Abstract
Introduction
For better or worse, humankind’s dominion over the planet is largely based on our ability to “cooperate both flexibly and in very large numbers” (Harari in Ji, 2023). This phenomenon is particularly evident in social media (SM), which has become a powerful tool for large-scale human interaction. SM refers to a group of internet-based applications that allow the creation and exchange of user-generated content (Kaplan & Haenlein, 2010). Unlike traditional forms of communication, SM offers flexibility in terms of audience size, personalisation, and content format. It supports multi-modal communication, incorporating a mix of sounds, images, text, videos, and even virtual/augmented reality, and can be tailored to meet individual privacy and efficiency needs (Ji, 2023).
With over 5 billion users (We Are Social Ltd, 2024), SM has become ubiquitous, transforming how people communicate, seek and share knowledge, and interact. Its rise as a novel platform for communication, a vast repository of user-generated data, and an arena for dynamic discourse has increasingly drawn the attention of researchers (Halaweh, 2018). While early SM studies attempted to understand the new phenomenon as well as its challenges and opportunities (Kaplan & Haenlein, 2010), more recently, researchers have turned their attention to its sociological aspects, in particular its ability to provide access to hidden and hard to access groups (Dosek, 2021; Halaweh, 2018; King et al., 2014).
Grounded Theory (GT), a qualitative social science research method, has gained traction in the field of SM studies. The emergence of GT in SM studies emerged around 2012 when Urquhart and Vaast (2012) began theorising about social media environments. Prior to this, GT had gained attention in the related field of information technology (IT), particularly for its usefulness in explaining situationally-based IT processes (Urquhart et al., 2010) and as a systematic and flexible way of generating theory from data (Bryant & Charmaz, 2007; Urquhart & Fernández, 2013).
This scoping review aims to map the GT research in the field of social media. In doing so, it provides a reference point for researchers interested in using grounded theory techniques, processes or methodology in their own social media research. In the introduction, we offer researchers new to GT an entry point to its vocabulary, methodology and processes.
For beginners, GT is probably best thought of as a prolific multicultural and multigenerational family of research methods with philosophical inheritances (Bryant, 2017) and hybridisations. Consequently, its underlying beliefs, language and practices are evolving, which means that, to some extent, there is no definitive shared common language (Bryant, 2017). To illustrate this point, the term GT itself can refer to a philosophically situated method, a logical research process, a constellation of tools or techniques (Bakker, 2019) or an explanation of abstract concepts (Birks & Mills, 2015; Charmaz, 2014) as they relate to the research question. Consequently, our task in this introduction is to create a common language for this article, introduce the main members of the GT family tree and explain common GT processes.
In this article, we use the term “GT method” to refer to a “theory/methods package” that includes a “set of epistemological and ontological assumptions, along with concrete practices” (Clarke & Star, 2008, p. 117) used to undertake research and cannot be interchanged (Clarke et al., 2018) thoughtlessly. The “GT process” refers to an interlinked chain of rigorous analytical practices, and “GT techniques” are analytical devices that serve the process (for a list of GT techniques, refer to Table 2). Finally, “a grounded theory”, or the product of the GT method and process, explains the relationships between abstract concepts and categories (Birks & Mills, 2015; Charmaz, 2014) inherent in participant experiences.
GT Method Family Tree
The first generation of GT, known as Classic GT (Glaser & Strauss, 1967), typically holds positivist assumptions (Sebastian, 2019; Singh & Estefan, 2018). Researchers using this method view the empirical world objectively and spend considerable time collecting and triangulating data (Artinian et al., 2009). This makes Classic GT suitable for generating theories with explanatory, predictive, and transferable power (Artinian et al., 2009; Birks & Mills, 2011). It is well-accepted in fields familiar with quantitative research (Bryant, 2017), like information systems (Orlikowski & Baroudi, 1991).
Strauss and Corbin (1990) introduced second-generation GT, which adopts pragmatist epistemology and interpretivist ontology (Sebastian, 2019). This approach links the beliefs and actions of individuals solving everyday problems (Morgan, 2020; Morse et al., 2021) and is influenced by Blumer’s (1969) symbolic interactionism (Charmaz, 2014; Schwalbe, 2020). It uses narrow research questions and focuses on participants’ actions. Unlike Classic GT, pragmatists view reality as indeterminate and conditional (Charmaz, 2014), making findings explanatory but not necessarily transferable or predictive.
Second-generation GT theorists, including Charmaz and Clarke, aimed to promote social justice (Clarke, 2012). They introduced constructivism, emphasising that research is not neutral (Charmaz, 2009). This methodology views research findings as tentative and highlights the researcher’s role and interpretation (Higginbottom & Lauridsen, 2014; Nagel et al., 2015).
The GT methods family continues to expand and evolve, bringing exciting new “theory/method package” additions such as Transformational GT (Redman-MacLaren & Mills, 2015). Novel methods may offer tools that interweave SM data into analyses, handle the vast quantity of data available online, or help us make sense of this new flexible style of mass communication.
Hallmark GT Processes
The field of GT values flexibility and encourages its theorists to develop their own theory/methods packages and processes for generating theory (Bryant, 2017; Charmaz, 2006; Glaser et al., 1967). Describing the many different GT processes and procedures would take a textbook, and readers are encouraged to consult the many available on the topic. For simplicity, we broadly address the commonalities among GT processes, irrespective of their philosophical underpinnings.
The first hallmark of the GT process is that an explicit hypothesis or theoretical framework is unnecessary to commence the empirical research (Bakker, 2019; Charmaz, 2014), meaning that it is popular among researchers drawn to broad questions about unexplored social terrain such as “How is social media used by health organisations as a method to communicate with the public?”.
The second hallmark of the GT process is that the analysis combines inductive logic with abductive reasoning (Bryant, 2017; Charmaz, 2014). This means that while the analysis moves from specific isolated accounts or observations to make broader generalisations (i.e., induction) (Bryant, 2017), it also involves tacking backwards and forwards between the data collection, analysis and conceptualisation. In other words, as the initial data is analysed, hypotheses emerge that direct the collection of new data or re-evaluation of old data (Clarke, 2014; Morgan, 2020). Abductive reasoning incorporates the third hallmark of GT, known as comparative analysis or “constant comparison”. This step pushes research from the empirical to the theoretical (Glaser et al., 1967). It involves comparing all the analytical material and raw data to explore similarities, differences, variations and surprises (Hallberg, 2006, as cited in Charmaz, 2014) and create more abstract concepts that encapsulate them (Bryant, 2017).
Theory generation is the final hallmark of GT, which we will discuss in this section. While there are nuances between the way different generations of GT understand theory, broadly, a grounded theory is an explanation of the relationships between abstract concepts and categories (Birks & Mills, 2015; Charmaz, 2014) as they relate to the research questions. Abstractions are created incrementally from the data analysis until one or more concepts explain the situation being researched (Bryant, 2017). During the final stages of data analysis, known as theoretical coding, researchers may draw on schemas, concepts or frameworks from broader literature to help sharpen their analysis or tell a unifying story about the data (Charmaz, 2014).
Aims
The aim of this scoping review is to map the application of GT in the field of SM research. By doing so, it provides a reference for researchers who are either new to GT or seeking to apply its techniques, processes, or methods in social media studies. Specifically, we aim to offer insights into how GT methods are used to investigate the unique, fast-evolving digital landscapes of social media platforms.
The review addresses the following questions: 1. What SM topics are being studied? 2. What are the indications for and limitations of using GT in SM studies? 3. What type of GT methods are used in SM studies and why?
Methods
Study Design
As constructivists, we were drawn to the scoping review methodology, over other types of reviews, for its epistemological refutation of absolute truth (Mak & Thomas, 2022; Thomas et al., 2020), and its enthusiastic acceptance of the co-creation “of new knowledge through connections between the research team, the team’s knowledge and past experiences, the data and the topic of interest” (Mak & Thomas, 2022, p. 561). Consequently, we approached the review iteratively and refined the research questions and inclusion/exclusion criteria considering the insights we have garnered from each other and the publication reviewers.
We used Levac et al.’s (2010) extension of Arksey and O’Malley’s (2005) scoping review framework to identify, analyse and synthesise the literature. This involved (1) crafting the search question, (2) balancing feasibility with breadth and comprehensiveness of the scoping process, (3) study selection, (4) data extraction, (5) incorporating numerical summary and qualitative thematic analysis, reporting results and consideration of implications.
The reporting protocol in this review uses the Preferred Reporting Items of Systematic reviews and Meta-Analysis (PRISMA) extension for scoping reviews PRISMA ScR (Tricco et al., 2018).
Eligibility Criteria
Articles included in this review report using a GT approach in SM research. Peer-reviewed journal articles were included if they were published between 2017–2024, written in English and available in full text online. The search terms are detailed in Figure 1. Eligibility Criteria Search Terms.
Inclusion and Exclusion Criteria.
Information Sources and Data Collection
The lead author implemented the initial search in June 2022 in three electronic databases: Social Sciences Citations Index (Web of Science), ProQuest Central and EBSCOhost EJS database. For comprehensiveness, a follow-up search was conducted in June 2024 to identify any additional articles published after the initial search.
Selection of Sources of Evidence
The search query yielded 323 articles. Citations were imported into Endnote and screened for duplicates. 83 articles were removed. Next, each article was screened manually by the lead author for the term “grounded theory”. Articles that used the word “grounded” in a context unrelated to grounded theory were excluded (
Quality Appraisal and Data Extraction
Codes Used in Excel to Demonstrate the Characteristics of Grounded Theory.
Critical Appraisal of Individual Sources of Evidence.
General Characteristics Found in 30 Shortlisted Articles Rated Medium and High.
The final shortlist presented in this review consists of 30 articles. Figure 2 summarises the selection process using the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA). Summary of the Identification of Studies via Databases (PRISMA 2020 Flow Diagram).
Data Charting Process
Authors 1 and 2 charted data from eligible studies using a standardised data abstraction tool designed for this study. They abstracted data on the study aim, indication for using GT, GT method used [explicitly stated or determined by analytical methods (e.g., axial coding) or references cited (e.g., Glaser and Strauss, Strauss and Corbin, Charmaz)], limitations of the research and contribution to the field (see Appendix A).
Thematic Analysis
The eligible studies were analysed using thematic analysis, as outlined by Braun et al. (2019). This process involved six distinct steps. First, Authors 1 and 2 familiarised themselves with the studies by reading and discussing their content and noting interesting patterns, potential connections, and insights relevant to the research questions.
Second, Authors 1 and 2 began generating inductive codes from the data. While the research questions served as a guiding framework, the authors approached the coding process without preconceived notions about the final outcomes. Coding was performed at both the semantic and latent levels, enabling the exploration of both explicit and implicit meanings within the documents.
Third, from the consolidated codes, Author 2 developed preliminary themes, which were then presented to Authors 1 and 3. They assisted in selecting themes that best told the story about the data.
The fourth and fifth steps involved refining and defining the themes collaboratively to ensure that each theme had a clear definition and boundary.
Finally, during the write-up stage (sixth step), the themes were further refined to ensure they remained grounded in the data and answered the research questions.
Results
Characteristics of Sources of Evidence
Each study’s aim, method, method rationale and limitations are presented in Appendix A.
Study Characteristics and Topics
Study Foci.
Summary of the Results
Broadly, the studies either focussed on people’s experiences of using SM or people’s behaviour on SM. When experiences were the focus, studies tended to collect data through interviews; when behaviours were the focus, they used data from SM.
Fourteen articles were positioned as grounded theory research (Alfonzo, 2023; Chaker et al., 2022; Davids & Brown, 2021; Etemadi et al., 2022; Farhangi & Alipour, 2021; Hu et al., 2021; Huang et al., 2021; Kasperiuniene, 2018; Ma et al., 2022; Moore et al., 2019; Nguyen et al., 2020; Throuvala et al., 2019; Zhang et al., 2020; Zingg et al., 2023), three were mixed methods research that incorporated grounded theory (Casey et al., 2022; Thomson & Greenwood, 2017; Zhang et al., 2021), nine were qualitative research that used grounded theory processes or techniques (Adorjan & Ricciardelli, 2021; Cavusoglu & Atik, 2021; Etter & Albu, 2021; Fincham, 2021; Hameleers, 2020; Sarikakis & Winter, 2017; Scheibling, 2020; Tripathi & Verma, 2018; Yang et al., 2021). Ten used theoretical lenses or paradigms to make sense of their findings (Adorjan & Ricciardelli, 2021; Brooks et al., 2021; Casey et al., 2022; Cavusoglu & Atik, 2021; Farhangi & Alipour, 2021; Fincham, 2021; Hameleers, 2022; Scheibling, 2020; Yang et al., 2021; Zhang et al., 2020). See Appendix A for more details.
Main Findings
The findings will be reported under the following headings, which align with the review questions: Indications of using GT in SM studies, Rationales for the GT method and Limitations of using GT in SM studies. We explore each theme below, integrating direct quotes from the articles to illustrate concepts. Where possible, we specify the GT method used in each study. For a detailed breakdown, refer to Appendix A, which provides this information as far as it was discernible.
Indications of Using GT in SM Studies
We identified three reasons why authors chose to use GT for their research projects, which we have labelled
Breaking New Ground
The impetus that most authors cited for their studies was, broadly, to break new ground. GT was viewed as an A further constraint is that many studies examine single aspects linked to social media. These narrow foci affect the generalisability of the research results to organisational realities (p. 2)
GT also offered
For some, breaking new ground denoted uncovering overlooked (Cavusoglu & Atik, 2021) or marginalised voices (Casey et al., 2022; Cavusoglu & Atik, 2021; Mkono, 2019; Nguyen et al., 2020) or examining people’s experiences in sensitive areas (Cavusoglu & Atik, 2021; Moore et al., 2019). Casey et al. (2022), who combined narrative inquiry and (broad) GT in a mixed method study, provide an example in their study on SM use of settler farming women in outback Queensland: Traditionally women’s work and “individual knowledge and social resilience to stressors like climate and climate change have been unrecognized and rendered invisible” and the personalized, recorded experiences of women working through this current drought are scant (pp. 196–197).
In these situations, one of the allures of GT appeared to be the open-ended nature of the interview process, where participants were given the power to
Identifying a Process/Generating Theory
Many authors appeared drawn to GT for its suitability in addressing
Others went further than process identification to advance existing theories. Brooks et al. (2021) did this in their constructivist GT study (which included elements of Corbin and Strauss’ analysis) on SM influencer’s acquisition of cultural capital. They constructed a definitional matrix and conceptual model of the influencer celebrification process and in doing so
Notably, some authors were particularly interested in drawing out the complexity of issues and commented that GT’s emergent theory was
Pragmatics
Many authors used GT techniques in their qualitative or mixed-method studies; their reasons for doing so appear multifaceted and sometimes unclear. A few researchers (Casey et al., 2022; Tripathi & Verma, 2018; Zhang et al., 2021) used GT processes in their respective mixed methods studies to move their findings from the empirical to the theoretical. Scheibling (2020) used GT’s abductive reasoning and sensitising concepts as an analytical device in their qualitative document analysis of men’s parental identities, experiences and ideologies on SM.
In other studies, we could not verify why GT techniques were more suitable than any other type of inductive qualitative analysis. Some examples include Sarikakis and Winter’s (2017) use of Strauss and Corbin’s coding process to analyse interviews in their qualitative study on user’s legal consciousness about privacy, Fincham’s (2023) use of open coding in their qualitative study of the values and concerns regarding SM news audience and Thomson and Greenwood’s (2017) use of GT analysis techniques in their Q methodology study.
We also found evidence suggesting that GT was selected because of its conformity with quantitative approaches to research, which are potentially more accepted in the SM field. Ma et al. (2022), who followed Corbin and Strauss’ process, provide an example in their research on enterprise SM use on employee performance: Furthermore, the grounded theory method has a good balance. It requires researchers to follow strict analytical procedures by a comprehensive utilization of their own empirical knowledge to understand the original materials, so as to achieve the perfect combination of subjective and objective research results (p. 485).
Rationale for GT Method
A variety of articles (e.g., Chaker et al., 2022; Davids & Brown, 2021; Hu et al., 2021; Mkono, 2019; Nguyen et al., 2020) clearly articulated why they had selected their approach to GT and how they followed the GT process. Davids and Brown (2021), for example, discuss how Strauss and Corbin’s method assists with organic, rigorous and systematic theory generation in electronically mediated social contexts. However, only two articles explicitly explained GT as a method (i.e., theory/methods package), outlining the philosophical underpinnings for their research.
First, Kasperiuniene (2018) used Charmaz’s (2014) constructivist GT approach, explaining that they chose co-constructivism for their research on collective creativity in social media processes. This was because
While they did not lean on GT as a method per se, it is worth mentioning that Scheibling (2020) used symbolic interactionism to philosophically amalgamate qualitative document analysis and grounded theory processes and techniques and, likewise, Etter and Albu (2021) used agential realism as a unifying epistemological and ontological lens for their qualitative study combining GT-style inductive analysis and situational analysis.
Limitations of Using GT in SM Studies
Despite authors highlighting GTs ability to draw out the complexity and nuance of a phenomenon, many noted the narrow focus of their exploratory study. This could be because studies were This article is not without its limitations. First, we restricted our empirical endeavour to the communication tactics of right-wing populist actors and citizens who are active on certain Facebook platforms. Right-wing populists do not have issue ownership over media critique, and it may be the case that other politicians (left-wing populist, parties in opposition) blame the media for being inaccurate and dishonest (p. 1150).
Chaker et al. (2022) also noted that their results were time-bound: Although our findings are based on the lived experiences of inside salespeople and sales managers, our interviews were retrospective and occurred at one particular time with each participant (p. 142).
Where these limitations were noted, the authors stated that their findings or theoretical models and frameworks should be treated with caution and could not be generalised to larger populations. They recommended follow-up research with more diverse populations (Kasperiuniene, 2018; Nguyen et al., 2020) to
Discussion
In this scoping review, we set out to map the GT research in the field of SM studies. We found GT studies emerging across a variety of SM topics, propelled by a need for more exploratory research that unpacks the complexity and nuance of (albeit tentative and narrowly focussed) human or psychological processes. Many researchers also appear to combine GT processes and techniques with other research methods and theoretical frameworks to move findings from the empirical to the theoretical. Finally, we found it challenging to determine which GT method was used because it was infrequently described. While there is much to address here, our discussion will centre on the contribution of GT to SM research and a broader conversation about methodological rigour. On this final point, our intention is not to commit methodological zealotry but to help SM researchers consider and, importantly, report their methodological decisions.
GT appears to have much to offer in the field of SM. First, as a well-established exploratory approach (Chun Tie et al., 2019) with a considered approach to controversial research topics with secrets and silences (Charmaz, 2014), GT shows great promise in mapping the uncharted social territory of SM. From a procedural perspective, GT also brings a rigorous approach to exploratory research. This is unsurprising given that Barney Glaser and Absolom Strauss created GT to bring a systematic approach to qualitative research (Bryant, 2017).
We also observed that SM is a rich data source for GT studies; however, at this stage, it appears mainly to use text-based data. We conjecture that the multi-modal nature of SM may spawn new GT theory/methods packages that interweave sounds, images, gifs, text, videos and virtual/augmented reality into analyses. In the meantime, we refer future researchers to Halaweh’s (2018) guidance on the collection of observational, live, or historical data for GT studies for assistance with such data collection.
An important finding of our review was that it was difficult to identify the specific GT method used in most articles, making it challenging to assess methodological rigour and interpret findings. This issue may stem from publication word-count limits as well as researcher or reviewer practices. While we do not support the “slavish attention to following a narrowly prescribed set of procedures” (Creamer, 2022, p. 43), we assert it is important for researchers to demonstrate a foundational grounding in methodology. The GT theory/methods package provides such grounding, guiding data collection, analysis, and reporting (Bryant, 2017) which is especially important for SM researchers combining new online data collection strategies with GT. Understanding the methodological foundations helps to combine and sequence research in mixed methods designs (Morse, 2016), promoting transparency and aiding critique. Moreover, working through these methodological considerations in this new multi-modal and expansive social environment will evolve GT. We contend that standardising (in broad terms) how the application of GT is described in studies could begin to address these issues.
This finding extends to our observations of the use of a priori theoretical frameworks to guide GT analysis. GT is an emergent research method that tries to avoid undue influence from outside factors such as theories, frameworks and discourses (Birks & Mills, 2015; Charmaz, 2006). Many GT scholars feel that using a priori theories or frameworks steers the GT study in a particular direction, potentially overlooking important issues (Charmaz, 2020; Clarke et al., 2018). However, others argue that it is appropriate to use a priori theoretical frameworks, particularly those dealing with social construction (e.g., race, class, gender), to assist researchers in managing their views, assumptions and biases (Mitchell, 2014) or to frame and advance the design of mixed methods GT studies (Creamer, 2022). To remain true to the inductive tenets of GT, we encourage GT researchers to use theories and frameworks as “sensitising concepts”. Sensitising concepts are broad ideas that provide initial, tentative thoughts about research topics (Blumer, 1969) and are used in GT to develop ideas about processes emerging from the data (Charmaz, 2014). Sensitising concepts must earn their way into the emergent analysis and be disregarded if they cannot. In other words, they can guide but should not hijack.
Limitations and Future Research Directions
Like all research, this scoping review has limitations. First, we undertook this review in 2022 with an update in 2024. It is likely that more studies have been published in the time it has taken to do this review. We also completed our literature search after the first round of database searching, limited our search to English and had relatively narrow search terms (i.e., we searched specifically for the term “social media”, not all the different forms of SM). Therefore, it is likely we have missed some articles; however, there is evidence that when ten or more studies are combined, there is a reduced risk of false conclusions (Nussbaumer-Streit et al., 2018); we found 30 articles.
Although it is not a limitation per se, in line with constructivism, the results of this review have been iteratively constructed through the authors’ lenses. Different reviewers may have constructed the findings in different ways.
This scoping review highlights that GT studies are generally small-scale, time-bound exploratory investigations often restricted to homogenous populations due to the reliance on convenience sampling. As a result, the findings should be interpreted cautiously and should not be generalised to broader populations. Future reviews could explore how many GT studies are subsequently expanded into larger-scale or longitudinal projects. Such follow-up investigations could provide insights into whether methodological advancements are required to support GT researchers in “scaling up” their projects to produce research with greater reliability, validity, and broader applicability.
Conclusion
In this scoping review, we set out to map the GT research in the field of SM studies. It is evident that GT is gaining traction and is offering a robust and innovative approach to exploring the uncharted social territory of SM. SM researchers are increasingly using GT to uncover common social processes, generate insightful theoretical frameworks, and draw out the complexity and nuance needed for a more holistic understanding of phenomena. Additionally, we found that authors are creatively incorporating GT techniques into other research methods, showcasing its potential versatility and adaptability. Although the rationale behind this integration was not always clear, it highlights the dynamic and evolving nature of research in this field and GT in general.
Very few articles articulated their GT method. SM researchers considering using or adapting GT methods or techniques are encouraged to thoughtfully consider and clearly report the methodological underpinnings of their research. This will not only assist readers in assessing rigour and interpreting findings but also contribute to the evolution of the GT method itself.
Supplemental Material
Supplemental Material - Using Grounded Theory Method in Social Media Studies: A Scoping Review
Supplemental Material for Using Grounded Theory Method in Social Media Studies: A Scoping Review by Gitte Galea, Lydia Mainey and Ritesh Chugh in International Journal of Qualitative Methods
Footnotes
Declaration of Conflicting Interests
Funding
Supplemental Material
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
