interest to management scholars
(Deephouse, 1999; Navis & Glynn, 2011; Zhao, Fisher, Lounsbury, & Miller, 2017), but this line
of research is contingent upon the ability to measure coherence and variation of strategic action
against the backdrop of a category. How to delineate categorical boundaries is thus a key
concern. Haans (2019) explored the optimal distinctiveness of firm positioning relative to
industry categories. He used topic modeling on texts from organizational websites to uncover the
strategic positioning of firms in Dutch creative industries. The method enabled him to calculate
both industry average and distinctiveness measures for individual firms. By using topic modeling
to induce bottom-up, positioning-based classifications, Haans (2019) was able to generate new
theoretical insights that diverged from prior research by suggesting that optimal distinctiveness
for organizations depends on the distinctiveness of other organizations. Thus, positioning-based
classification, as identified through topical analysis has strategic implications. In related work,
scholars have used topic modeling to develop important conceptual infrastructure in the form of

29
inductive classifications for research on industry intelligence and competitive dynamics (Guo,
Sharma, Yin, Lu, & Rong, 2017; Shi, Lee, & Whinston, 2016).
A third topic in this area identifies topic modeling as a means to derive categories of risk
perception in finance. Such studies build on a long history of debates about the impact of
corporate disclosures on investor behavior (Fama & French, 1993). Researchers have struggled
to classify how risk factors are communicated and perceived by companies, analysts, and
investors. In contrast to the established method of using predefined dictionaries for content
analysis to quantify risk types (e.g., Campbell, Chen, Dhaliwal, Lu, & Steele, 2014 using the
schema: idiosyncratic, systematic, financial, tax, litigation), researchers have applied
unsupervised learning methods to financial texts to inductively classify risk factors. For example,
Bao and Datta (2014) applied LDA to induce risk types from corporate 10-K forms, and then
tested these against risk perceptions of investors, advancing theory by showing that the topic
modeling-induced risk meanings better predicted investor perceptions of risk. Huang, Lehavy,
Zang, and Zheng (2017) were able to extend this analysis to inductively identify risk factors and
other economically interpretable topics within analyst reports and corporate conference calls,
providing additional insights into how analysts both discover relevant information and interpret it
on behalf of investors. In both of these papers, scholars used topic modeling to extend textual
analyses of corporate financial disclosures by moving beyond the “how” (i.e., volume, sentiment,
and length) to the level of topical meaning in terms of “what is the meaning of what is being
said.” Topic modeling thus has enabled researchers to develop better classification systems based
on the textual data being sampled.
Another topic focuses on meanings associated with cultural events that are not captured
by formal documents and artifacts. Miller (2013) used topic modeling to capture meanings

30
around the nature of violence during the Qing Dynasty in China. Instead of relying on a fixed set
of categories, the method enabled him to induce an original typology of violence based on the
Qing administrator’s perceptions of unrest. Similarly, Ahonen (2015) applied topic modeling
techniques to challenge existing theory by inductively identifying the sources of legal traditions
across countries. The author considered differences in legal language in government budgeting
legislation as a basis for distinguishing between legal traditions. Both studies offer an approach
to overcome biases associated with interpreting cultural events.
In similar articles, scholars have used topic modeling to study topic-based classifications
in patent data (Kaplan & Vakili, 2015; Suominen, Toivanen, & Seppänen, 2017; Venugopalan &
Rai, 2015). The practice of mapping knowledge structures in science is in its infancy, and the use
of topic modeling has the potential to change how scientific fields are classified (see Song, Heo,
& Lee, 2015; Song & Kim, 2013; Yau, Porter, Newman, & Suominen, 2014) since topic
modeling analyses do not perfectly correspond to formal systems of classification (Cho et al.,
2017; Kaplan & Vakili, 2015). Topic modeling analyses also may reveal insights when used in
conjunction with other forms of analysis such as citation and co-authorship patterns. As such,
topic modeling can yield more fine-grained classifications and extend classic bibliometric and
content analysis methods.
The papers we reviewed in this section map the knowledge spaces and dynamics of
academic fields. Topic modeling enables scholars to compare latent topics in particular
documents with pre-existing bodies of knowledge and quantitatively measure broad trends in
meaning, thus providing a counterpoint or corroboration of coding performed exclusively by
humans. Because topic modeling is a rendering process based on human and algorithmic efforts,
employing it to map knowledge spaces uncovers latent classification systems that may or may

31
not overlap with more formal classifications. Our review of papers in this subject area has
resulted in the discovery of new concepts that can be used to better understand phenomena in a
variety of management research streams.
Understanding online audiences and products. For the last two decades, management
theorists have been particularly interested in understanding how audiences evaluate firms and
products in research on cultural entrepreneurship (Martens, Jennings, & Jennings, 2007; Navis &
Glynn, 2010, 2011), status (Podolny, 1993), categories (Hannan et al., 2007; Zuckerman, 1999),
and now, with the expansion of the Internet, understanding how these dynamics may change in
online contexts (Mollick, 2014). These scholars have sought to understand the deeper patterns
and meanings of producer communications and theorize audiences’ reactions (e.g., Cornelissen
et al., 2015). Nevertheless, isolating nuances both in the meanings of sensegiving
communications (e.g., about products) and the responses of heterogeneous audiences remains
difficult.
Topic modleing modeling has been taken up by researchers—particularly in marketing—
to analyze the cognitive content of online discourse about products and the behavior of online
consumers as audiences. This subject area of understanding online audiences and products has
emerged out of four topics: the nature of online consumer profiles (#12), online consumer brand
recognition and preferences (#23), online customer evaluations and responses to them (#29), and
enhanced topic modeling techniques on products and audiences (#13).
The first topic, the nature of online consumer profiles, has been advanced by
conceptualizing consumers based on the clicking patterns of different online groups (Trusov, Ma,
& Jamal, 2016), the network of related brands and brand tags clicked on by consumers (Netzer et
al., 2012), and communities of consumers defined based on common virtual market participation

32
(e.g., portals) or similar patterns of geo-location markers (Zhang, Moe, & Schweidel, 2017). In
these studies, topics were rendered not just from a “bag of words” across a corpus of documents,
but from a “bag of behaviors” across a corpus of activities. This conceptual pivot maps roles to
“topics” of behaviors. For example, click patterns for a group across diverse products/services
during a particular time period offer unobtrusive measures of both a latent set of consumer
profiles and their associated behaviors. Marketing studies using topic modeling have also
uncovered evaluations by consumers in new ways. For instance, Zhang et al.’s work (2017) on
elite universities revealed that the willingness to tweet—and, even more importantly, retweet—
about topics associated with a university reinforces the elite university status hierarchy.
Ironically, the most elite of the elites receive more tweet outs and retweets, not only from their
own members, but also from members of other universities. Management scholars interested in
categories (Durrand & Paolella, 2015; Vergne & Wry, 2014) and communities (Marquis &
Davis, 2007) might use these re-conceptualized online consumer communities to broaden
theorization and measures of their core constructs. Scholars might also use online endorsements
(clicks and tweets) to complement other forms of analyst assessments (Giorgi & Weber, 2015;
Zuckerman, 2001).
A second topic is online brand recognition and preference. Here, scholars conceptualize
brands not just as specific offerings with cachet, but as the associated networks of audiences
linked to those products along with the sets of user-generated tags employed by audiences to
identify brand groups. For example, Nam, Joshi, and Kannan (2017) used topic modeling to
render representative topics based on user-generated “social tags” from the shared bookmarking
service Delicious. They then examined how Apple customers linked and endorsed Apple
products via product tags, such as, “mac,” “phone,” and “Apple,” all of which were linked to

33
“Apple Corporation.” The brand in its fullest form (Apple), then, was the overall network of
linked tags used by customers. Similarly, Netzer, Feldman, Goldenberg, and Fresko (2012) used
car brand clicks on the online forum Edmonds.com to identify co-occurring words in topics
about different car brands. The clusters of words (topics) revealed overlaps, evolving brand
clusters, and “semantic networks” (i.e., meaningful text-based attributes) that differentiated
brands. In addition, Netzer et al. (2012) were able to anticipate brand switches within and across
these topic-based networks. They did so by studying changes in discussions about and
associations among brands in these topic networks (also see Tirunillai & Tellis, 2014). These
rendering moves do not differ significantly from management theory approaches to fashion and
design (Dalpiaz, Rindova, & Ravasi, 2016) and exemplar categories (Zhao, Ishihara, Jennings, &
Lounsbury, 2018); management scholars working in this vein might broaden their
understandings of how meaning is associated with brands and use topic modeling to augment
their measures of templates and categories. In addition, given the association of brand and
identity (Navis & Glynn, 2010; Raffaelli, 2018), management scholars might use group brand
identification (as measured by topic preferences) to track identity formation and evolution.
A third topic focuses on the dynamics of influencing online consumers, or in other words,
how agency is exercised online and with what effects. Marketing scholars, by and large, believe
that online consumers are more difficult to understand and influence because they are
decentralized, diverse, and switch often. Research identified as related to the topic of online
consumer responses suggests that learning adjustment is due to latent structural modifications
around topics captured by analyzing online forum data. For example, Puranam, Narayan and
Kadiyali (2016) used topic modeling to analyze all New York City restaurant reviews before and
after the implementation of a regulation that required posting calorie counts; their results

34
demonstrate a shift in online consumer evaluations, and in their view, food consumption patterns
in New York City. More recently, Wang and Chaudhry (2018) examined online hotel ratings,
and the effects of managers’ responses to posive and negative customer reviews. They used
LDA to generate a measure of response tailoring by comparing the content of managers’
responses to a baseline value. Highly tailored managerial responses to negative reviews were
considered by customers to be a form of high-quality complaint management; in contrast,
tailored responses to positive reviews were considered to be overly promotional (hence,
backfired on management). The use of topic modeling techniques to capture consumer
evaluations and adjustments is of interest to management scholars engaged in cultural analysis
and neo-structuralism research (DiMaggio, 2015; Lounsbury & Ventresca, 2003; Mohr &
Bogdanov, 2013), because a bedrock assumption in these culture-oriented approaches is that
agency is less observable and more distributed. Topic modeling of online reviews across
audiences can also help capture actor adjustments around latent structures (e.g., see Hannigan et
al., 2019; Heugens & Lander, 2009). In addition, longitudinal, affect-based topic modeling might
enrich studies of performance adjustment (Greve, 2003), anchoring (Ballinger & Rockman,
2010), and event analysis (Morgeson, Mitchell, & Liu, 2015).
A final topic in this subject area is focused on improving topic modeling of online
audiences and products to capture nuances of communication and audience responses (#13). The
groundbreaking and oft-cited work by Lee and Bradlow (2011) regarding automated online
reviews has several features that have become norms for rendering with topic modeling, such as
using triangulation (e.g., with k-means clustering and MDS), mapping structures, thinking about
“fit” with algorithms, and examining change over time. Recently, Guerreiro, Rita, and Trigueros
(2016) and Jacobs, Donkers, and Fek (2016) introduced correlational topic models, sentence-

35
based models, and hierarchical topic models to demonstrate the utility of using some supervision
and structure in topic model rendering. Along similar lines, Büschken and Allenby (2016) used
sentences and phrases rather than words as inputs for LDA to show that topics based on them
might exhibit less change (i.e., be “sticky”) over time. Because management researchers are
curently interested in understanding the interface of such methods and derived topics and
meaning (DiMaggio, 2015; Schmeidel et al., 2018), Büschken and Allenby’s (2016) work poses
an interesting rendering question for management researchers: Is stickiness a product of using
sentences (the method) or is it due to linguistic meaning being constructed at the sentence-
(rather than word-) level by online consumers?
To summarize, using topic modeling to analyze online audiences and products enables
management scholars to think more deeply about the nature of online audiences (e.g., as click-
based profiles, virtual networks, and computer-mediated communities); to reconceptualize
products as distributed brands tied to evolving individual and category identities; and to capture
the more subtle means by which audiences evaluate online products, and correspondingly
understand how organizations might adjust in real time to those evaluations. In addition, the
refinement of topic models of online audiences creates modeling standards for other topic
modeling research, and encourages scholars to think more deeply about the meaning given to
products by online audiences.
Analyzing frames and social movements. Topic modeling also has been used to analyze
frames and understand the dynamics of social movements. Management scholars have long been
interested in symbolic management (Zajac & Fiss, 2006; Zajac & Westphal, 1994; Zott & Huy,
2007), such as understanding how investors respond to organizational framing efforts (Giorgi &
Weber, 2015; Rhee & Fiss, 2014), theorizing the political dynamics associated with different

36
framing strategies within firms (Kaplan, 2008b), and understanding the dynamics of social
movements (e.g., Benford & Snow, 2000). This research requires scholars to identify frames—
epistemological devices that actors use to organize experiences by answering the question posed
by Goffman (1974, p. 8): “What is it that’s going on here?”
Topic modeling methods have helped scholars expand theoretical boundaries in this area
by providing an empirical method for inductively uncovering latent frames and then
understanding the dynamics associated with frame proliferation and effectiveness. Our topic
modeling analysis revealed four topics in this subject area: understanding how frames influence
political processes (#27); the relationship between frames, context, and audience (#6);
understanding field-level relationships between organizations, discourses, and strategies (#17);
and social movement strategies, networks and actions (#11).
The first topic relates to how frames influence political processes. Frames enable actors
to “render what would otherwise be a meaningless aspect…into something that is meaningful”
(Goffman, 1974, p. 21). Scholars are particularly interested in the often political and contested
dynamics associated with framing (e.g., Fiss & Hirsch, 2005; Kaplan, 2008b). An exemplar
article showing how topic modeling can contribute to this research stream is Fligstein et al.’s
(2017) study of the Federal Open Market Committee’s decision-making processes in public
meetings. Specifically, they sought to develop a theory to explain how the committee failed to
appropriately perceive the risks to the economy in the months leading up to the financial crisis.
In addition to confirming the existence of macroeconomics as a master frame, their topic
modeling approach revealed the existence and application of a banking frame and a finance
frame. By focusing on the specific events—the housing bubble and the financial crisis—the
researchers were able to track which frames came to dominate Fed committee discussions at the

37
time of each event. The authors thus used topic modeling to develop a theory that explains how a
predominant frame can blind actors involved in decision-making processes.
A second topic explores the relationship between frames, context, and audience. Actors
use distinct frames to advance their interests (Kaplan, 2008b) and seek to create effective frames
through mechanisms such as frame alignment (Snow, Rochford, Word, & Benford, 1986) or
frame resonance (Snow & Benford, 1988). In an exemplar article, Levy and Franklin (2014) used
topic modeling as a means of identifying distinct discursive frames. Specifically, they used a
study of political contention in the U.S. trucking industry regarding hours of service to
inductively analyze the frames that emerged from a study of comments on a public website. They
were able to use topic modeling to uncover distinct differences between individual and
organizational uses of frames in the debate, showing how different parties used different frames
to promote their interests. Uncovering nuanced distinctions in framing content deployed by
different parties over time can help researchers generate new theory about the influence of
communication content and techniques on political processes.
The third topic relates to research on field-level relationships between organizations,
discourse, and strategy. Specifically, to understand framing effects, it is often necessary to move
beyond the content of a specific frame. To illustrate, Bail, Brown, and Mann (2017) explored the
relationship between conversational and emotional styles in advocacy work—seeking to
incorporate sentiment analysis into our understanding of frames. The authors used topic
modeling to classify the types of topics raised by autism advocates and used LIWC to capture
sentiment and bias in normalized spaces. This unique combination of topic modeling and LIWC
sentiment analysis enabled them to reveal the cognitive and emotional “currents” running
through advocacy groups, and to show how the ability to “dispatch messages that contribute to a

38
phase shift [between emotional and cognitive-focused communication]” ultimately leads to more
effective results (Bail et al., 2017, p. 1205). Thus, topic modeling has enhanced our
understanding of frame effectiveness in the context of broad field-level relationships between
organizations, discourse, and strategy.
Similarly, the fourth topic relates to researchers’ attempts to understand the relationship
between social movement strategies, networks, and actions. For example, Almquist and Bagozzi
(2017) sought to understand the network relationships between radical environmental activists in
the United Kingdom. Based on a longitudinal corpus of a radical social movement’s texts, they
identified the centrality of network ties and then used structural topic modeling to locate the
groups and the positions they took on various radical issues, thereby enabling them “to evaluate
whether the presence of a given group tie (or cluster member) significantly increases the
attention dedicated to a given topic” (Almquist & Bagozzi, 2017, p. 26). By combining structural
topic modeling and network analysis, the authors were able to classify subnetworks of actors to
develop a better theoretical account of the discursive actions and network relationships of social
movements by mapping unseen or hidden ties. Put another way, topic modeling generates
theoretical artifacts that facilitate researchers’ efforts to connect the content of communications
with other theoretical constructs.
In summary, topic modeling provides several benefits that have led to significant
theoretical advancements related to frames and framing. First, topic modeling has helped
researchers strengthen their understanding of frames. For example, scholars can use topic
modeling to track the prominence of researcher-derived high-level frames for large corpora over
an extended period of time. Additionally, the algorithmic nature of topic modeling approaches
ensures the replicability of identified frames. Second, the inductive nature of many topic

39
modeling techniques enables the discovery of unanticipated frames and audiences that use them,
providing a powerful opportunity for scholars to generate new theory. Specifically, topic
modeling methods enable researchers to understand the dynamics associated with the co-
presence of competing voices within a single text (i.e., heteroglossia, Bakhtin, 1982), which
provides researchers with a way to study multiple competing or collaborative frames. Finally,
topic modeling facilitates the creation of new theory since it produces theoretical artifacts that
can be paired with other forms of analysis such as sentiment analysis or network analysis.
Understanding cultural dynamics. Management scholars have sought to leverage
psychological and sociological research on culture—“the interaction of shared cognitive
structures and supra-individual cultural phenomena (material culture, media messages, or
conversation, for example) that activate those structures” (DiMaggio, 1997, p. 264)—to explain
diverse phenomena. For example, in research on institutional logics (e.g., Thornton, Ocasio, &
Lounsbury, 2012), strategic action fields (e.g., Fligstein & McAdam, 2011), and professions
(e.g., Abbott, 1988), scholars have theorized the evolution and impact of cultural meanings at the
level of an institutional field. In research on organizational culture (e.g., Hatch, 1993) and
organizational identity (e.g., Gioia & Thomas, 1996), scholars have theorized the evolution and
impact of cultural meanings at the level of the organization. In research on cultural
entrepreneurship (e.g., Lounsbury & Glynn, 2001, 2019; Martens et al., 2007) and institutional
work (e.g., Lawrence, Suddaby, & Leca, 2009), scholars have attempted to understand how
individuals leverage cultural material to achieve strategic objectives. In all of these areas,
researchers have attempted to theorize both the dynamics of cultural influences and the evolution
of cultural concepts.
Overall, this research on culture has faced significant challenges. One such challenge

40
relates to the measurement of cultural constructs. For example, scholars have defined
institutional logics as “the socially constructed, historical pattern of material practices,
assumptions, values, beliefs, and rules by which individuals produce and reproduce their material
subsistence, organize time and space, and provide meaning to their social reality” (Thornton &
Ocasio, 1999, p. 804). But in empirical studies, it has been harder to specify them. A second
challenge is to understand the temporal dynamics associated with culture. For example, in
cultural entrepreneurship research, scholars attempt to understand how entrepreneurial
organizations are able to legitimate a new market category over an extended period of time (e.g.,
Navis & Glynn, 2010). Researchers also attempt to connect cultural meanings with events and
actions, for example, by connecting the content of organizational discourse with changes in
organizational networks and broader social discourse (Bail, 2012).
Scholars have used topic modeling methods to push the boundaries of our understanding
of such cultural dynamics. Our analysis reveals five themes in this research: understanding the
professionalization of a field (#2), using topic modeling to analyze big data to understand
cultural trends (#5), understanding dynamics associated with literary meanings (#9),
understanding how cultural meanings change over time (#19), and understanding the evolution
of cultural trends (#28). Topic modeling has enabled scholars to generate novel theory by
providing an operational means to identify cultural concepts and then trace the evolution of those
concepts over time and across different locations of social space.
The first topic in this area revolves around developing new theory about the
professionalization of fields. Specifically, Croidieu and Kim (2018) theorized the rise of alternate
fields and quasi-professions by studying the emergence of U.S. wireless radio broadcasting field
and the “lay professional legitimation” of amateur radio operators from 1899 to 1927. To

41
understand the legitimation process for amateur operators, the authors had to gather a wide,
diverse constellation of documents from various archival sources: U.S. government regulations,
radio operators from the era, radio corporations, and the New York Times. They analyzed the
distribution of topics over time and by audience to determine the meanings of those patterns
using historical (or case) records. This process enabled the authors to identify first- and second-
order mechanisms by period. They paired topic modeling of diverse archival materials with
standard historical reading and complementary content analysis to create and defend a theoretical
account of professionalization based on historical data.
A second topic focuses on how big data can be used to understand cultural trends. These
articles describe and illustrate nuances of the processes scholars use to extract meanings from
large corpora. For example, Wagner-Pacifici, Mohr, and Breiger (2015) summarized a special
issue in Big Data & Society on assumptions of sociality that synthesized the results of several
other subjects. First, they highlighted the importance of recognizing that big data methods,
unreflexively applied, can lead to biased results. Second, they discussed the importance of the
interpretive role of analysts who use big data and related methods to generate theory. Third, they
emphasized how big data methods require a move away from traditional deductive science,
highlighting their inherently inductive and abductive nature. Finally, they showed how analyzing
big data requires scholars to ask fundamental questions such as “What is a thing? What is an
agent? What is time? What is context? What is cause?” (Wagner-Pacifici et al., 2015, p. 5). Thus,
scholars must reflexively consider the cultural implications of studying big data.
Interestingly, in sociological research that has provided analogical inspiration for
management scholars, Mohr and Bogdanov (2013) used topic modeling to analyze literary
meanings. In the humanities, Tangherlini and Leonard (2013) introduced a technique called sub-

42
corpus topic modeling to compare canonical texts with broader literature and societal discourse.
Specifically, they used the technique to “develop a well-curated topic model of a sub-corpus”
and then used “the ensuing model to discover passages from the large, unlabeled corpus”
(Tangherlini & Leonard, 2013, p. 728). To illustrate the utility of their method, they showed how
topics associated with Charles Darwin’s intellectual ideas penetrated “into the broader literary
world” (Tangherlini & Leonard, 2013, p. 735). They thus used topic modeling to understand
topics associated with well-known texts and then applied the outputs to analyze other, less well-
known cultural meanings.
Another evident topic focuses on how cultural meanings evolve over time. An example of
this can be seen in the work of DiMaggio et al. (2013), who identified the frames invoked and
crafted by news outlets in their coverage of the public controversy surrounding the U.S.
government’s support of artists and art organizations. The authors rendered corpora using data
from five mainstream media outlets; after applying unsupervised LDA to isolate and link topics,
they inductively identified different frames. Their results reveal not only the differences across
frames by time period, but also how a single text produced by these media outlets might use
multiple frames. Applying a fractional multinomial logit analysis, they calculated the expected
relative prominence of topics based on their LDA analysis. By further aggregating those topics
into particular topic groupings, then classifying them as conflict or comparison frames, they were
able to reveal the likely link between the relative increase in conflict topics that accompanied the
growing sentiment against public funding for U.S. arts organizations starting in the 1980s. The
authors thus used topic modeling to identify different frames of cultural meaning in the public
sphere and then showed how these meanings changed over time.
A final topic looks at the impact of cultural meanings on societal actions. For instance,

43
Marshall (2013) sought to understand the evolution of cultural trends by contrasting how
different academic theories of demography unfolded over a 60-year period in Great Britain and
France. Specifically, she used correlated topic modeling (to account for the assumption that
topics in her corpus might be correlated across documents) to understand how concepts
associated with fertility were understood (and unfolded) differently in different cultural contexts.
She used topic modeling to identify topics, measure the prevalence of those topics in the corpus,
and then connect those topics to the dominant theories of demography in effect during that time.
The topic modeling analysis enabled her to identify differences between the responses of French
and British academics to changing demographics during the study period. Topic modeling thus
enables scholars to trace the evolution of cultural trends by connecting the prevalence of themes
in discourse to historical events.
Overall, topic modeling has provided management scholars with a new methodology for
generating novel insights about cultural dynamics. First, topic modeling provides a means to
develop an unbiased understanding of the prevalence of distinct cultural concepts over an
extended period of time, thereby enabling scholars measure cultural concepts more precisely.
Second, topic modeling enables scholars to compare a well-known subset of knowledge to
broader corpora that might reflect that knowledge structure more generally, thereby enabling
scholars to develop new theories and link constructs that previously had been difficult to
connect, both empirically and theoretically. Similarly, topic modeling enables researchers to see
how different meanings within the discourse surrounding a particular topic exist and shift over
time. Finally, topic modeling can connect shifts in discourse to broader cultural trends.
NEW TRENDS RELATED TO TOPIC MODELING AND RENDERING

44
Many new trends in management and computer science research are relevant to
management scholars’ use of topic modeling to render corpora, topics, and theoretical artifacts
(see Figure 2). Each trend within a rendering process has a unique trajectory that is important to
discuss and respect. For instance, some trends broaden specific rendering processes (e.g.,
creating corpora), whereas others deepen them (e.g., fitting topic models). Trends also involve
some of the aforementioned management subject areas. In this section, we discuss not only
trends, but also their implications for rendering and building management knowledge.
Trends in Rendering Corpora
As management researchers embrace approaches that move beyond dictionary-centric
content analysis, corpus selection becomes an even more critical step in topic modeling research.
Recent methods papers on text analysis reveal a broad effort to engage more closely both with
computational linguistics and NLP (Kobayashi et al., 2018; Schmeidel et al., 2018). These efforts
were precipitated by an important shift toward conceptualizing corporal dimensions to enable
comparison.
Corpus linguistics. Within management, this trend of engaging with computational
linguistics is most evident in a recent special issue of Organizational Research Methods
(Tonidandel, King, & Cortina, 2018) on big data and modern data analytics. This special issue
demonstrates the arc of pre-processing corpora as a precursor to higher order text analyses with
big data (Kobayashi et al., 2018; Schmiedel et al., 2018). However, many of these pre-processing
techniques were highlighted several years earlier by Pollach (2012), who pointed management
researchers to a branch of linguistics known as “corpus linguistics” to show how word patterns
can lead to meaningful insights by virtue of the corpora in which they appear. Techniques for

45
analyzing corpora themselves—both qualitatively and quantitatively—include word frequency
lists, keyword-in-context searches, the comparison of corpora, word collocations, and statistical
methods for assessing word-frequency patterns.
Pollach (2012) originally positioned corpus linguistics techniques as methodological
innovations for content analysis. In very recent work, Kobayashi et al. (2018, p. 1) took a
broader approach, suggesting that such pre-processing considerations represent a “fundamental
logic” of mining “text data.” As part of that mining, papers in this vein have stressed the
imperative of pre-processing as “wrangling” text data into a corpus (Braun, Kuljanin & DeShon,
2018). Schmiedel et al. (2018) have laid out some steps that recognize the fundamental
importance of data collection and cleaning in topic modeling analysis. Theoretically speaking,
these papers draw on core ideas from linguistics, such as the famous distributional hypothesis
(Firth, 1957)—that is, “words that occur in the same contexts tend to have similar meanings”
(Turney & Pantel, 2010, p. 142). Inferring meanings, in other words, depends on the context
created by the corpus. As a result, these recent papers are raising the bar in terms of the level of
sophistication and reporting standards required for scholars who use topic modeling and other
text analysis methods.
In fact, we built our rendering process on the insight that corpora curation has
implications for theoretical work because meaning is inferred from context. A source corpus
begins as natural language, which can be messy and thus requires selecting and trimming. These
two steps standardize documents, which then enable topics in the corpus to be rendered at a
higher level of abstraction. Moretti (2013) called this “distant reading,” where a corpus can be
fully and adequately represented in terms of topics. Sharpening this reading requires iteration; for
this reason, our rendering process has an arrow pointing back from rendering topics to rendering

46
corpora. The trends we identified in pre-processing point to the adaption of techniques from
corpus linguistics for the purposes of corpus curation, thereby expanding the toolkit for
rendering.
NLP. Innovations in NLP are advancing how scholars prepare and preprocess the words
in corpora. NLP research highlights two key concerns: first, as the base unit of meaning, a token
(a word, parts of words, or phrase combining words) is a function of grammar; and, second,
structures of grammar are embedded in sentences, which have co-dependencies across words and
paragraphs within a document. Uttered meanings correspond to parts of speech. For example, the
meaning of the token Google changes based on whether it is a noun (i.e., referring to the
company or software), or a verb (i.e., referring to use of the search engine), and can be referred
to in a similar manner through a pronoun in a subsequent sentence. Thus, a token as a unit of
meaning may be a word or multiple words (i.e., a phrase) (Chomsky, 1956).
NLP research suggests that latent meaning in texts can be captured by bigrams, or two-
word units rather than individual words, as in the standard “bag of words” approach (Manning,
Raghavan, & Schütze, 2010). Some management researchers have therefore shifted the unit of
analysis to a “bag of sentences” (Bao & Datta, 2014; Büschken & Allenby, 2016). Determining
the boundary of analysis is technically tricky. For example, because a sentence break is not just a
function of searching for the full stop character (i.e., “.”), researchers have developed NLP
methods to determine sentence boundaries in a common task called sentence segmentation (Kiss
& Strunk, 2006). Moreover, advanced deep learning algorithms (e.g., neural networks) are being
introduced that go beyond “bag of words” approaches altogether to consider syntactic position
and context when identifying linguistic structures such as constituency and dependency parsing
representations (Manning et al., 2014). Deep learning is an unsupervised algorithm that can be

47
trained on large text corpora to “learn” latent structures, including semantic compositionality
(Socher et al., 2013) within texts (or other kinds of data) that can then be used for explanatory or
predictive purposes.
Additional advances have improved the precision of identifying tokens. For example,
mentions of individual actors may be standardized by employing NLP technologies such as
Named Entity Recognition (Mohr et al., 2013) and co-reference resolution (Manning et al.,
2014). The former is an NLP method that can automatically identify entities based on their
appearance in texts and can annotate analytical codes as actors, organizations, and countries. The
latter is an NLP tool that can extend Named Entity Recognition to pronouns and other references
to entities across sentences. Standardizing entities to resolve ambiguities inherent in manifested
natural language facilitates machine-based reading.
Approaches to making such transformations are particularly salient in topic modeling
because this trimming determines the token unit upon which topics are established (Schmiedel et
al., 2018). These decisions regarding rendering corpora have theoretical implications. The NLP
methods discussed here are largely inductive tools, with machine learning algorithms annotating
texts. While inductive methods have become more widely accepted in management journals,
there is still considerable risk of over-fitting findings to the data if scholars generalize too
quickly (i.e., engage in “theoretical over-fitting”) (Tchalian, 2019). Thus, researchers must
continue to check the validity of such annotating.
Non-Western languages. Another new corpus-rendering trend that touches upon these
developments in corpus linguistics is the treatment of languages that are structurally dissimilar to
most Western languages—in particular, languages without spaces between words (or, scriptio
continua), including many Southeast Asian writing systems (e.g., Thai, Burmese, Lao) and those

48
that use Chinese characters (i.e., Chinese and Japanese). Treatment of these languages is not
straightforward. For example, each Chinese clause can be recognized as a group of characters.
Each Chinese character corresponds to a syllable; although some characters represent individual
(i.e., one-syllable) words, many words consist of more than one character. These linguistic
features make pre-processing necessary to ensure effective topic modeling and theorizing,
thereby enabling the algorithm to identify the tokens that comprise the texts.
The traditional content-analytical method of using pre-set dictionaries to match characters
with possible words in the corpus confronts computational problems, and the permutations and
ambiguities of language often lead to poor results. Customized dictionaries improve fit, but still
yield substantial inaccuracies (Allen et al., 2017; Slingerland, Nichols, Neilbo, & Logan, 2017).
Today, statistical and machine learning models are complementing, if not replacing, pre-set
dictionaries. These models build internal lists of words by training algorithms through iterative
learning. This training can be performed using extant language libraries (e.g., the People’s Daily
Language Library) to segment unknown texts.
The introduction and development of these methods has opened the door to employing
topic models to investigate a wide range of novel data sources and cultures. For example, Huang
et al. (2015) used topic modeling to analyze one of China’s biggest online social network
platforms, Weibo, to track the real-time ideation process of suicide, which is traditionally
assessed by surveys and interviews and thus suffers self-reporting and retrospective biases. Their
approach has shed new light on future studies of various ideation processes such as
entrepreneurial ideation.
Such word segmentation processes also make comparative analysis and theorization of
multiple-language corpora feasible. In particular, with appropriate pre-processing, topic models

49
can be used to analyze the diffusion and translation of new ideas, frames, and categories crossing
national borders. For example, the cross-national diffusion of CSR has attracted scholarly
attention (Kim & Bae, 2016; Lim & Tsutsui, 2012). But identifying the extent to which CSR has
been locally translated and innovated would require fine-grained analysis of multiple-language
corpora, which topic modeling can facilitate. Because the topic outputs from non-English
corpora must be translated into their English equivalents to be used in comparison and
theorization, and because the cultural context still matters for those identified topics, such
comparative projects are best developed by teams with at least one researcher who knows the
language and culture and can apply that knowledge to help validate the rendering of the corpora.
Summary. New trends in rendering corpora hold great promise for addressing the
technical and theoretical limitations of current topic modeling approaches. They show that
corpus selection as well as lemmatizing and other forms of corpus preparation have theoretical
implications, and therefore must be explicitly discussed in methods sections of papers, likely
under the aegis of “data pre-processing.” The use of foreign languages only magnifies these
challenges, just as they do in any form of archivalism applied to other cultures.
Trends in Rendering Topics
Researchers are continuing to refine how topics are rendered in an effort to manage the
degree of supervision required and how fit can be defined. In Figure 2, we show how the
rendering of topics revolves around the criteria for identifying robust, applicable topics (i.e.,
around supervision and fit criteria). Supervision and fitting, in turn, depend on the form of
theorizing taken—inductive, abductive, or deductive—with induction aligning with less
supervision and fitting than deduction.

50
Integrating topic rendering with other approaches. Many scholars today are finding
that topic modeling works best when integrated with other methods of analysis, which has
implications for the rendering of topics. One recent style of work covered by labels such as “big
qual” (Davidson, Edwards, Jamieson, & Weller, 2019) and “RiCH (Reader in Control of
Hermeneutics)” (Breiger, Wagner-Pacifici, & Mohr, 2018) gives the interpretive human reader
primacy, but leans on the affordances of computational tools for forming rich representations of
topics. Other styles in recent work integrate topic modeling with more traditional deductive
methods (Haans, 2019; Hankammer, Antons, Kleer, & Piller, 2016; Roberts et al., 2014), where
topics are rendered according to a logic of variable coherence. Topic modeling in these
correlational analyses seems to rely on a parsimony principle, where topics are presented in
papers as tables with applied labels and fewer than 10 highly associated words per topic (i.e.,
Schmiedel et al., 2018). Our reading of this trend reveals that the dominant method in the
research design affects how topics are rendered.
Recent trends in topic modeling within management research have also shifted attention
toward alternate ways of capturing latent patterns to reveal new (sometimes provisional)
meaning structures that change over time. The LDA-based analyses we reviewed in this paper
mostly followed a pattern of rendering one set of topics in a corpus. Through iterative steps in
the rendering process, Hannigan et al. (2019) found that a key topic in a scandal’s media
coverage was changing due to the disclosure of a social control agent’s judgements of
wrongdoing. To overcome this challenge, they split their corpus in two, rendering topics across
each sub-corpus. They used the word-topic matrices from both models to find comparable topics,
which they subsequently used as independent variables representing media effects of a scandal in
event history models at different time periods. Similar efforts to periodize data can be seen in

51
work by Croidieu and Kim (2018). We see such efforts as contextualizing topics in ongoing
theoretical concerns.
As another example, Cho et al. (2017) embedded topic modeling with other commonly
used methods of conducting a literature review. The concept of topic was used to approximate an
“author community” of researchers exhibiting certain topics prominently in their work. This
framing affected the logic of how they rendered topics. They rendered latent author communities
using topic modeling against those derived using bibliometric network analysis to show
similarities and differences in approaches, but this comparison governed the validity of topics
rendered. Alternative analytical approaches that help generate theory (Bail, 2012; Kennedy,
2008), especially emergence processes, also promise the ability to better articulate latent patterns
to reveal hierarchical linguistic structure (Mohr et al., 2013). Therefore, the rendering of topics is
part of the overall theory generation process itself.
Structural topic modeling. Just as LDA disrupted latent semantic indexing (LSI),
scholars are attempting to modify LDA by improving fit algorithms and making it more
structured and systematic. One major development is structural topic modeling (STM) (Bail et
al., 2017; Roberts et al., 2014; Schmiedel et al., 2018), which extends LDA by incorporating
meta-data about documents, such as who wrote each text and when or where they were written.
This information can be re-applied to the topic estimation procedure and help improve model fit.
In so doing, STM enables researchers to identify relationships not just between topics and
documents, but also between the producers of documents and the texts and topics. It can be used
in a linear regression framework to analyze specific meta-data (as covariates) to identify
statistically significant relationships to each topic. It can also be used in mixed methods
approaches such as with critical discourse analysis to tie textual data analyzed using topic models

52
with richer qualitative analysis (Vaara, Aranda, Etchanchu, Guyt, & Sele, 2019).
In recent working papers appearing in Academy of Management Annual Meeting
Proceedings, researchers have adopted mixed STM approaches. For instance, Aggarwal, Lee,
and Hwang (2017) used topic modeling to operationalize review diversity in Yelp reviews to
show that status gains are correlated with higher-quality reviews and non-elite conformity to
those same reviews. Likewise, Karanovic, Berends, and Engel (2018) used topic modeling to
study actors’ perceptions of “platform capitalism” (Davis, 2016) in a popular online forum for
Uber drivers. Their analysis reveals consistent patterns in a large corpus representing over
120,000 forum posts and shows that drivers’ reactions can both contribute to and critically
evaluate the legitimacy of a new organizational form, despite being imposed from above.
Hierarchical LDA. Another promising extension to LDA topic modeling is hierarchical
LDA (hLDA) (Blei, Griffiths, & Jordan, 2010). While LDA traditionally requires that a
researcher set the number of topics (the k parameter), hLDA can generate the optimal number of
topics based on other researcher-defined parameters, such as the number of hierarchical levels
and number of terms per topic. While different software implementations of hLDA use different
algorithms to generate the hierarchical models, generally speaking, the hLDA algorithm
generates a set of sub-topics after identifying an aggregate topic. The algorithm then “reshuffles
the deck” by reclassifying documents or document segments into synthetic document groupings
and rerunning the algorithm for each grouping to generate additional sub-topics. The result is a
hierarchy representing the topics and sub-topics, or sub-dimensions, of the texts being analyzed.
The ability to generate a hierarchical representation of the internal structure of a
discourse can provide substantial theoretical insights. Tchalian, Glaser, Hannigan, and
Lounsbury (2019) are using hLDA to identify the competing and complementary messaging

53
efforts of stakeholders in the emergent electric vehicle (EV) industry: automobile manufacturers,
newspaper reporters, automotive experts, and government officials. The hierarchical structure of
the hLDA output is enabling Tchalian et al. (2019) to trace both the longitudinal appearance of
different topics involved with the construction of the emergent EV category and their
prominence within the discourse. This approach allows them to define the theoretical concept of
“institutional attention”—the field-level convergence that both isolates and aggregates the
various interests involved in the social construction of the EV as a market category. The
hierarchical arrangement of topics in their paper and others (e.g., Smith, Hawes, & Myers, 2014),
reveals not only the primacy of ideas over time, but also the socio-cognitive meaning structures
emphasized in cultural sociology (Mohr, 1998) and content analysis (Duriau et al., 2007), thus
highlighting the great potential of topic modeling approaches for generating novel theoretical
insights.
Summary. Advances in rendering topics have broadened topic modeling’s use by pairing
it with other techniques, and deepened its use by creating variants that structure topics (e.g.,
hLDA). Rendering topics, at least for the near future, appears sufficiently robust to work with
developments in near variants such as NLP and specific machine processing algorithms (i.e.,
“trained” algorithms in specific domains). These trends have the potential to extend the
theoretical deltas we identified in our analysis of management subject areas. However, applying
new algorithms for topic modeling and determining proper logics of fit and validity also raises
important questions about research design. For example, use of STM reinforces critical decisions
about appropriate measurement and variation in econometric based approaches, and hLDA
simply shifts a researcher’s interpretive choices from determining the number of topics to
deciding the number of levels and words per topic. These advances demonstrate that the most

54
powerful path of development in topic modeling is not to displace, but rather complement
traditional research designs by enabling the use of different approaches to abstract and measure
phenomena using text.
Trends in Rendering Theoretical Artifacts
Trends in rendering theoretical artifacts may offer the richest, most open-ended area of
development in the field. Three trends are of particular interest: delineating latent structures,
mapping new meaning, and blending AI with human supervision to generate new artifacts. Each
trend has been pursued using a range of theorizing approaches from inductive to deductive, and
each has the ability to both extend and build theory, as indicated by the iterative arrows in Figure
2.
Latent structures and the “new structuralism.” Increasingly, scholars are using topic
modeling to assess structural relations in fields (Bail, 2014; Jha & Beckman, 2017; McFarland et
al., 2013). Structural artifacts formed through rendering may enable theorists to identify new
mechanisms for uncovering organizational or institutional structures, including those flexible
enough to allow for a variety of instantiations in studies of fields (Lounsbury & Ventresca,
2003). The central thread relates to the use of topic modeling to map cultural dynamics around
social structures. A macro approach involves mapping the meaning structures that comprise
business environments (Pröllochs & Feuerriegel, 2018), knowledge profiles of firms (Suominen
et al., 2017), emerging fields (Hannigan & Casasnovas, 2019), and political issues (Kim et al.,
2018). Researchers have modeled the topics and rhetorical attributes of scientific articles, in turn
finding links between the hidden topic structure of scientific communities as “thought
collectives” and impacts on knowledge consumption patterns (Antons et al., 2018). Others have

55
identified the “backstage” influences of stakeholder groups in the sustainability movement in
higher education and have used measures of discursive distance to identify field-level coherence
(Augustine & King, 2017).
More micro approaches involve modeling the formation of social network ties using
topic-based proximity measures (Lee, Qui, & Whinston, 2016), or tracking the signatures of
content authorship using author-topic models (Rosen-Zvi, Griffiths, Steyvers, & Smyth, 2004).
Scholars are using these micro approaches to revisit a classic question in social science: How are
social structures and meanings co-constituted? Lee et al. (2016) considered the mechanism of
homophily in network formation by topic modeling texts of user-generated biographies and their
associated tweets. In turn, they found that people with similar topic vectors were more likely to
check-in to the same locations and form similar online social network ties. Rosen-Zvi et al.
(2004) used an extension to LDA to model the contents of documents and authors’ interests.
They created the “author-topic model” artifact which can be used to compare documents for
similarity and applied to automatically match paper authors to reviewers. In each of these papers,
researchers used topic modeling to render and theorize structural dimensions as artifacts.
Scholars are extending the new structuralist approach by using topic modeling to analyze
dynamics of culture and meaning (Lounsbury & Glynn, 2019; Mohr & Bogdanov, 2013). The
simultaneous rendering of topics and contents of identified topic clusters reveals how social
structure and meanings can be co-constituted at the field level. An example of a classic approach
in this style of work is an exploration of “grass-fed beef” (Weber, Patel, & Heinze, 2013) as a
construct that conveys particular meanings and describes the evolving structure of a market.
Topic modeling enables social structures and meanings to be studied in new ways. Hannigan and
Casasnovas (2019) used topic modeling and Named Entity Recognition to map the co-occurrence

56
of actors and topics appearing in media coverage to identify the spatial and temporal
arrangements of an emerging field. Following classic works in the new structuralist tradition
(i.e., Mohr & Duquenne, 1997), Hannigan and Casasnovas created incidence matrices of topic
and actor co-occurrence and used them to generate maps of hierarchical Galois lattice structures.
These lattice artifacts are visual maps that demonstrate co-constitution by showing the nesting of
substructures formed through two modes of analysis. Mohr and Duquenne (1997) used lattices to
show how practices and meanings co-constituted institutional logics, whereas Hannigan and
Casasnovas (2019) used lattices to reveal the types of actors and topics co-constituting spatial
and temporal arrangements in field formation. Advances in relational topic modeling (RTM)
(Chang & Blei, 2009; Gerlach, Peixoto, & Altmann, 2018) that identify document networks are
also being used to render more document-based theoretical artifacts, perhaps representing
different audience perspectives. These audience perspectives, including those captured using
STM, enable latent structures among knowledge creators to be identified.
Bringing back meaning. Whilst topic modeling provides tools for extracting and
presenting constellations of words and phrases that appear in patterns across documents in
corpora, the question of whether such topics represent meaning structures is an important one
(Mohr, 1998). During the initial analytical stage, analysts interpret topics based on logics of fit
and interpretability. However, presenting topics without careful concern for theoretical artifacts
risks presenting disembodied arguments about meaning. Thus, a naive machine learning analysis
may omit important distinctions if applied crudely. An important topic modeling trend thus
centers on how to capture meaning and meaning structures.
Organizational scholars have long been interested in studying meanings, particularly in
light of recent concerns about measuring the construction and deployment of culture (i.e.,

57
Gehman & Soublière, 2017; Lounsbury & Glynn, 2019; Weber & Dacin, 2011). Whilst topic
modeling-based research promises the potential to study cultural dynamics with increased scale
and precision, scholars acknowledge that the technique must be paried with a respect for
symbolic and social boundaries (Lounsbury & Glynn, 2019; Mohr et al., 2013). For example,
Mohr et al. (2013) pointed to Burke’s (1945) classic analytical structure of the pentad to study
scenes of action. They used topic modeling and NLP to study the pentad in a corpus of U.S.
national security documents. Analytically, they used named entity recognition to map actors,
topic modeling to identify scenes, and NLP-based semantic grammar parsers to identify acts.
Other scholars have described the utility of applying related computational methods such as
semantic network analysis to contextualize topic modeling through theoretical artifacts (Carley
& Kaufer, 1993; Diesner & Carley, 2005). Combined with a concern for theoretical artifacts,
topic modeling thus opens the door to rendering modes of meaning, such as observing
connotations and denotations of an institutional field.
Blending topic modeling and AI. A third fertile area of enhancing the theoretical
artifacts built with topic modeling lies at the intersection of artifacts derived from artificial
intelligence (AI) and those derived from topic model rendering. AI and the deep learning models
on which it is built can be blended with topic models in at least two ways. First, in the class of AI
models known as “deep neural networks,” two relevant methods enable blending with topic
modeling: convolutional neural network (CNN) methods and recurrent neural network (RNN)
methods. Unlike machine learning models such as LDA that use minimal inferences about
context, these models retain more contextual information and thus are becoming increasingly
relevant for social science researchers. They are more appropriate for dealing with streaming
data such as Facebook updates and Amazon reviews, in which local contexts (e.g., prior words in

58
a word sequence) affect the position of each topic term (Jin, Luo, Zhu, & Zhuo, 2018).
Combining these methods with topic models may enable a more complex and dynamic rendering
of theoretical artifacts such as frames, logics and the latent value orientations discussed above.
When applied to large text corpora, both CNN and RNN are particularly effective in managing
the tradeoff of specificity, enabling the analysis and modeling of latent structures that better
balance under- and over-fitting. Moreover, they may help generate entirely new theoretical
artifacts to help identify and explain social and role structure, partisanship, ideological
contestation, discursive fields, and other socio-cultural structures and institutional regimes more
dynamically.
Second, deep learning can be integrated with topic models to analyze images—alone or
along with verbal text—which opens a new path to rendering theoretical artifacts. Whereas
verbal text is descriptive, linear, additive and temporal, images and visual features are embodied,
spatial, holistic and simultaneous, which defies conventional analytical techniques. The
integration of deep learning into topic models creates potential for future theoretical development
that considers both visual features and verbal text (Krizhevsky, Sutskever, & Hinton, 2012). In
particular, scholars have argued that the role of visual features in the process of
institutionalization is significant, but largely under-examined (Meyer, Jancsary, Höllerer, &
Boxenbaum, 2017).
In other words, deep learning helps manage tradeoffs around specificity and
configuration, and represents an effective solution to the ever-present issue of theoretical
parsimony, but it also comes with a caution. Because deep learning is a computationally
inductive modeling tool, many of its operationalizations are “black boxed,” making its feature
permutations challenging to reconstruct mathematically. It ironically highlights the tradeoff of

59
human supervision and reinforces the need to apply it along with other analytical techniques
within a mixed-methods approach to generating theoretical artifacts.
Summary. All three new trends in topic modeling—eliciting latent structures, capturing
meaning, and using AI to help generate theoretical artifacts—open up new avenues for theory
building. They complement the agnostic assumptions about meaning that are embedded in the
LDA algorithm and, in this way, echo how trends related to corpora selection and trimming and
to supervising and fitting topics are helping scholars overcome some of topic modeling’s foibles
while preserving its power. In particular, by revealing latent patterns and meaning structures,
topic modeling is increasingly able to generate social, cultural, and political constructs that
define evolving cultural meanings, discursive fields, and political action.
FROM THE BALCONY
Topic modeling, a method adapted from computer science, “represents a novel tool for
analyzing large collections of qualitative data in a scalable and reproducible way” (Schmiedel et
al., 2018, p. 3; see also Kobayashi et al., 2018). Our review reveals that topic modeling has been
used in surprisingly diverse ways by management scholars, demonstrating that it is a malleable
methodological and theoretical tool for tackling a variety of research questions. Although many
papers we examined described the technical underpinnings of the LDA algorithm, we found that
topic modeling practices are part of an often-implicit process of rendering corpora, topic models,
and theoretical artifacts from raw data. We applied topic model rendering in this review to curate
and make sense of the topic modeling corpus in the management literature. Our analysis reveals
that topic modeling is gaining steam in management research (see Figure 1), particularly in five
areas: detecting novelty and emergence, developing inductive classification systems,

60
understanding online audiences and products, analyzing frames and social movements, and
understanding cultural dynamics. Topic modeling has both strengthened knowledge in each area
and enabled scholars to explore subjects in new ways. The current trends in rendering with topic
modeling have only increased the value added by the technique. We now wish to briefly consider
the topic modeling field in management research from a broader perspective, touching on
important challenges and debates that will shape the direction of research and the evolution of
the domain.
Challenges and Debates
Perhaps the biggest challenge in the near future stems from how topic modeling has
helped open the door to a plethora of work based on the quantitative structural study of meaning
(Mohr, 1998; Ventresca & Mohr, 2002). Emergent classification systems based on meaning
structures, such as those we have examined in topic modeling research, provide a reflexive
contrast to others recognized and used to parse meaning in materialized structures, such as patent
classification, risk typification, and industrial categorization. In this sense, we see management
moving in a direction that reflects current trends in cultural sociology, political science, and
linguistics; a machine learning approach like topic modeling can reveal shared cultural meanings
that in turn can be integrated into the analytical process alongside traditional socio-cultural
variables and constructs. Our identified trends in topic modeling reveal that this integration is
indeed occurring. Thus, topic modeling is not necessarily disrupting or displacing existing
methods, so much as augmenting and extending them.
By highlighting the different modes of studying meaning (Mohr et al., 2013), we also
acknowledge to the views of semiotics and qualitatively-oriented scholars who have long

61
recognized that meanings are grounded in practice and take on different levels of ambiguity. In
the debates around semiotics and modeling, it is important to recognize that topic modeling
combines the poetic (or connotative) with the semantic (or denotative) meanings of words in
topics and subjects; although the words in “bags” are independent, they are combined in
proximity and recognized in context. Integrating machine reading within studies of meaning
necessitates a discussion around the trade-offs of standardizing content and linking to theoretical
artifacts. This also highlights that topic modeling practice in management is a deeply theoretical
endeavor. Now that topic modeling algorithms are becoming more readily available through
toolkits in R, Python, and other open source software, we worry that topic modeling risks being
pigeon-holed as an LDA algorithm and “black boxed” as just another textual analysis technique.
By attending to the rendering process, we hope we have helped scholars understand the choices
inherent in the creation and pre-processing of corpora, the parameters used in the topic models
themselves, and in the creation of theoretical artifacts from the analysis. Indeed, by articulating
the rendering process, we have highlighted how topic modeling using machine learning
algorithms actually foregrounds analysts’ interpretive decisions and theory work.
Ultimately, theory is paramount for grounding claims around meaning. Our review has
emphasized that incorporating topic modeling in a theoretical manner entails careful engagement
with the cultural ecology of a social space. Our definition of the rendering process was created
along these lines; particularly when employing topic modeling to study the meanings of a social
space, one cannot neglect its structural foundations. The ecology imagery evokes connotations of
a structured space, contoured by theoretical concerns of social structure, such as boundaries,
stratification, and reputations of actors. This also invokes the imagery by philosophers of science
in assemblage theory, where a socio-cultural ecology is constituted by relationships formed

62
through processes of encoding meanings, such as stratification and territory (DeLanda, 2006).
The assemblage theory approach to conceptualizing knowledge-based fields is relevant to
our consideration of the researcher generating knowledge alongside algorithms with machine
learning. Such work is not performed by the human or the machine alone; rather, it is a combined
effort. We reflect on how assemblage theory has illustrated the institution of science operating
against the backdrop of two ideal styles of action—“nomadic” versus “state”—where the former
is paradigm breaking and smooth, concerned with variation and problematization, and the latter
is striated and contoured, concerned with precision and advances in structured fields of
knowledge (Jensen & Rödje, 2010; Deleuze & Guattari, 1987). Machine learning approaches
that are not configured with contextual structural knowledge may be nomadic—that is, overly
fluid and rendering meaning structures across fields, only looking for what is statistically
significant, but not necessarily socially or culturally significant. Understanding these ideal
“nomadic” and “state” approaches to scientific endeavors can help us understand the ideal types
of machine learning reading (nomadic: naive, fast, fluid, distant) and human-only reading (state:
careful, slow, narrowly focused, deep). Our hope is that by delineating the rendering process, we
are striking a middle ground between the two; in reflexively using machine learning tools in this
manner, the analyst can see possibilities (latent meaning structures) against materialized social
structures (formal classification systems).
To render meaning in this manner is to engender engagement with data, where the
researcher zooms in and zooms out based on distant reading (Moretti, 2013) and representations
of meaning structures. By conceptualizing topic modeling as part of a rendering process, we
hope that we have also avoided the fear that social science researchers are just “squeezing [their]
unstructured texts, sounds, or images into some special-purpose data model” (Underwood, 2015,

63
p. 1). Instead, researchers employ rendering processes for topic modeling as a “discovery
strategy” to infer meaning. This blending of formal analytical methodologies with an interpretive
focus helps reveal meanings and is echoed in an emerging stream of work in organizational
theory that Ventresca and Mohr (2002) labeled “new archivalism.”
Nevertheless, one challenge remains: as topic modeling has diffused into management
research, the practices for applying it have not remained static. Indeed, by adapting this method,
management scholars have contributed the rendering process itself. We see this contribution as
being aligned with movements that draw upon formal methods to generate representations of
meanings, which can then be analyzed in a plethora of ways (Brieger et al., 2018; Davidson et
al., 2019; Ventresca & Mohr, 2002). We found that many authors did indeed use computational
modeling tools in a manner similar what Ventresca and Mohr described in 2002; however, we
also found that the process of rendering goes further, particularly as it relates to rendering
meanings. In our opinion, topic modeling tends to naturally ally more with mixed approaches to
studying text (Brieger et al., 2018; Davidson et al., 2019; Ventresca & Mohr, 2002). Moreover,
because meaning schema (i.e., dictionaries, coding categories, etc.) are rejected a priori, the
technique often seems to be more inductive in nature.
Of course, this is by no means the only mode of theorizing enabled through topic
modeling. Other work has been more abductive in nature. For example, Fligstein et al.’s (2017)
frame analysis helps explain how the Federal Open Market Committee underestimated the risks
to the economy leading up to the 2008 financial crisis; their research design enabled them to use
topic modeling to connect hypotheses to texts via a combination of qualitative and quantitative
techniques. Indeed, topic modeling has also been used with partially deductive forms of
theorizing (e.g., see Haans, 2019; Kaplan & Vakili, 2015).

64
As a final, cumulative point, we think that the flexibility of topic modeling—its utility in
creating corpora, its ability to be paired with different quantitative and qualitative methods, and
its applicability in variety of theoretical approaches—underpins its power and promise for
management research. By surfacing topic modeling’s flexibility, we hope our detailed
exploration of the rendering process has persuaded the reader, at least to some extent, to consider
engaging with topic modeling in order to build new management theory.

65
REFERENCES
Abbott, A. (1988). The system of professions: An essay on the division of expert labor (1st ed.).
Chicago, IL: University of Chicago Press.
Abrahamson, E. (1996). Management fashion. Academy of Management Review, 21(1), 254–
285.
Abrahamson, E., & Fairchild, G. (1999). Management fashion: Lifecycles, triggers, and
collective learning processes. Administrative Science Quarterly, 44(4), 708–740.
Aggarwal, V., Lee, M. K., & Hwang, E. (2017). Status gains and subsequent effects on
evaluations. Academy of Management Proceedings, 2017(1), 16355.
Ahonen, P. (2015). Institutionalizing big data methods in social and political research. Big Data
& Society, 2(2), 1–12.
Al Sumait, L., Barbará, D., Gentle, J., & Domeniconi, C. 2009. Topic significance ranking of
LDA generative models. In Machine learning and knowledge discovery in databases, vol.
5781(pp. 67–82). Berlin, Heidelberg: Springer.
Allen, C., Luo, H., Murdock, J., Pu, J., Wang, X., Zhai, Y., & Zhao, K. (2017). Topic modeling
the Han dian ancient classics. ArXiv preprint arXiv:1702.00860. Retrieved from
https://arxiv.org/abs/1702.00860
Almquist, Z. W., & Bagozzi, B. E. (2017). Using radical environmentalist texts to uncover
network structure and network features. Sociological Methods & Research.
https://doi.org/10.1177/0049124117729696
Alvesson, M., & Kärreman, D. (2000). Taking the linguistic turn in organizational research:
Challenges, responses, consequences. Journal of Applied Behavioral Science, 36(2), 136–
158.
Alvesson, M., & Kärreman, D. (2000). Varieties of discourse: On the study of organizations
through discourse analysis. Human Relations, 53(9), 1125–1149.
Anand, N., & Peterson, R. A. (2000). When market information constitutes fields: Sensemaking
of markets in the commercial music industry. Organization Science, 11(3), 270–284.
Antons, D., Joshi, A. M., & Salge, T. O. (2018). Content, contribution, and knowledge
consumption: Uncovering hidden topic structure and rhetorical signals in scientific texts.
Journal of Management. https://doi.org/10.1177/0149206318774619
Antons, D., Kleer, R., & Salge, T. O. (2016). Mapping the topic landscape of JPIM, 1984–2013:
In search of hidden structures and development trajectories. Journal of Product
Innovation Management, 33(6), 726–749.
Argote, L., & Greve, H. R. (2007). “A behavioral theory of the firm”: 40 years and counting:
Introduction and impact. Organization Science, 18(3), 337–349.
Arora, A. (1995). Licensing tacit knowledge: Intellectual property rights and the market for
know-how. Economics of Innovation and New Technology, 4(1), 41–60.
Arora, A., Gittelman, M., Kaplan, S., Lynch, J., Mitchell, W., & Siggelkow, N. (2016).
Question-based innovations in strategy research methods. Strategic Management Journal,
37(1), 3–9.
Augustine, G., & King, B. G. (2017). Behind the scenes: A backstage look at field formation
within sustainability in higher education. Academy of Management Proceedings, 2017(1),
15788.

66
Azzopardi, L., Girolami, M., & van Risjbergen, K. (2003). Investigating the relationship
between language model perplexity and IR precision-recall measures. Presented at the
26th Annual International ACM SIGIR Conference on Research and Development in
Information Retrieval. Association for Computing Machinery. Toronto, Ontario.
Bail, C. A. (2012). The fringe effect: Civil society organizations and the evolution of media
discourse about Islam since the September 11th attacks. American Sociological Review,
77(6), 855–879.
Bail, C. A. (2014). The cultural environment: Measuring culture with big data. Theory and
Society, 43(3–4), 465–482.
Bail, C. A., Brown, T. W., & Mann, M. (2017). Channeling hearts and minds: Advocacy
organizations, Cognitive-emotional currents, and public conversation. American
Sociological Review, 82(6), 1188–1213.
Bakhtin, M. M. (1982). The dialogic imagination: Four essays. Austin, TX: University of Texas
Press.
Ballinger, G. A., & Rockmann, K. W. (2010). Chutes versus ladders: Anchoring events and a
punctuated-equilibrium perspective on social exchange relationships. Academy of
Management Review, 35(3), 373–391.
Bansal, P., & Corley, K. (2011). The coming of age for qualitative research: Embracing the
diversity of qualitative methods. Academy of Management Journal, 54(2), 233–237.
Bao, Y., & Datta, A. (2014). Simultaneously discovering and quantifying risk types from textual
risk disclosures. Management Science, 60(6), 1371–1391.
Barley, S. R. (1983). Semiotics and the study of occupational and organizational cultures.
Administrative Science Quarterly, 28(3), 393–413.
Barry, D. (1997). Telling changes: From narrative family therapy to organizational change and
development. Journal of Organizational Change Management, 10(1), 30–46.
Baumer, E. P. S., Mimno, D., Guha, S., Quan, E., & Gay, G. K. (2017). Comparing grounded
theory and topic modeling: Extreme divergence or unlikely convergence? Journal of the
Association for Information Science and Technology, 68(6), 1397–1410.
Bendle, N. T., & Wang, X. (2016). Uncovering the message from the mess of big data. Business
Horizons, 59(1), 115–124.
Benford, R. D., & Snow, D. A. (2000). Framing processes and social movements: An overview
and assessment. Annual Review of Sociology, 26, 611–639.
Bennardo, G., & de Munck, V. C. (2014). Cultural models: Genesis, methods, and experiences.
Oxford, UK: Oxford University Press.
Berelson, B. (1952). Content analysis in communication research. New York, NY: Free Press.
Blanchard, S. J., Aloise, D., & Desarbo, W. S. (2017). Extracting summary piles from sorting
task data. Journal of Marketing Research, 54(3), 398–414.
Blei, D. M. (2012). Probabilistic topic models. Communications of the ACM, 55(4), 77–84.
Blei, D. M., Griffiths, T. L., & Jordan, M. I. (2010). The nested Chinese restaurant process and
Bayesian nonparametric inference of topic hierarchies. Journal of the ACM, 57(2), 7:1–
7:30.
Blei, D. M., & Lafferty, J. D. (2007). A correlated topic model of science. Annals of Applied
Statistics, 1(1), 17–35.
Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent Dirichlet allocation. Journal of Machine
Learning Research, 3(Jan), 993–1022.

67
Boden, D., & Zimmerman, D. H. (Eds.). (1991). Talk and social structure: Studies in
ethnomethodology and conversation analysis. Berkeley, CA: University of California
Press.
Boje, D. M. (1995). Stories of the storytelling organization: A postmodern analysis of Disney as
“Tamara-Land.” Academy of Management Journal, 38(4), 997–1035.
Borgman, C. L. (2015). Big data, little data, no data: Scholarship in the networked world.
Cambridge, MA: MIT Press.
Box, G. E. P. (1979). Robustness in the strategy of scientific model building. In R. L. Launer &
G. N. Wilkinson (Eds.), Robustness in statistics (pp. 201–236). New York, NY:
Academic Press.
Brannen, M. Y. (2004). When Mickey loses face: Recontextualization, semantic fit, and the
semiotics of foreignness. Academy of Management Review, 29(4), 593–616.
Breiger, R. L., Wagner-Pacifici, R., & Mohr, J. W. (2018). Capturing distinctions while mining
text data: Toward low-tech formalization for text analysis. Poetics, 68, 104–119.
Bucheli, M., & Wadhwani, R. D. (Eds.). (2014). Organizations in time: History, theory,
methods. Oxford, UK: Oxford University Press.
Burke, K., 1945. A grammar of motives. Berkeley, CA: University of California Press.
Büschken, J., & Allenby, G. M. (2016). Sentence-based text analysis for customer reviews.
Marketing Science, 35(6), 953–975.
Buurma, R. S. (2015). The fictionality of topic modeling: Machine reading Anthony Trollope’s
Barsetshire series. Big Data & Society. https://doi.org/10.1177/2053951715610591
Campbell, J. L., Chen, H., Dhaliwal, D. S., Lu, H., & Steele, L. B. (2014). The information
content of mandatory risk factor disclosures in corporate filings. Review of Accounting
Studies, 19(1), 396–455.
Cao, L., & Fei-Fei, L. (2007). Spatially coherent latent topic model for concurrent segmentation
and classification of objects and scenes. In Computer Vision, 2007. IEEE 11th
International Conference on ICCV (pp. 1-8). Piscataway, NJ: IEEE.
Carley, K.M., Kaufer, D., 1993. Semantic connectivity: An approach for analyzing semantic
networks. Communication Theory, 3(3), 183–213.
Chang, J., & Blei, D. (2009). Relational topic models for document networks. In Artificial
intelligence and statistics (pp. 81–88). Retrieved from
http://proceedings.mlr.press/v5/chang09a.html
Chang, J., Boyd-Graber, J., Gerrish, S., Wang, C., & Blei, D. M. (2009). Reading tea leaves:
How humans interpret topic models. In Y. Bengio, D. Schuurmans, J.D. Lafferty, C.K.I.
Williams, & A. Culotta (Eds.), Advances in neutral information processing systems, vol.
22. Retrieved from https://papers.nips.cc/paper/3700-reading-tea-leaves-how-humans-
interpret-topic-models.pdf
Charmaz, K. (2014). Constructing grounded theory (2nd ed.). Thousand Oaks, CA: Sage.
Cho, Y.-J., Fu, P.-W., & Wu, C.-C. (2017). Popular research topics in marketing journals, 1995–
2014. Journal of Interactive Marketing, 40, 52–72.
Chomsky, N. (1956). Three models for the description of language. IRE Transactions on
Information Theory, 2(3), 113–124.
Corley, K. G., & Gioia, D. A. (2004). Identity ambiguity and change in the wake of a corporate
spin-off. Administrative Science Quarterly, 49(2), 173–208.

68
Cornelissen, J. P., Durand, R., Fiss, P. C., Lammers, J. C., & Vaara, E. (2015). Putting
communication front and center in institutional theory and analysis. Academy of
Management Review, 40(1), 10-27.
Croidieu, G., & Kim, P. H. (2018). Labor of love: Amateurs and lay-expertise legitimation in the
early U.S. radio field. Administrative Science Quarterly, 63(1), 1–42.
Crossley, S. A., Dascalu, M., & McNamarac, D. S. (2017). How important is size? An
investigation of corpus size and meaning in both latent semantic analysis and latent
Dirichlet allocation. In 30th International Florida Artificial Intelligence Research Society
Conference, FLAIRS 2017. Menlo Park, CA: AAAI Press.
Dalpiaz, E., Rindova, V., & Ravasi, D. (2016). Combining logics to transform organizational
agency: Blending industry and art at Alessi. Administrative Science Quarterly, 61(3),
347–392.
Davidson, E., Edwards, R., Jamieson, L., & Weller, S. (2019). Big data, qualitative style: A
breadth-and-depth method for working with large amounts of secondary qualitative data.
Quality & Quantity, 53(1), 363-376.
Davis, G. F. (2016). Can an economy survive without corporations? Technology and robust
organizational alternatives. Academy of Management Perspectives, 30(2), 129–140.
Deephouse, D. L. (1999). To be different, or to be the same? It’s a question (and theory) of
strategic balance. Strategic Management Journal, 20(2), 147–166.
Deerwester, S., Dumais, S. T., Furnas, G. W., Landauer, T. K., & Harshman, R. (1990). Indexing
by latent semantic analysis. Journal of the American Society for Information Science,
41(6), 391.
DeLanda, M. (2006). A new philosophy of society: Assemblage theory and social complexity.
London, UK: A&C Black.
Deleuze, G., & Guattari, F. (1987). A Thousand Plateaus. New York, NY: Continuum.
Denzin, N. K., & Lincoln, Y. S. (1994). Handbook of qualitative research. Thousand Oaks, CA:
Sage.
Denzin, N. K., & Lincoln, Y. S. (2011). The SAGE handbook of qualitative research. Thousand
Oaks, CA: Sage.
Dey, I. (1995). Reducing fragmentation in qualitative research. In U. Kelle (Ed.), Computer-
aided qualitative data analysis: Theory, methods, and practice (pp. 69–79). London:
Sage.
Diesner, J., Carley, K.M. (2005). Revealing social structure from texts: Meta-matrix text analysis
as a novel method for network text analysis. In V. K. Narayanan & D. J. Armstrong
(Eds.), Causal mapping for research in information technology (pp. 81–108). Hershey,
PA: IGI Global.
Dilthey, W., & Jameson, F. (1972). The rise of hermeneutics. New Literary History, 3(2), 229–
244.
DiMaggio, P. J. (1997). Culture and cognition. Annual Review of Sociology, 23, 263–287.
DiMaggio, P. J. (2015). Adapting computational text analysis to social science (and vice versa).
Big Data & Society, 2(2), 2053951715602908.
DiMaggio, P. J., Nag, M., & Blei, D. (2013). Exploiting affinities between topic modeling and
the sociological perspective on culture: Application to newspaper coverage of U.S.
government arts funding. Poetics, 41(6), 570–606.

69
DiMaggio, P., Nag, M., & Blei, D. (2013). Exploiting affinities between topic modeling and the
sociological perspective on culture: Application to newspaper coverage of U.S.
government arts funding. Poetics, 41(6), 570–606.
Durand, R., & Khaire, M. (2017). Where do market categories come from and how?
Distinguishing category creation from category emergence. Journal of Management,
43(1), 87–110.
Durand, R., & Paolella, L. (2013). Category stretching: Reorienting research on categories in
strategy, entrepreneurship, and organization theory. Journal of Management
Studies, 50(6), 1100–1123.
Duriau, V. J., Reger, R. K., & Pfarrer, M. D. (2007). A content analysis of the content analysis
literature in organization studies: Research themes, data sources, and methodological
refinements. Organizational Research Methods, 10(1), 5–34.
Dyer, T., Lang, M., & Stice-Lawrence, L. (2017). The evolution of 10-K textual disclosure:
Evidence from latent Dirichlet allocation. Journal of Accounting and Economics, 64(2),
221–245. https://doi.org/10.1016/j.jacceco.2017.07.002
Edmondson, A. C., & Mcmanus, S. E. (2007). Methodological fit in management field research.
Academy of Management Review, 32(4), 1246–1264.
Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management
Review, 14(4), 532–550.
Etter, M., Ravasi, D., & Colleoni, E. (2018). Social media and the formation of organizational
reputation. Academy of Management Review.
Evans, J. A., & Aceves, P. (2016). Machine translation: Mining text for social theory. Annual
Review of Sociology, 42(1), 21–50.
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds.
Journal of Financial Economics, 33(1), 3–56.
Firth, J. R. (1957). Papers in linguistics. London, UK: Oxford University Press.
Fiss, P. C. (2007). A set-theoretic approach to organizational configurations. Academy of
Management Review, 32(4), 1180–1198.
Fiss, P. C., & Hirsch, P. M. (2005). The discourse of globalization: Framing and sensemaking of
an emerging concept. American Sociological Review, 70(1), 29–52.
Fleming, L. (2001). Recombinant uncertainty in technological search. Management Science,
47(1), 117–132.
Fligstein, N., & McAdam, D. (2011). Toward a general theory of strategic action fields.
Sociological Theory, 29(1), 1–26.
Fligstein, N., Stuart Brundage, J., & Schultz, M. (2017). Seeing like the Fed: Culture, cognition,
and framing in the failure to anticipate the financial crisis of 2008. American Sociological
Review, 82(5), 879–909.
Forbes, D. P., & Kirsch, D. A. (2011). The study of emerging industries: Recognizing and
responding to some central problems. Journal of Business Venturing, 26, 589–602.
Gehman, J., Glaser, V. L., Eisenhardt, K. M., Gioia, D., Langley, A., & Corley, K. G. (2018).
Finding theory-method fit: A comparison of three qualitative approaches to theory
building. Journal of Management Inquiry, 27(3), 284–300.
Gehman, J., & Soublière, J.-F. (2017). Cultural entrepreneurship: From making culture to
cultural making. Innovation, 19(1), 61–73.
Gerlach, M., Peixoto, T. P., & Altmann, E. G. (2018). A network approach to topic models.
Science Advances, 4(7), eaaq1360.

70
Gioia, D. A., & Chittipeddi, K. (1991). Sensemaking and sensegiving in strategic change
initiation. Strategic Management Journal, 12(6), 433–448.
Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking qualitative rigor in inductive
research: Notes on the Gioia methodology. Organizational Research Methods, 16(1), 15–
31.
Gioia, D. A., & Thomas, J. B. (1996). Identity, image, and issue interpretation: Sensemaking
during strategic change in academia. Administrative Science Quarterly, 41(3), 370–403.
Giorgi, S., & Weber, K. (2015). Marks of distinction: Framing and audience appreciation in the
context of investment advice. Administrative Science Quarterly, 60(2), 333–367.
Glaser, B. G., & Strauss, A. L. (1967). Discovery of grounded theory: Strategies for qualitative
research. New York, NY: Routledge.
Glaser, V., Fiss, P. C., & Kennedy, M. T. (2011). Rhetoric and resonance: Framing strategies for
institutionalizing new market conceptions. Academy of Management Proceedings,
2011(1), 1–6.
Goffman, E. (1974). Frame analysis: An essay on the organization of experience. Cambridge,
MA: Harvard University Press.
Greve, H. R. (2003). Organizational learning from performance feedback: A behavioral
perspective on innovation and change. Cambridge, UK: Cambridge University Press.
Grimmer, J., & Stewart, B. M. (2013). Text as data: The promise and pitfalls of automatic
content analysis methods for political texts. Political Analysis, 21(3), 267–297.
Guerreiro, J., Rita, P., & Trigueiros, D. (2016). A text mining-based review of cause-related
marketing literature. Journal of Business Ethics, 139(1), 111–128.
Guo, L., Sharma, R., Yin, L., Lu, R., & Rong, K. (2017). Automated competitor analysis using
big data analytics: Evidence from the fitness mobile app business. Business Process
Management Journal, 23(3), 735–762.
Haans, R. F. J. (2019). What’s the value of being different when everyone is? The effects of
distinctiveness on performance in homogeneous versus heterogeneous categories.
Strategic Management Journal, 40(1), 3–27.
Hankammer, S., Antons, D., Kleer, R., & Piller, F. (2016). Researching mass customization:
Mapping hidden structures and development trajectories. In Academy of Management
Proceedings, 2016(1), 10900.
Hannan, M. T., & Carroll, G. R. (1992). Dynamics of organizational populations: Density,
legitimation, and competition. Oxford, UK: Oxford University Press.
Hannan, M. T., & Freeman, J. (1977). The population ecology of organizations. American
Journal of Sociology, 82(5), 929–964.
Hannan, M. T., Pólos, L., & Carroll, G. R. (2007). Logics of organization theory: Audiences,
codes, and ecologies. Princeton, NJ: Princeton University Press.
Hannigan, T. R. & Casasnovas, G. (2019). New structuralism and field emergence: The co-
constitution of meanings and actors in the early moments of impact investing. Research
in the Sociology of Organizations.
Hannigan, T. R., Bundy, J., Graffin, S. D., Wade, J. B., & Porac, J. (2015). The social
construction of scandal: The role of media in the British parliamentary expense affair.
Academy of Management Proceedings, 2015(1), 14966.
Hannigan, T. R., Porac, J. F., Bundy, J., Wade, J. B., & Graffin, S. D. (2019). Crossing the line
or creating the line: Media effects in the 2009 British MP expense scandal. Working
paper.

71
Hardy, C. (2001). Researching organizational discourse. International Studies of Management &
Organization, 31(3), 25–47.
Hatch, M. J. (1993). The dynamics of organizational culture. Academy of Management Review,
18(4), 657–693.
Hatch, M. J., & Schultz, M. (2017). Toward a theory of using history authentically: Historicizing
in the Carlsberg Group. Administrative Science Quarterly, 62(4), 657–697.
Heugens, P. P., & Lander, M. W. (2009). Structure! Agency! (and other quarrels): A meta-
analysis of institutional theories of organization. Academy of Management Journal,
52(1), 61-85.
Houghton, J. P., Siegel, M., Madnick, S., Tounaka, N., Nakamura, K., Sugiyama, T., Nakagawa,
D., & Shirnen, B. (2017). Beyond keywords: Tracking the evolution of conversational
clusters in social media. Sociological Methods & Research.
https://doi.org/10.1177/0049124117729705
Hu, D. J., & Saul, L. K. (2009). A probabilistic topic model for music analysis. In Proceedings of
NIPS, vol. 9 (pp. 1–4). Retrieved from
http://cseweb.ucsd.edu/~dhu/docs/nips09_abstract.pdf
Huang, A. H., Lehavy, R., Zang, A. Y., & Zheng, R. (2017). Analyst information discovery and
interpretation roles: A topic modeling approach. Management Science, 64(6), 2833–2855.
Huang, X., Li, X., Zhang, L., Liu, T., Chiu, D., & Zhu, T. (2015). Topic model for identifying
suicidal ideation in Chinese microblog. In 29th Pacific Asia Conference on Language,
Information and Computation (pp. 553–562).
Humphreys, A., & Wang, R. J.-H. (2018). Automated text analysis for consumer research.
Journal of Consumer Research, 44(6), 1274–1306.
Ignatow, G., & Robinson, L. (2017). Pierre Bourdieu: Theorizing the digital. Information,
Communication & Society, 20(7), 950–966.
Jacobs, B. J. D., Donkers, B., & Fok, D. (2016). Model-based purchase predictions for large
assortments. Marketing Science, 35(3), 389–404.
Jensen, C. B. & Rödje, K. (2010). Deleuzian Intersections: Science, Technology, Anthropology.
New York, NY: Berghahn Books
Jha, H. K., & Beckman, C. M. (2017). A patchwork of identities: Emergence of charter schools
as a new organizational form. In Emergence, vol. 50 (pp. 69–107). Bingley, UK: Emerald
Publishing.
Jin, M., Luo, X., Zhu, H., & Zhuo, H. H. (2018). Combining deep learning and topic modeling
for review understanding in context-aware recommendation. In Proceedings of NAACL-
HLT 2018 (pp. 1605–1614). doi: 10.18653/v1/N18-1145
Jockers, M. L., & Mimno, D. (2013). Significant themes in 19th-century literature. Poetics,
41(6), 750–769.
Jurafsky, D., & Martin, J. H. (2008). Speech and language processing (2nd ed.). Upper Saddle
River, N.J: Pearson.
Kaminski, J., Jiang, Y., Piller, F., & Hopp, C. (2017). Do user entrepreneurs speak different?
Applying natural language processing to crowdfunding videos. In Proceedings of the
2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems (pp.
2683–2689). New York, NY: ACM.
Kaplan, S. (2008a). Cognition, capabilities, and incentives: Assessing firm response to the fiber-
optic revolution. Academy of Management Journal, 51(4), 672–695.

72
Kaplan, S. (2008b). Framing contests: Strategy making under uncertainty. Organization Science,
19(5), 729–752.
Kaplan, S., & Vakili, K. (2015). The double-edged word of recombination in breakthrough
innovation. Strategic Management Journal, 36(10), 1435–1457.
Karanovic, J., Berends, H., & Engel, Y. (2018). Is platform capitalism legit? Ask the workers.
Academy of Management Proceedings, 2018(1), 17038.
Kennedy, M. T. (2005). Behind the one-way mirror: Refraction in the construction of product
market categories. Poetics, 33(3–4), 201–226.
Kennedy, M. T. (2008). Getting counted: Markets, media, and reality. American Sociological
Review, 73(2), 270–295.
Kennedy, M. T., Chok, J. I., & Liu, J. (2012). What does it mean to be green? The emergence of
new criteria for assessing corporate reputation. In T. G. Pollock & M. L. Barnett (Eds.),
Oxford Handbook of Corporate Reputation. Oxford, UK: Oxford University Press. doi:
10.1093/oxfordhb/9780199596706.013.0004
Kennedy, M. T., & Fiss, P. C. (2013). An ontological turn in categories research: From standards
of legitimacy to evidence of actuality. Journal of Management Studies, 50(6), 1138–
1154.
Kim, H., Ahn, S.-J., & Jung, W.-S. (2018). Horizon scanning in policy research database with a
probabilistic topic model. Technological Forecasting and Social Change.
https://doi.org/10.1016/j.techfore.2018.02.007
Kim, S., & Bae, J. 2016. Cross-cultural differences in concrete and abstract corporate social
responsibility (CSR) campaigns: Perceived message clarity and perceived CSR as
mediators. International Journal of Corporate Social Responsibility, 1, 1–14.
Kinney, A. B., Davis, A. P., & Zhang, Y. (2018). Theming for terror: Organizational adornment
in terrorist propaganda. Poetics, 69, 27–40.
Kiss, T., & Strunk, J. (2006). Unsupervised multilingual sentence boundary detection.
Computational Linguistics, 32(4), 485–525.
Kitchin, R., & McArdle, G. (2016). What makes big data, big data? Exploring the ontological
characteristics of 26 datasets. Big Data & Society, 3(1), 1–10.
Kline, S. J., & Rosenberg, N. (1986). An overview of innovation. In R. Landau & N. Rosenberg
(Eds.), The positive sum strategy: Harnessing technology for economic growth (pp. 275–
306). Washington, DC: National Academy of Sciences.
Kluyver, T., Ragan-Kelley, B., Pérez, F., Granger, B. E., Bussonnier, M., Frederic, J., Corlay, S.
(2016). Jupyter Notebooks—a publishing format for reproducible computational
workflows. In F. Loizides & B. Schmidt (Eds.), Positioning and power in academic
publishing: Players, agents and agendas (pp. 87–90). Clifton, VA: IOS Press. doi:
10.3233/978-1-61499-649-1-87
Kobayashi, V. B., Mol, S. T., Berkers, H. A., Kismihók, G., & Den Hartog, D. N. (2018). Text
classification for organizational researchers: A tutorial. Organizational Research
Methods, 21(3), 766–799.
Krippendorff, K. (1980). Content analysis (1st ed.). Beverly Hills, CA: Sage.
Krippendorff, K. (2004). Content analysis: An introduction to its methodology (2nd ed.).
Thousand Oaks, CA: Sage.
Krippendorff, K. (2012). Content analysis: An introduction to its methodology (3rd ed.).
Thousand Oaks, CA: Sage.

73
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep
convolutional neural networks. In Proceedings of the 25th International Conference on
Neural Information Processing Systems, vol. 1 (pp. 1097–1105). Red Hook, NY: Curran
Associates Inc.
Kuhn, T. S. (1996). The structure of scientific revolutions (3rd ed.). Chicago, IL: University of
Chicago Press.
Lakoff, G. (1970). Linguistics and natural logic. Synthese, 22(1–2), 151–271.
Langley, A. (1999). Strategies for theorizing from process data. Academy of Management
Review, 24(4), 691–710.
Langley, A., & Abdallah, C. (2011). Templates and turns in qualitative studies of strategy and
management. In Building methodological bridges (pp. 201–235). Bingley, UK: Emerald
Group Publishing.
Lasswell, H. D. (1948). Power and personality. New York, NY: W.W. Norton.
Lasswell, H. D., & Lerner, D. (1952). The comparative study of symbols. Palo Alto, CA:
Stanford University Press.
Lawrence, T. B., Suddaby, R., & Leca, B. (2009). Institutional work (1st ed.). Cambridge, UK:
Cambridge University Press.
Lee, G. M., Qiu, L., & Whinston, A. B. (2016). A friend like me: Modeling network formation in
a location-based social network. Journal of Management Information Systems, 33(4),
1008–1033.
Lee, H., Kwak, J., Song, M., & Kim, C. O. (2015). Coherence analysis of research and education
using topic modeling. Scientometrics, 102(2), 1119–1137.
Lee, T. Y., & Bradlow, E. T. (2011). Automated marketing research using online customer
reviews. Journal of Marketing Research, 48(5), 881–894.
Levy, K. E. C., & Franklin, M. (2014). Driving regulation: Using topic models to examine
political contention in the U.S. trucking industry. Social Science Computer Review, 32(2),
182–194.
Lim, A., & Tsutsui, K. 2012. Globalization and commitment in corporate social responsibility:
Cross-national analyses of institutional and political economy effects. American
Sociological Review, 77(1): 69-98.
Liu, Y., Mai, F., & MacDonald, C. (2018). A big-data approach to understanding the thematic
landscape of the field of business ethics, 1982–2016. Journal of Business Ethics.
https://doi.org/10.1007/s10551-018-3806-5
Locke, K. D. (2001). Grounded theory in management research. London: Sage.
Loewenstein, J., Ocasio, W., & Jones, C. (2012). Vocabularies and vocabulary structure: A new
approach linking categories, practices, and institutions. Academy of Management Annals,
6(1), 41–86.
Lounsbury, M., & Glynn, M. A. (2001). Cultural entrepreneurship: Stories, legitimacy, and the
acquisition of resources. Strategic Management Journal, 22(6/7), 545–564.
Lounsbury, M., & Glynn, M. A. (2019). Cultural entrepreneurship: A new agenda for the study
of entrepreneurial processes and possibilities. Cambridge, UK: Cambridge University
Press.
Lounsbury, M., & Ventresca, M. (2003). The new structuralism in organizational theory.
Organization, 10(3), 457–480.
McCallum, A. K. (2002). MALLET: A machine learning for language toolkit. Retrieved from
http://mallet.cs.umass.edu.

74
Manning, C., Raghavan, P., & Schütze, H. (2010). Introduction to information retrieval. Natural
Language Engineering, 16, 100–103.
Manning, C. D., & Schütze, H. (1999). Foundations of statistical natural language processing
(1st ed.). Cambridge, MA: MIT Press.
Manning, C. D., Surdeanu, M., Bauer, J., Finkel, J., Bethard, S. J., & McClosky, D. (2014). The
Stanford CoreNLP Natural Language Processing Toolkit. In Proceedings of the 52nd
Annual Meeting of the Association for Computational Linguistics: System
Demonstrations (pp. 55-60). https://www.aclweb.org/anthology/P14-5010
Mantere, S., & Ketokivi, M. (2013). Reasoning in organization science. Academy of
Management Review, 38(1), 70–89.
Marciniak, D. (2016). Computational text analysis: Thoughts on the contingencies of an evolving
method. Big Data & Society, 3(2).
Marquis, C., Glynn, M. A., & Davis, G. F. (2007). Community isomorphism and corporate social
action. Academy of Management Review, 32(3), 925–945.
Marshall, E. A. (2013). Defining population problems: Using topic models for cross-national
comparison of disciplinary development. Poetics, 41(6), 701–724.
Martens, M. L., Jennings, J. E., & Jennings, P. D. (2007). Do the stories they tell get them the
money they need? The role of entrepreneurial narratives in resource acquisition. Academy
of Management Journal, 50(5), 1107–1132.
Mattmann, C. A. (2013). Computing: A vision for data science. Nature, 493(7433), 473.
McFarland, D. A., Ramage, D., Chuang, J., Heer, J., Manning, C. D., & Jurafsky, D. (2013).
Differentiating language usage through topic models. Poetics, 41(6), 607–625.
Mehrotra, R., Sanner, S., Buntine, W., & Xie, L. (2013). Improving LDA topic models for
microblogs via tweet pooling and automatic labeling. In Proceedings of the 36th
International ACM SIGIR Conference on Research and Development in Information
Retrieval (pp. 889–892). New York, NY: ACM.
Meyer, J., & Hannan, M. T. (1979). National development and the world system: Educational,
economic, and political change, 1950-1970. Chicago, IL: University of Chicago Press.
Meyer, R. E., Jancsary, D., Höllerer, M. A., & Boxenbaum, E. (2017). The role of verbal and
visual text in the process of institutionalization. Academy of Management Review, 43(3),
392–418.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook.
Thousand Oaks, CA: Sage.
Miles, M. B., Huberman, A. M., & Saldaña, J. (2013). Qualitative data analysis (3rd ed.).
Thousand Oaks, CA: Sage.
Miller, G. A., Beckwith, R., Fellbaum, C., Gross, D., & Miller, K. J. (1990). Introduction to
WordNet: An on-line lexical database. International Journal of Lexicography, 3(4), 235–
234.
Miller, I. M. (2013). Rebellion, crime and violence in Qing China, 1722–1911: A topic modeling
approach. Poetics, 41(6), 626–649.
Mimno, D. (2012). Computational historiography: Data mining in a century of classics journals.
Journal on Computing and Cultural Heritage, 5(1), 3:1–3:19.
Mimno, D., Wallach, H. M., Talley, E., Leenders, M., & McCallum, A. (2011). Optimizing
semantic coherence in topic models. In Proceedings of the Conference on Empirical
Methods in Natural Language Processing (pp. 262–272). Stroudsburg, PA: Association
for Computational Linguistics.

75
Moe, W. W., Netzer, O., & Schweidel, D. A. (2017). Social media analytics. In B. Wierenga &
R. van der Lans (Eds.), Handbook of Marketing Decision Models (pp. 483–504). New
York, NY: Springer International.
Moe, W. W., & Schweidel, D. A. (2017). Opportunities for innovation in social media analytics.
Journal of Product Innovation Management, 34(5), 697–702.
Mohr, J. W. (1998). Measuring meaning structures. Annual Review of Sociology, 24, 345–370.
Mohr, J. W., & Bogdanov, P. (2013). Introduction—topic models: What they are and why they
matter. Poetics, 41(6), 545–569.
Mohr, J. W., & Duquenne, V. (1997). The duality of culture and practice: Poverty relief in
New York City, 1888–1917. Theory and Society, 26(2/3), 305–356.
Mohr, J. W., Wagner-Pacifici, R., Breiger, R. L., & Bogdanov, P. (2013). Graphing the grammar
of motives in national security strategies: Cultural interpretation, automated text analysis
and the drama of global politics. Poetics, 41(6), 670–700.
Mollick, E. (2014). The dynamics of crowdfunding: An exploratory study. Journal of Business
Venturing, 29(1), 1–16.
Momeni, A., & Rost, K. (2016). Identification and monitoring of possible disruptive
technologies by patent-development paths and topic modeling. Technological
Forecasting and Social Change, 104, 16–29.
Moretti, F. (2013). Distant reading. New York, NY: Verso.
Morgeson, F. P., Mitchell, T. R., & Liu, D. (2015). Event system theory: An event-oriented
approach to the organizational sciences. Academy of Management Review, 40(4), 515–
537.
Mützel, S. (2015). Facing big data: Making sociology relevant. Big Data & Society, 2(2),
2053951715599179.
Nag, M. (2015). Meaning is relational: The changing contexts of the keyword “risk” in the New
York Times using a bag-of-tuples topic model. Presented at the Text II Conference,
Princeton, NJ.
Nag, R., & Gioia, D. A. (2012). From common to uncommon knowledge: Foundations of firm-
specific use of knowledge as a resource. Academy of Management Journal, 55(2), 421–
457.
Nam, H., Joshi, Y. V., & Kannan, P. K. (2017). Harvesting brand information from social tags.
Journal of Marketing, 81(4), 88–108.
Navis, C., & Glynn, M. A. (2010). How new market categories emerge: Temporal dynamics of
legitimacy, identity, and entrepreneurship in satellite radio, 1990–2005. Administrative
Science Quarterly, 55(3), 439–471.
Navis, C., & Glynn, M. A. (2011). Legitimate distinctiveness and the entrepreneurial identity:
Influence on investor judgments of new venture plausibility. Academy of Management
Review, 36(3), 477–499.
Nelsen, B. J., & Barley, S. R. (1997). For love or money? Commodification and the construction
of an occupational mandate. Administrative Science Quarterly, 42(4), 619–653.
Nelson, L. K. (2017). Computational grounded theory: A methodological framework.
Sociological Methods & Research, 0049124117729703.
Netzer, O., Feldman, R., Goldenberg, J., & Fresko, M. (2012). Mine your own business: Market-
structure surveillance through text mining. Marketing Science, 31(3), 521–543.

76
Newman, D., Noh, Y., Talley, E., Karimi, S., & Baldwin, T. (2010). Evaluating topic models for
digital libraries. In Proceedings of the 10th Annual Joint Conference on Digital Libraries
(pp. 215–224). New York, NY: ACM.
Ocasio, W. (1997). Towards an attention-based view of the firm. Strategic Management Journal,
18, 187–206.
Oh, J., Stewart, A. E., & Phelps, R. E. (2017). Topics in the Journal of Counseling Psychology,
1963–2015. Journal of Counseling Psychology, 64(6), 604–615.
Peirce, C. S. (1958). Collected papers of Charles Sanders Peirce. (A. W. Burks, Ed.).
Cambridge, MA: Harvard University Press.
Pennebaker, J.W., Boyd, R.L., Jordan, K., & Blackburn, K. (2015). The development and
psychometric properties of LIWC2015. Austin, TX: University of Texas at Austin.
Pentland, B. T., & Feldman, M. S. (2005). Organizational routines as a unit of analysis.
Industrial and Corporate Change, 14(5), 793–815.
Pfarrer, M., Pollock, T., & Rindova, V. (2010). A tale of two assets: The effects of firm
reputation and celebrity on earnings surprises and investors’ reactions. Academy of
Management Journal, 53(5), 1131–1152.
Phillips, N., Lawrence, T. B., & Hardy, C. (2004). Discourse and institutions. Academy of
Management Review, 29(4), 635–652.
Podolny, J. M. (1993). A status-based model of market competition. American Journal of
Sociology, 98(4), 829–872.
Pollach, I. (2012). Taming textual data: The contribution of corpus linguistics to computer-aided
text analysis. Organizational Research Methods, 15(2), 263–287.
Porac, J. F., Wade, J. B., & Pollock, T. G. (1999). Industry categories and the politics of the
comparable firm in CEO compensation. Administrative Science Quarterly, 44(1), 112–
144.
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, 52(5), 856–862.
Prein, G., & Kelle, U. (1995). Using linkages and networks for theory building. In U. Kelle
(Ed.), Computer qualitative data analysis: Theory, methods, and practice (pp. 62–68).
London, UK: Sage Publications.
Pröllochs, N., & Feuerriegel, S. (2018). Business analytics for strategic management: Identifying
and assessing corporate challenges via topic modeling. Information & Management.
https://doi.org/10.1016/j.im.2018.05.003
Puranam, D., Narayan, V., & Kadiyali, V. (2017). The effect of calorie posting regulation on
consumer opinion: A flexible latent Dirichlet allocation model with informative priors.
Marketing Science, 36(5), 726–746.
Raffaelli, R. (2018). Technology reemergence: Creating new value for old technologies in Swiss
mechanical watchmaking, 1970–2008. Administrative Science Quarterly,
0001839218778505. https://doi.org/10.1177/0001839218778505
Ragin, C. C. (2008). Redesigning social inquiry: Fuzzy sets and beyond. Chicago, IL: University
of Chicago Press.
Rao, H., Monin, P., & Durand, R. (2003). Institutional change in Toque Ville: Nouvelle cuisine
as an identity movement in French gastronomy. American Journal of Sociology, 108(4),
795–843.

77
Rhee, E. Y., & Fiss, P. C. (2014). Framing controversial actions: Regulatory focus, source
credibility, and stock market reaction to poison pill adoption. Academy of Management
Journal, 57(6), 1734-1758.
Roberts, M. E., Stewart, B. M., Tingley, D., Lucas, C., Leder‐Luis, J., Gadarian, S. K.,
Albertson, B., & Rand, D. G. (2014). Structural topic models for open-ended survey
responses. American Journal of Political Science, 58(4), 1064–1082.
Röder, M., Both, A., & Hinneburg, A. (2015). Exploring the space of topic coherence measures.
In Proceedings of the Eighth ACM International Conference on Web search and Data
Mining (pp. 399–408). New York, NY: ACM.
Rosen-Zvi, M., Griffiths, T., Steyvers, M., & Smyth, P. (2004). The author-topic model for
authors and documents. In Proceedings of the 20th Conference on Uncertainty in
Artificial Intelligence (pp. 487–494). Arlington, VA: AUAI Press.
Rothenberg, A. (2014) Flight from wonder—An investigation of scientific creativity. New York,
NY: Oxford University Press.
Ruckman, K., & McCarthy, I. (2017). Why do some patents get licensed while others do not?
Industrial and Corporate Change, 26(4), 667–688.
Saussure, F. de (1959). Course in general linguistics. New York, NY: McGraw-Hill.
Schmiedel, T., Müller, O., & vom Brocke, J. (2018). Topic modeling as a strategy of inquiry in
organizational research. Organizational Research Methods, 3(1).
https://doi.org/10.1177/1094428118773858
Shi, Z., Lee, G. M., & Whinston, A. B. (2016). Toward a better measure of business proximity:
Topic modeling for industry intelligence. MIS Quarterly, 40(4), 1035–1056.
Shimizu, T. (2018). Coordinating collective framing process with heterogeneous actors in
technology standard development. Presented at the 78th Annual Meeting of the Academy
of Management, Chicago, IL.
Sievert, C., & Shirley, K. (2014). LDAvis: A method for visualizing and interpreting topics. In
Proceedings of the Workshop on Interactive Language Learning, Visualization, and
Interfaces (pp. 63–70). Stroudsburg, PA: Association for Computational Linguistics.
Slingerland, E., Nichols, R., Neilbo, K., & Logan, C. (2017). The distant reading of religious
texts: A “big data” approach to mind-body concepts in early China. Journal of the
American Academy of Religion, 85(4), 985–1016.
Smith, A., Hawes, T., & Myers, M. (2014). Hiéarchie: Visualization for hierarchical topic
models. In Proceedings of the Workshop on Interactive Language Learning,
Visualization, and Interfaces (pp. 71–78). Baltimore, MD: Association for Computational
Linguistics.
Snow, D. A., & Benford, R. D. (1988). Ideology, frame resonance, and participant mobilization.
In International Social Movement Research, vol. 1 (pp. 197–218). Greenwich, CT: JAI
Press.
Snow, D. A., Rochford, E. B., Worden, S. K., & Benford, R. D. (1986). Frame alignment
processes, micromobilization, and movement participation. American Sociological
Review, 51(4), 464–481.
Socher, R., Perelygin, A., Wu, J., Chuang, J., Manning, C. D., Ng, A., & Potts, C. (2013).
Recursive deep models for semantic compositionality over a sentiment treebank. In
Proceedings of the 2013 Conference on Empirical Methods in Natural Language
Processing (pp. 1631–1642). Retrived from
https://nlp.stanford.edu/~socherr/EMNLP2013_RNTN.pdf

78
Song, M., Heo, G. E., & Lee, D. (2015). Identifying the landscape of Alzheimer’s disease
research with network and content analysis. Scientometrics, 102(1), 905–927.
Song, M., & Kim, S. Y. (2013). Detecting the knowledge structure of bioinformatics by mining
full-text collections. Scientometrics, 96(1), 183–201.
Sørensen, J. B., & Stuart, T. E. (2000). Aging, obsolescence, and organizational innovation.
Administrative Science Quarterly, 45(1), 81–112.
Stevens, K., Kegelmeyer, P., Andrzejewski, D., & Buttler, D. (2012). Exploring Topic
Coherence over Many Models and Many Topics. Proceedings of the 2012 Joint Conference
on Empirical Methods in Natural Language Processing and Computational Natural
Language Learning, 952–961.
Strauss, A. C., & Corbin, J. M. (1998). Basics of qualitative research: Techniques and
procedures for developing grounded theory (2nd ed.). Los Angeles, CA: Sage.
Strothotte, T., & Schlechtweg, S. (2002). Non-photorealistic computer graphics: Modeling,
rendering, and animation. San Francisco, CA: Morgan Kaufmann.
Suominen, A., Toivanen, H., & Seppänen, M. (2017). Firms’ knowledge profiles: Mapping
patent data with unsupervised learning. Technological Forecasting and Social Change,
115, 131–142.
Tangherlini, T. R., & Leonard, P. (2013). Trawling in the sea of the great unread: Sub-corpus
topic modeling and humanities research. Poetics, 41(6), 725–749.
Tchalian, H. (2019). Microfoundations and recursive analysis: A mixed-methods framework for
language-based research, computational methods, and theory development. In
Microfoundations of Institutions.
Tchalian, H., Alsudais, A., & Ocasio, W. (2018). Bringing values back in: Cultural persistence
in institutional vocabularies. Working paper.
Tchalian, H., Glaser, V. L., Hannigan, T. R., & Lounsbury, M. (2019). Institutional attention:
Cultural entrepreneurship and the dynamics of category construction. Working paper.
Thornton, P. H., & Ocasio, W. (1999). Institutional logics and the historical contingency of
power in organizations: Executive succession in the higher education publishing industry,
1958-1990. American Journal of Sociology, 105(3), 801–843.
Thornton, P. H., Ocasio, W., & Lounsbury, M. (2012). The institutional logics perspective: A
new approach to culture, structure and process. New York, NY: Oxford University
Press.
Timmermans, S., & Tavory, I. (2012). Theory construction in qualitative research: From
grounded theory to abductive analysis. Sociological Theory, 30(3), 167–186.
Tirunillai, S., & Tellis, G. J. (2014). Mining marketing meaning from online chatter: Strategic
brand analysis of big data using latent Dirichlet allocation. Journal of Marketing
Research, 51(4), 463–479.
Tolbert, P. S., & Zucker, L. G. (1996). The institutionalization of institutional theory. In S.
Clegg, C. Hardy, & W. Nord (Eds.), Handbook of organization studies (pp. 175–190).
London: Sage.
Tonidandel, S., King, E. B., & Cortina, J. M. (2018). Big data methods: Leveraging modern data
analytic techniques to build organizational science. Organizational Research Methods,
21(3), 525-547.
Toubia, O., & Netzer, O. (2016). Idea generation, creativity, and prototypicality. Marketing
Science, 36(1), 1–20.

- 79 -
Trajtenberg, M. (1990). A penny for your quotes: Patent citations and the value of innovations.
RAND Journal of Economics, 21(1), 172–187.
Trusov, M., Ma, L., & Jamal, Z. (2016). Crumbs of the cookie: User profiling in customer-base
analysis and behavioral targeting. Marketing Science, 35(3), 405–426.
Turney, P. D., & Pantel, P. (2010). From frequency to meaning: Vector space models of
semantics. Journal of Artificial Intelligence Research, 37, 141-188.
Tzoukermann, E., Klavans, J. L., & Strzalkowski, T. (2005). Information retrieval. In R. Mitkov
(Ed.), Oxford Handbook of Computational Linguistics (pp. 1–12). Oxford, UK: Oxford
University Press.
Underwood, T. (2015). The literary uses of high-dimensional space. Big Data & Society, 2(2), 1–
6.
Uzzi, B., Mukherjee, S., Stringer, M, & Jones, B. (2013) Atypical combinations and scientific
impact. Science, 342(6157):468–472.
Vaara, E. (2010). Taking the linguistic turn seriously: Strategy as a multifaceted and
interdiscursive phenomenon. In J. Baum & J. B. Lampel (Eds.), The globalization of
strategy research (pp. 29–50). Bingley, UK: Emerald Group Publishing.
Vaara, E., Aranda, A., Etchanchu, H., Guyt, J., and Sele, K. (2019). How to make use of
structural topic modeling in critical discourse analysis? Working paper.
Ventresca, M. J., & Mohr, J. W. (2002). Archival research methods. In J. A. C. Baum (Ed.),
Blackwell companion to organizations (pp. 805–828). Oxford, UK: Blackwell.
Venugopalan, S., & Rai, V. (2015). Topic based classification and pattern identification in
patents. Technological Forecasting and Social Change, 94, 236–250.
Vergne, J.-P., & Wry, T. (2014). Categorizing categorization research: Review, integration, and
future directions. Journal of Management Studies, 51(1), 56-94.
Wagner-Pacifici, R., Mohr, J. W., & Breiger, R. L. (2015). Ontologies, methodologies, and new
uses of big data in the social and cultural sciences. Big Data & Society, 2(2),
2053951715613810.
Wang, X., Bendle, N. T., Mai, F., & Cotte, J. (2015). The Journal of Consumer Research at 40:
A historical analysis. Journal of Consumer Research, 42(1), 5–18.
Wang, X., McCallum, A., & Wei, X. (2007). Topical n-grams: Phrase and topic discovery, with
an application to information retrieval. In Proceedings of the Seventh IEEE International
Conference on Data Mining (ICDM 2007) (pp. 697–702). Piscataway, NJ: IEEE. doi:
10.1109/ICDM.2007.86
Wang, Y., & Chaudhry, A. (2018). When and how managers’ responses to online reviews affect
subsequent reviews. Journal of Marketing Research, 55(2), 163–177.
Washington, M., & Zajac, E. J. (2005). Status evolution and competition: Theory and evidence.
Academy of Management Journal, 48(2), 282–296.
Weber, K., Patel, H., & Heinze, K. L. (2013). From cultural repertoires to institutional logics: A
content-analytic method. In M. Lounsbury & E. Boxenbaum (Eds.), Institutional Logics
in Action, Vol. 39B (pp. 351–382). Bingley, UK: Emerald Group Publishing.
Weber, K., & Dacin, M. T. (2011). The cultural construction of organizational life:
Introduction to the special issue. Organization Science, 22(2), 287–298.
Weber, R. P. (1990). Basic content analysis (2nd ed.). London: Sage.
Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of Management
Review, 14(4), 490–495.

- 80 -
Whorf, B. L. (1956). Language, thought, and reality: Selected writings of Benjamin Lee Whorf.
Cambridge, MA: Technology Press of Massachusetts Institute of Technology.
Wilson, A. J., & Joseph, J. (2015). Organizational attention and technological search in the
multibusiness firm: Motorola from 1974 to 1997. Bingley, UK: Emerald Group
Publishing.
Yau, C.-K., Porter, A., Newman, N., & Suominen, A. (2014). Clustering scientific documents
with topic modeling. Scientometrics, 100(3), 767–786.
Zajac, E. J., & Fiss, P. C. (2006). The symbolic management of strategic change: Sensegiving
via framing and decoupling. Academy of Management Journal, 49(6), 1173–1193.
Zajac, E. J., & Westphal, J. D. (1994). The costs and benefits of managerial incentives and
monitoring in large U.S. corporations: When is more not better? Strategic Management
Journal, 15(S1), 121–142.
Zhang, Y., Moe, W. W., & Schweidel, D. A. (2017). Modeling the role of message content and
influencers in social media rebroadcasting. International Journal of Research in
Marketing, 34(1), 100–119.
Zhao, E. Y., Fisher, G., Lounsbury, M., & Miller, D. (2017). Optimal distinctiveness:
Broadening the interface between institutional theory and strategic management.
Strategic Management Journal, 38(1), 93–113.
Zhao, E. Y., Ishihara, M., Jennings, P. D., & Lounsbury, M. (2018). Optimal distinctiveness in
the console video game industry: An exemplar-based model of proto-category evolution.
Organization Science, 29(4), 588–611.
Zott, C., & Huy, Q. N. (2007). How entrepreneurs use symbolic management to acquire
resources. Administrative Science Quarterly, 52(1), 70–105.
Zuckerman, E. W. (1999). The categorical imperative: Securities analysts and the illegitimacy
discount. American Journal of Sociology, 104(5), 1398–1438.

- 81 -
Figure 1
A Comparative Assessment of Topic Modeling’s Use
Note: The charts show the number of unique articles published in the social sciences (white bars) and the business/
management literature (black bars) in Scopus and the Web of Science.
4,765
11,398
33,223
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
2003 2005 2007 2009 2011 2013 2015 2017
Total use
Content Analysis
1,199
4,126
11,015
0
2,000
4,000
6,000
8,000
10,000
12,000
2003 2005 2007 2009 2011 2013 2015 2017
Total use
Grounded Theory
1,509
4,034
8,411
0
2,000
4,000
6,000
8,000
10,000
2003 2005 2007 2009 2011 2013 2015 2017
Total use
Natural Language Processing
13
126
1,000
0
200
400
600
800
1000
1200
2003 2005 2007 2009 2011 2013 2015 2017
Total use
Top ic M od eli ng

- 82 -
Figure 2
Topic Modeling Rendering in Theory-Building Spaces

- 83 -
Table 1
Topic Modeling Conceptual Terms
Conceptual
Terms
Definition in the Context of Rendering with Topic Modeling
Algorithm
A process or unambiguous set of rules to be followed, usually by a
computer. An automated processing technique for distilling data inputs into
topic modeling elements (clusters, weights, similarities).
Big textual data
Data characterized by large volume (a million or more words), high variety
(diverse sources), and high temporality (many periods).
Coherence
A quantitative metric for topic quality. Clear and well-bounded topic(s)
with evident criteria for classification of other text or topics within it. Based
on pairs of words in a topic that have high co-document frequencies.
Dictionary
The set of meaningful key words to be used to assess the content and
meaning of a corpus. The basis for annotating words in a text as a code
category.
Disambiguation
A process of using the context to adjudicate between different meanings (or
readings) of a word beyond its literal definition.
Fit
Criteria for how many topics are derived, how they are related, and what
they might mean.
Heteroglossia
Multiple styles of word-use in a single text reflecting different perspectives
or styles of expression.
hLDA
Hierarchical latent Dirichlet allocation—a form of structured LDA.
KWIC
Key words in context; embedding or considering words in their relationship
with other words in a corpus and in a socio-cultural condition.
Lemmatizing
Transforming a word into its dictionary form. In practice, different
lemmatizing methods convert words to their singular forms or by using a
higher-level synonym from a linguistic thesaurus.
LDA
Latent Dirichlet allocation, in which documents are assumed to draw
content from a latent set of topics with probability-based parameters that
can be adjusted to determine those topics.
LIWC
Linguistic inquiry and word count (aka “Luke”) is a dictionary-based,
positive- and negative-affect word frequency program designed to capture
content and affective meaning.
LSI
Latent semantic indexing (LSI) is an algorithm which uses linear algebra to
perform dimensionality reduction and convert texts to a matrix form.
LSVDs
Lasswell Value Dictionary tags
Perplexity
A quantitative metric for the quality of a topic model based on the number
of topics selected. In general, perplexity is a statistical measure of how well
a model fits based on splitting data into a training set and test set. In LDA
topic modeling, it is a relative measure of topic fit; better models have
lower perplexity scores.
Polysemy
Words that have multiple meanings or uses.
Relationality
Words whose meanings are contextually dependent.
Rendering
The process of generating provisional knowledge by iterating between
selecting and trimming raw textual data, applying algorithms and fit criteria

- 84 -
Conceptual
Terms
Definition in the Context of Rendering with Topic Modeling
to surface topics, and creating and building with theoretical artifacts, such
as processes, causal links or measures.
Selecting
Selecting documents (e.g., using sampling) and forms of text to be assessed.
Smoothing
Applying LDA-related algorithms to reduce the number of and disparity
among topics, normally through iteration.
Stemming
The conversion of text segments (words) to their root word forms.
Stop words
Words that serve a less important role in meaning construction (i.e., articles
such as “the” or “a”).
Theoretical
artifact
A construct, conceptual association, process, causal linkage, mechanism or
measure.
Token
The smallest, disaggregated, distinct bit of textual data (normally a noun)
used in analysis.
Topic
A bag of words that frequently appear together across documents; the
derived word(s) from a topic in topic modeling representing word tokens.
Trimming
Reducing textual data and specific words into useful tokens, normally by
lemmatizing and/or stemming; a form of text normalization.

- 85 -
Table 2
Management Subject Areas Enhanced by Topic Modeling Research
Subject Area
Topics
Exemplars
Key Contributions
Detecting
novelty and
emergence
Understanding shifts in
patent citations (#25: patent,
technology, knowledge,
technological, citation,
identify, path, base, cite,
highly)
Kaplan &
Vakili (2015)
Provides a means to
disentangle the cognitive
content of novel
innovations from the
outcomes associated
with innovations
Measuring topics to
understand innovation
(#24: idea, weight,
distribution, edge, measure,
base, node, combination,
average, semantic)
Toubia &
Netzer (2016)
Provides a means of
empirically measuring
different theoretical
dimensions of creativity
to develop new
understandings of idea
generation
Using topic models to
understand managerial
cognition through technology
problems, search and
attention
(#1: problem, search,
structure, attention, concept,
process, exist, unit, create,
general)
Wilson &
Joseph (2015)
Provides a way for
researchers to understand
the dynamics of
managerial attention
relative to background
knowledge.
Understanding knowledge
dynamics
(#14: scientific, impact,
focus, app, knowledge,
article, content, find,
rhetorical, attribute)
Antons et al.
(2018)
Provides a means to
theorize how latent
knowledge structures
undergird innovative
activities
Understanding emerging
organizational forms
(#10: form, identity,
community, logic,
organizational, actor,
institutional, application,
distinct, school)
Jha & Beckman
(2017)
Provides a method for
theorizing the
relationships between
constructs at different
levels of analysis, such
as organizational identity
and institutional logics

- 86 -
Subject Area
Topics
Exemplars
Key Contributions
Developing
inductive
classification
systems
Understanding dynamics of
meanings and networks in
knowledge fields
(#34: article, journal, field,
publish, year, citation,
scholar, papers, author,
paper)
Wang et al.
(2015)
Provides a means to
discover emerging trends
in knowledge fields by
enabling researchers to
identify different
dimensions of
knowledge and connect
these dimensions with
other theoretical
constructs
Understanding how
categories affect competitive
dynamics
(#18: firm, category,
industry, performance,
position, distinctiveness,
competitor, show, level,
competitive)
Haans (2019)
Provides a means to
measure differentiation
associated with cultural
concepts in strategic
action
Understanding the
relationships between risk
and investment
(#31: information, analyst,
report, investor, risk,
discovery, interpretation,
manager, role, find)
Huang et al.
(2017)
Provides a way for
researchers to compare
disparate forms of data
such as written reports
and transcripts of
conference calls
Inducing underlying
meanings associated with
cultural events
(#32: major, rebellion, job,
event, state, report, case,
crime, level, related)
Miller (2013)
Provides a way to
overcome human biases
associated with
interpreting cultural
events
Classifying sets of data and
consumers
(#4: make, pile, task, datum,
set, summary, consumer,
sort, propose, item)
Blanchard,
Aloise, &
Desarbo (2017)
Introduces a new
technique that can be
used to address a classic
consumer behavior
problem of sorting
Understanding
online audiences
and products
The nature of online
consumer profiles (#12: user,
content, message, social-
media, consumer, influence,
individual, role, activity,
platform)
Trusov et al.
(2016)
Provides a means for
conceptualizing
customers as click
groups, networks, and
online communities

- 87 -
Subject Area
Topics
Exemplars
Key Contributions
Online brand recognition and
preference
(#23: brand, approach, car,
text-mining, map, keyword,
association, mention, tag,
consumer)
Netzer et al.
(2012)
Helps capture brand
network attributes and
evolving brand linkages
Online customer evaluations
and responses to them
(#29: review, response,
rating, health, restaurant,
post, hotel, regulation, find,
treatment)
Wang &
Chaudry (2018)
Maps the co-occurrence
of reviews and responses
in real time to
understand performance
adjustment effects
Improving topic modeling of
online audiences and
products (#13: product,
dimension, customer,
consumer, attribute,
purchase, market, prediction,
review, online)
Jacobs et al.
(2016)
Refines topic selection
and supervision criteria,
as well as fit criteria
(e.g., smoothing,
correlation, and
hierarchy across topics)
Analyzing
frames and
social
movements
Understanding how frames
influence political processes
(#27: financial, fomc,
economy, price, market,
hypothesis, macroeconomic,
primary, discussion, real)
Fligstein et al.
(2017)
Provides a means to
identify and measure the
deployment of different
frames in political
activities
The relationship between
frames, context, and
audience
(#6: frame, context,
audience, important,
framing, make, process, give,
individual, part)
Levy &
Franklin (2014)
Enables researchers to
identify distinct
discursive frames
Understanding field-level
relationships between
organizations, discourse, and
strategies
(#17: organization, theme,
individual, effort, people,
comment, strategy, day,
term, field)
Bail et al.
(2017)
Provides a means to
capture sentiment and
bias in normalized
spaces

- 88 -
Subject Area
Topics
Exemplars
Key Contributions
Social movement strategies,
networks, and actions
(#11: group, network,
identify, radical, movement,
pair, environmental, action,
strategy, finding)
Almquist &
Bagozzi (2017)
Provides a means to map
unseen or hidden ties
Understanding
cultural
dynamics
Understanding the
professionalization of a field
(#2: amateur, field,
professional, public, space,
radio, actor, theme,
expertise, expert)
Croidieu & Kim
(2018)
Provides a method for
inductively analyzing a
corpus as part of a
longitudinal case study
Using topic modeling to
analyze big data to
understand cultural trends
(#5: social, conversation,
big-data, language, theory,
cognitive, public, shift,
meaning, emotional)
Wagner-Pacifici
et al. (2013)
Articles that explicitly
describe and illustrate
how to use topic
modeling to extract
meanings from large
corpora
Understanding dynamics
associated with literary
meanings
(#9: work, author, write,
literary, passage, read,
corpus, series, gender, stm)
Tangherlini &
Leonard (2013)
Enables researchers to
identify and compare
meanings across
different sub-corpora
over time
Understanding how cultural
meanings change over time
(#19: art, support, term,
percent, view,
recombination, newspaper,
assign, agency, grant)
DiMaggio et al.
(2013)
Enables researchers to
analyze shifts in cultural
meanings over time
Understanding the evolution
of cultural trends
(#28: time, period, trend,
change, fertility, population,
country, context, british,
demographic)
Marshall (2013)
Uses methods such as
correlated topic
modeling to connect
changes in cultural
meaning over time with
quantitative data