By Authors:
Mia Chang-Zunion – ESCP Business School
Stine Grodal – Northeastern University
Interviewers:
Fabio Busicchia – Polytechnic University of Milan, School of Management
Tailun Chen – Zhejiang University
Article link: https://journals.sagepub.com/doi/full/10.1177/00018392241240319
Q1: How did the idea for this study emerge and evolve? Was the origin of this research more phenomenon- or theory-driven?
This study was definitely phenomenon-driven. We were initially impressed by the remarkable advances in science that have enabled individuals to live with HIV, leading to our initial research interest in this setting. Beyond its medical aspect, the evolution of AIDS as a category was also a socially controversial phenomenon, making it an interesting setting to study some questions related to economic sociology.
Initially, we applied topic modeling techniques on newspaper data, and puzzling themes arose. In particular, we were perplexed by the early categorization of AIDS as a “gay disease”. To investigate this puzzle, we identified oral histories from physicians who had participated during the early part of the epidemic. These oral histories shifted the study toward understanding the broader context of this disease during its early years. AIDS was initially termed “gay-related immunodeficiency.” At one point the appearance of a baby affected by this disease challenged extant beliefs on the disease. However, many doctors and most of the public refused the idea that a baby could have AIDS, and resisted updating the category AIDS beyond the initial stigmatized participants. This was puzzling to us and we began to ponder which consequences this case might have for our theories of category persistence and updating. It was one of those moments which exemplify how in qualitative research it is important to start with a real-world phenomenon to avoid the confirmation bias that comes from approaching data with a specific theory. Focusing on an empirical puzzle allows for more novel theoretical insights to emerge.
Q2: How different is the paper now compared to the first draft in 2019 through the review process? What shifted the most—theoretical lens, process model—and what drove those changes?
During the review process, we were lucky that we did not have to change the theoretical lens completely. Our paper kept its original framing as a categorization paper, but based on the feedback from the reviewers the focus shifted from category coherence to category updating, This led us to debate whether the main focus of the story was category updating or persistence. To address this debate, we collected more data, including media and congressional records and politicians’ frame of this disease.
In revising the paper, we also added more years of data to expand on the empirical story to ultimately show a longer period of category evolution. Initially, our paper ended with the scientific discovery of the virus. However, we wondered if the story of the categorization of AIDS ended there? We found that while the scientific discovery of the virus led to a causal update of the category among a group of core scientists then the majority still clung onto the idea that AIDS was a gay disease, which therefore posed no risk to the general public. We puzzled about what was driving this lack of category updating. In understanding this phenomenon we were aided by the advice of the reviewers who has prompted us to examine the moral dimension of the category. We found that making a distinction between the causal and the moral dimensions of the category allowed us to explain the puzzling phenomenon. To examine the direction between the causal and the moral dimensions of the category, we went back to the data and found interesting empirical support in that story. Based on new findings, we introduced distinctions between those who updated the category early (“vocal minority”) and those who persisted in their initial understandings (“silent majority”).
Additionally, we spent much time thinking about the concept of silencing during the revision. It was difficult to empirically demonstrate something that did not appear in the data and leave a paper trail. It became clear that silencing is more challenging to show than overt contestation, and overcoming this methodological hurdle added much depth to our findings.
Another significant change was when one reviewer asked, “Where is the political story here?” The early version of the paper was mostly focused on how physicians updated the category, but this reviewer was interested in the social and political dynamics surrounding the category AIDS. To answer the reviewer’s question, we collected more data, including congressional reports, and examined the categorization process from a more social lens, which was very generative. This examination allowed us to see how the categorization and the concept of AIDS changed not only as scientists updated their ideas but also as the general public’s understanding evolved alongside it.
Q3: In your view, what are the central takeaways of this research for informing academicians, managers, and policy-makers?
The major takeaway is that we cannot assume that the general public will update their pervious belief on category automatically alongside with the emergence of scientific evidence. This is actually what occurs in the world nowadays. Think about what was happening during the pandemic, the spread of misinformation and scientific denialism. We saw a parallel to how people resisted scientific facts, such as in the case of vaccines and climate change. Academicians and policymakers cannot just denounce the silent majority for not updating their beliefs. We need to identify how to convince them to change, and this cannot be achieved through rational arguments alone. To change societal views, we need to show that aligning these moral values with new scientific understandings is possible. This highlights the importance of understanding how scientific ideas are connected to moral beliefs when trying to change public opinion.
Q4: Some accidental events can create an emotional hype that eventually breaks the silence and pushes the update of the category (e.g., Rock Hudson). Do chance and luck also play an essential role in category emergence?
The announcement of Rock Hudson’s AIDS diagnosis, as the first major celebrity to publicly disclose his condition, triggered a significant shift in media and public attention to AIDS. However, the structural pressures for changes were already in place before this event. Hudson’s announcement acted as a catalyst that made these underlying changes visible. In this sense, while chance events like Hudson’s diagnosis can disrupt the equilibrium, they do not fundamentally create the change. There was already a broader trend occurring—many people were unknowingly affected by HIV, but it was not widely acknowledged until Hudson’s story brought it to the forefront. Due to the many people who were infected, if it had not been Rock Hudson, it had been another person who was perceived to be on the other side of the moral boundary that would have stepped forward and thus punctuated the moral boundary. The important thing is that it was the punctuation of the moral boundary—not the initial scientific discovery of the virus—which changed public opinion about AIDS and recategorized it as not only being a gay disease.
We thus theorized that the response from the society at large was more closely tied to an update of the moral dimension of the category rather than the causal dimension—the scientific discovery itself. Essentially, the change in public perception was driven by a moral re-framing of the disease, not just the scientific breakthrough. Hudson’s diagnosis was the visible manifestation of a shift that had been building silently underneath the surface.
Q5: Considering what happened in the COVID-19 pandemic, is there a possibility that the linear process of breaking moral boundary (and the updating of category) to be reversed back by some social movements?
This is an excellent question to study in the current context. Historically, there was an assumption that social progress, like advancements in women’s rights and LGBTQ+ rights, followed a linear trajectory, with society progressing towards more inclusive norms. However, category updating is not a linear process; the accumulation of knowledge, how we update our knowledge and social norms can be reversed by, for instance, political elections or societal events.
Social changes are often a battle of opposing forces. Progress in one direction can provoke a reaction from groups that feel their values or social dominance are being undermined. These back-and-forth dynamics are a fundamental part of social change. While it may seem like a reversal is occurring, it is essential to recognize that these struggles can still lead to progress, albeit nonlinearly. Even though there may be backlash against LGBTQ+ rights, the situation today is far better than in the 1980s, when LGBTQ+ people faced intense discrimination. Despite possible pushback, we have come a long way, and the trajectory of change is still moving forward, even if it is marked by tension and setbacks.
Q6: Several studies call for using or integrating archival data in management studies. What are the advantages of using archival data for qualitative research? What types of research questions are best suited for archival data?
Archival data in qualitative research allows researchers to study phenomena across different places and over extended periods, where ethnography or interviews would be difficult or impossible. Archival data enables the study of large-scale events, such as the Arab Spring, where researchers cannot physically be in all the countries simultaneously or observe the events in real-time. This makes archival data ideal for studying phenomena that unfold in multiple locations simultaneously or across different historical contexts.
Furthermore, archival data allows for detailed longitudinal analysis. Researchers can track social processes or phenomena over time, which is especially useful for studying things that develop over years or decades. For example, archival data allows us to trace the evolution of a company’s strategies or political movements. By accessing documents, reports, or other records left behind by actors involved in a social process, we can trace actions, decisions, and changes over time, often in more detail than ethnographic data would allow.
However, there are some limitations. Archival data, while providing insights into actions and decisions, does not always capture the meaning behind those actions. Interview data allows researchers to explore the thought processes and meanings people attribute to their actions, something archival data may lack. Additionally, because archival data is public, it reflects how people or organizations want to be perceived, which may introduce biases in the data.
Q7: You choose to present the ‘Process of Data Analysis’ precisely by showing steps to come up with a data structure. To what extent does showing the abductive process strengthen the resulting process theory?
The main reason for having such a figure is to strengthen the rigor and the transparency of the research process. Having a methods figure helps address questions from readers and reviewers like “What were the actual steps that you went through?” or “Why did you make the conclusions that you did?”. However, the figure should also not be judged as a template: different studies should have used other methods and processes to arrive at the insights. Having a methods figure enhances readability and helps readers understand how we came from data to theory. They could follow the main decisions that we made. The figure provided transparency by having a visual expression that was sometimes difficult to convey in the text itself about how we came from one step to the other. We cannot spend pages and pages describing the main turning points, so we need to focus on the most important events so that the reader can follow where we started and what reasoning led us to our final conclusions. Constructing these data analysis, figures and writing the method section thus involve much abstraction. During the research process you generally have a couple of significant “Aha moments” when you go through your data analysis process, making you think differently about your research topic. So, try to focus on those key turning points in the research process. To facilitate creating such an accounts years after it took place scholars can keep a research diary which details key decisions that were made.
Q8: What were the main obstacles in conducting the categorization process from a large amount of data to develop the theory? What actions were most critical to overcoming these challenges?
The big challenge was “seeing the forest for the trees.” Working with archival data can lead to an overwhelming amount of information because archival data are often heterogeneous and abundant. In our case, isolating what was relevant and essential from what was irrelevant and trivial in the data was challenging. The advantage of addressing the data abductively is that the abductive process does not start with detailed data analysis as in induction. Induction begins with comprehensive open coding for the entire data set. It is simply not practical for the vast archival dataset.
The abductive process flips that around. You have to start with a hunch and try to look for something that is an anomaly or something puzzling by skimming through the data. You are constructing a scaffold of your data. To do that, you need to figure out what data you have and its main building blocks while searching for anomalies. Then, you will try to verify whether this anomaly really exists. You try to find data that supports or rejects the idea that there is an anomaly. Once you have the anomaly, you will search the data for elements that can explain that anomaly. When you have identified a potential explanation for your anomaly, then you begin to more thoroughly analyze the data. It starts by scanning through the data and looking for examples that will either support or oppose your explanation. If you find data which oppose your explanation, then you will move on to try to identify a new explanation. If you find data which supports your explanation when you will try to find data to elaborate on the explanation, which becomes more and more specific over time. At the very end when you have gained a lot of confidence in the explanation then you might engage in structured coding to show your explanation more precisely.
Another problem in some categorization studies is the relationship among different levels of analysis in the dataset. Individuals can talk. However, scholars want to theorize about the category at the macro-level, but the category itself does not “talk.” Then, you need to view individual statements by informants as representing the macro categorical construct. An important part of this process is often triangulations among informants’ statements and different data sources, for example by comparing informants statements with news and media articles. So, you must figure out how to build a scaffold from data. You need to have a structured approach to collecting, analyzing, and triangulating between different data sources.
The essence of qualitative archival work lies in the intense interpretive process of analyzing diverse data sources. What defines us as qualitative researchers is not the kind of data that we use (whether it is interviews, participant observations, images, artifacts or even quantitative numbers), but our theory-building approach. Qualitative scholars distinguish themselves by developing new theories through interpretation, regardless of the data type—the process makes us qualitative researchers.
Q9: How can qualitative researchers balance transparency in the methodology section with leaving enough room for findings and contributions?
Transparency is all about trying to show the primary turns you took. Because you do not have unlimited space, your strategy should be to show “Aha moments”, where you collected new data and realized that you need to shift the analysis dramatically in this way. It is like if you want to create a diagram of how you bicycled across the United States, then you would also only be able to draw a macroscopic representation of the significant turns you made in your path—not all the little turns that you took up a winding road. Put yourself in the mind of the reviewer and the reader. Think about the major points you took that will allow the reader to understand and follow what you did.
Writing a qualitative paper is all about not getting lost in detail. You need to describe the overall pattern, not every little turn you took, even if you thought it essential. Even if you spend three months on it, it might not make it into the methods section simply because in the end it did not result in a major turn in your analysis. You could have an appendix to go into more detail about it.
This is not just about writing the methods section; the whole process of writing from methods to findings involves much abstraction. You read hundreds of interesting things from your data, but you may mention it in one sentence sometimes. At the end of the day, you need to condense all these tiny details into just one sentence so that readers can read it and get a basic sense of what was going on. It is a very dense kind of writing process.
Q10: Physicians are one of the critical actors in the process of AIDS updating; what are some shaping actions they take during that process? How might these actions inspire today’s minority experts addressing challenges such as climate change or anti-intellectualism?
One important implication of our study was that if these minority experts want to do something and make a change, they cannot just focus on the causal mechanisms and the science itself. They must understand how these causal understandings are embedded in larger categorical structures. Both casual and moral dimensions need to be updated simultaneously because people seek coherence in their categorical understandings of the world, and you cannot expect them to update one element of that category if it creates dissonance within the category.
The advice here is that if you want to tackle climate change and anti-intellectualism, the most important thing is to address the moral or value-based dimensions. Instead of saying, “You were an idiot for not accepting our scientific fact,” professionals need to understand how climate change relates to all the values people hold. They should understand why believing in climate change might threaten these beliefs. Professions must puncture the moral boundary, for example, by showing people that for example green technology is not opposed to other values that people might hold like a more traditional American way of life. You cannot just be pushing the idea of climate change without understanding how people’s beliefs about it relate to many other values they hold.
Q11: This is a good example of “impactful research”. Can you give any advice for junior faculty who want to publish impactful research in journals like ASQ?
It is crucial to focus on an essential empirical phenomenon, instead of an esoteric interest that can only reflect a small domain. As organizational scholars, junior and senior alike, it is essential to become more selective when choosing research projects to focus on contexts important for society and businesses.
As organizational scholars, we may feel that studying a niche phenomenon can contribute to theory in an interesting way. However, to communicate those findings to the general public, we must translate them from one context to another. As scholars, we are apt at doing this translation work. But if you talk to practitioners, they are much more embedded inside their context and see less similarity between different contexts than we do.
It is crucial to study prevalent, significant, and plentiful contexts in the real world. Maybe we have some “pet area” we are interested in. But we honestly owe it to society and to the people who pay our salaries, who pay for research, to study phenomena that are prevalent in the real world, and where it is easier for practitioners to understand why they are essential and impactful.
In short, try to find an empirical area that is big, plentiful, and important to practitioners, and then also focus on starting theories from the phenomenon rather than from the theory.
Interviewer bios:
Fabio Busicchia is a PhD Candidate at the School of Management of Politecnico di Milano, Italy. Fabio has conducted research activities at Bayes Business School, London. He is also a lecturer in Industrial Economics at the Graduate School of Management. His research activities focus on industrial dynamics in nascent industries, market entry strategies, and technology legitimacy. Fabio has taken an interest in the drone industry and studied its evolution.
Tailun Chen is a Ph.D. candidate in Technology and Innovation Management at the School of Management, Zhejiang University. His research interests involve the non-market strategy, category emergence, innovation diffusion and strategic change.