Why the Qualitative Versus Quantitative Divide is Overrated

When I got a job as a lecturer in qualitative research methods, it seemed obvious to me that I was a qualitative researcher. Since then, five years of teaching qualitative methods and research design to qualitative and quantitative students from across the social sciences have made me growingly uncomfortable with the "qualitative" versus "quantitative" binary categorisation. It often feels like I am pushing categories inherited from previous (established) generations onto new generations. We teach students to problematise categories but ask them to take these for granted. We ask students to be cautious about elements of their socialisation that may lead to social reproduction and come in the way of thinking differently, yet repeat potentially rigid ways of thinking methodology. This practice does not match my pedagogy let alone the values that guide my academic work and the way I see my life.
Two types of literature have helped me fuel this reflection. On the one hand, literature has developed teaching material to help identify “what is qualitative” and support teachers of “qualitative” modules to answer this question (Flick et al., 2004; Martyn Hammersley, 2013). Some authors also tend to directly engage the differences between qualitative and quantitative approaches, for example in the context of mixed-method research (see Baškarada and Koronios, 2018; Creswell and Plano Clark, 2017). On the other hand, researchers have provocatively attempted to nuance these differences by stressing the idea that qualitative dimensions of research are “present in quantitative work as well” (Aspers and Corte, 2019: 9413), shifting the conversation regarding what qualitative research is by arguing that “the answer does not matter but the question is important” (Small, 2021), or “transcend[ing] or even subvert[ing] the so-called qualitative-quantitative divide” (Mason, 2006: 9).
In this blog post, I introduce the main points that helped me challenge my socialisation about what is qualitative research and challenge my perceptions about the qualitative versus quantitative divide. Following my own journey, I synthesise what I believed was something specifically qualitative about research before introducing the argument that made me question this belief: in terms of data, method, research design, and standards and values followed by research. As we will see, these ideas are not individual and are in fact discourses commonly found in the methodological literature.
Through analytically organising four arguments that helped me debunk my own assumptions, I hope that this piece can help readers put into perspective their own in an accessible and easy-to-digest way, for example as a brainstorming exercise for a much-needed conversation in methodology seminars. As such, I did not write this piece as a way to denunciate colleagues still finding good use in the binary or as a way to put forward a unique version of what I define as qualitative. This short piece also does not aim to provide an exhaustive review of all the ways qualitative research has been defined in the literature (for readers interested in such literature review, see for example (Aspers and Corte 2019)). Rather I wanted to contribute to put forward the need to unveil, question, and problematise taken-for-granted assumptions and blinders that may come with usual methodological categorisation to avoid further reifying these positions. I wanted to write this essay as an invitation and a provocation to challenge categories that unfortunately sometimes get in the way of not only useful methodological cross-pollination but also healthy academic (pluralistic) environments, in the hope that we can socialise next generations in less entrenched and rigid categories.

I. “Qualitative” (versus “Quantitative”) data

One definition of qualitative research I commonly encountered is that research is qualitative when it uses qualitative data. Non-numerical data for example include “rich description given by people (NOT necessary randomly selected – more a kind of a purposeful sample), via interview transcripts, archival documents, descriptive observations, historical and non-historical documents, archaeological-type of ancient languages, though also, through emerging data sources including visual data such as photos and generated art, as well as reflexive logbooks” (Tenenbaum et al., 2011: 349). As such qualitative data is defined as non-numerical in opposition to quantitative data which is defined as numerical (Kalra et al., 2013; Nassaji, 2020).
However, this otherwise straightforward distinction between qualitative and quantitative data does not necessarily hold under scrutiny. On one hand, how we approach so-called ‘qualitative’ data is often implicitly quantitative. Indeed, as Benoit notes:
“… relations such as stronger or greater imply, whether this is made explicit or not, a relative degree of quantity, even if the characteristic being compared is discussed in purely qualitative terms. The act of comparison, therefore, naturally and readily lends itself to quantification.” (Benoit 2005, 10)
On the other hand, quantitative data is rarely purely numerical in all the stages of the research process, a classic example of this being textual data. Used for quantitative analysis, textual data is non-numerical by nature and must be converted into numbers to be analysed quantitively before being transformed into words again in the process of analysis. Moreover, one can note that in the current context of increased computational power and the democratisation of specialised softwares, less and less statistical knowledge and direct manipulation of this quantification process is required in the hand of the researchers as researchers can growingly adopt packages and models that do this conversion for them (see for example Quanteda Guru for quantitative text analysis or H2O AutoML for machine learning).

II. “Qualitative” (versus “Quantitative”) methods

A second conventional way of distinguishing qualitative and quantitative research I encountered deals with the type of methods used. If the research employs a method of data collection (e.g., interviews, participant observation) perceived as qualitative, then the project is deemed qualitative (Flick, 2017). If the research employs a method of data analysis perceived as qualitative (e.g., discourse analysis, visual analysis) then the project is deemed qualitative (Dey, 1993).
But what do make some methods to be labelled as qualitative in the first place? Qualitative methods of data collection are often associated with direct interactions with human beings – qualitative methods “study persons by directly interacting with them” (Aluwihare-Samaranayake, 2012: 65)– as well as immersion – one goes in the field and spends time with the social groups studied. Qualitative methods of analysis are often associated with the study of perceptions and meanings; one does not only investigate what is but also/rather how people experience things and talk about them (Denzin & Lincoln, 2011).
 
However, in contrast with this neat division, one can see that some methods labelled as ‘quantitative’ also directly engage with human participants – e.g., experiments – or aim to understand perceptions – e.g., surveys. In addition, many qualitative methods of collection and analysis do not involve the researcher being physically present or directly interacting with the population studied – e.g., qualitative research focusing on documents or social media data.

III. “Qualitative” (versus “Quantitative”) ways of doing research

A third element used to define qualitative research and to define it in contrast with quantitative research deals with ways of doing research, which some may refer to as epistemological dimensions. Qualitative methods are often qualified as ‘interpretative’ in contrast to ‘a natural scientific model in quantitative research’ which some refer to as positivism ((Clark et al., 2008: 266) cited in (Lamont, 2021, p. 95)). These terms are not often precisely defined. However, one implication of this distinction is that quantitative research is put forward as capable of providing evidence demonstrating causal links in a way that qualitative research cannot (Gunter, 2013). Finally, scholarship sometimes defines and distinguishes qualitative and quantitative research by emphasising the former’s predilection to focus on micro-processes and the latter being best equipped to investigate macro phenomena (Kelle, 2001: 103).
While such perceptions are commonly circulated, there is an established body of literature showing that these distinctions are far from clear-cut. Seminal texts such as King et al. (1994) have long argued that “the logic of inference, or positivism, unites qualitative and quantitative approaches to research.” (Lamont, 2021: 95). Moreover, qualitative methods are growingly mobilised to investigate causality qualitatively, such as is the case with process tracing (Waldner 2015). Finally, many qualitative projects aim to make macro claims, for example by developing elaborated multi-method qualitative research designs to study complex objects such as globalization and internationalization processes. Growing arguments for ‘Big Qual’ datasets also reflect a trend towards macro qualitative work (Davidson et al., 2019; Taylor & Schroeder, 2015).

IV. "Qualitative sensitivities” unique to “Qualitative” Research

​Finally, qualitative research is often defined by research values and attention to specific dimensions of research considered characteristic of the qualitative tradition. Some authors refer to them as “qualitative sensitivities” (Tanweer et al., 2021) but these could also be interpreted as specific standards for evaluating research quality. The following points are routinely presented as specific to qualitative research and less emphasized in quantitative research:
Interpretation. Interpretation is considered an important dimension of knowledge production within qualitative research. As such, Tanweer et al. (2021, p. 1) emphasise the “interpretivist lens” as one of the “qualitative sensitivities” they put forward. According to them, the interpretivist lens is “concerned with the construction of meaning in a given context”. In my understanding, this applies not only to the world we study but also to how we approach the process through which we conduct research. Every time we transform data into findings we engage in a process of “this means that” – a meaning-making process – that is one of the most important steps of research. As such, any act of analysis is an interpretative act (author, 2019) and interpretation should be taken seriously when it comes to validity and quality of research as this act, constitutive of the jump from data to results is where many errors and biases become embedded in our work.
The human element of research. Two dimensions are summarized under this point. On one hand, the human element of research concerns who is researched and how often with the consideration of avoiding producing harm via our research. This concern has led to an established body of qualitative research literature focusing on research ethics (Kostovicova and Knott, 2022). On the other hand, the human element of research deals with how elements of our socialization, trajectory, and positionality, affect knowledge production and what we should do about it. Here, the keyword is reflexivity, understood "as the practice of making conscious and explicit our practices, assumptions and dispositions" (author). Beyond these key dimensions, qualitative scholarship has also introduced socio-psychological concepts to the methodological literature to foster engagement with the human dimension of research and guide researchers in this regard. One such is “emotional intelligence” defined by (Salovey and Mayer, 1990) "as being able to monitor and regulate feelings to guide thought and action through five basic competencies: self-awareness, self-regulation, motivation, empathy, and social skills."
Context. Finally, the importance of the context in which the phenomenon studied happens has also traditionally been put forward as an important question in qualitative research (Miller and Dingwall, 1997). This means approaching social phenomena as context-dependent and mobilising primary and secondary sources to produce contextualisation in order to identify how so. Namely, it follows the idea that social problems can only be understood and explained in relation to the contexts (social, political, historical, economic, textual, cultural…) in which they emerge. In that sense, sensitivity to context helps us make sure that claims match empirical evidence in order to avoid problems such as overgeneralising or misattributing characteristics to human subjects that in fact do not match their reality (see for example the literature investigating the problem of Eurocentrism in social sciences).
In regard to qualitative sensitivities and how they are mobilised to distinguish qualitative from quantitative work, I make a normative argument. Qualitative research has distinguished itself for being sensitive and careful about the human element of research, how we interpret, and the contexts in which what study happens. It is unclear why these criteria are only relevant for qualitative research and why they should not be valued and promoted within social science research at large. Indeed, these elements seem both likely to increase the validity and quality of research and engage with socio-political implications of research which are growingly put forward beyond qualitative circles, for example in computational social science within and beyond academia.

Conclusion

When I teach methodology, students often ask me what is qualitative research and how it differs from quantitative research. This question is at the forefront of students' learning methodology and how I am asked to organise and label my modules and lectures. In contrast, it seems that established colleagues asked these questions less and less the further into their careers. Identities become comfortably established across the binary as their research programme becomes easily identifiable through these terms.
In this blog post, I aimed to show that without being completely obsolete, the qualitative versus quantitative binary is largely overrated and does not deserve to organise curricula, research programmes, and academic fields in such an extensive way. Social science methodologists pride themselves on the rigour and quality of their analysis. However, when it comes to their own research practice, the lack of precision of the categories we commonly use is striking. Identities and traditions prevail over critical thinking and updating our frameworks to case-by-case situations. The same standards and questioning attitude with which we uphold ourselves when it comes to categories of observation and measurement should also be implemented when it comes to the categories we use in our everyday professional life, as these impact how we work together across the knowledge boundaries thus created.
To conclude, I invite the readers to reflect on the way they relate to the categories qualitative and quantitative and their traditional opposition when it comes to research methodology. Here is a combination of factual, reflective and normative questions to open the floor to the readers and help foster a collective conversation:
 
•       Can you relate to the common perceptions highlighted in this research note? Do you use the categories qualitative versus quantitative often in your everyday work? How so?
•       What are the potential benefits of using (and opposing) the binary qual. versus quant.?
•       Is it useful to mediate the field of methodology through this binary? What are the potential negative consequences of doing so?
•       Shall we still keep socialising the new generations in these old categories? Why so? Does it contribute to producing better knowledge and healthier academic environments?
•       What are the sociological, psychological, political and economic dynamics behind the ongoing success of the use of this binary in social sciences? Who does it benefit? To whose detriment?
•       Why do we keep saying we are qualitative? What does it bring us?

References

​Aluwihare-Samaranayake D (2012) Ethics in Qualitative Research: A View of the Participants’ and Researchers’ World from a Critical Standpoint. International Journal of Qualitative Methods 11(2): 64–81.
Aspers P and Corte U (2019) What is Qualitative in Qualitative Research. Qualitative Sociology 42(2). Springer New York LLC: 139–160.
Baškarada S and Koronios A (2018) A philosophical discussion of qualitative, quantitative, and mixed methods research in social science. Qualitative Research Journal 18(1): 2–21.
Clark T, Foster L, Sloan L, et al. (2008) Bryman’s Social Research Methods.
Creswell JW and Plano Clark VL (2017) Designing and Conducting Mixed Methods Research | SAGE Publications Ltd.
Davidson E, Edwards R, Jamieson L, et al. (2019) Big data, qualitative style: a breadth-and-depth method for working with large amounts of secondary qualitative data. Quality and Quantity 53(1). Springer Netherlands: 363–376.
Denzin NK and Lincoln YS (2011) The Sage Handbook of Qualitative Research. SAGE Publications, Inc.
Dey Ian (1993) Qualitative Data Analysis: A User-Friendly Guide for Social Scientists. New York, NY: Routledge.
Flick U (2017) The Sage Handbook of Qualitative Data Collection. SAGE Publications Ltd.
Flick U, Kardorff E Von and Steinke I (2004) What is Qualitative Research? An Introduction to the Field. A Companion to Qualitative Research (May). Epub ahead of print 2004.
Gunter B (2013) The quantitative research process. A Handbook of Media and Communication Research: Qualitative and quantitative methodologies: 2nd edition. Taylor and Francis: 237–264.
Kalra S, Pathak V and Jena B (2013) Qualitative research. Perspectives in Clinical Research 4(3). Medknow: 192.
Kelle U (2001) Sociological Explanations between Micro and Macro and the Integration of Qualitative and Quantitative Methods. Forum Qualitative Sozialforschung 2(1).
Kostovicova D and Knott E (2022) Harm, change and unpredictability: the ethics of interviews in conflict research. Qualitative Research 22(1): 56–73.
Lamont C (2021) Research Methods in International Relations. SAGE Publications Ltd.
Martyn Hammersley (2013) What Is Qualitative Research. London: Bloomsbury.
Mason J (2006) Mixing methods in a qualitatively driven way. Qualitative Research 6(1).
Miller Gale and Dingwall Robert (1997) Context and Method in Qualitative Research. Context and Method in Qualitative Research. SAGE Publications Ltd: 1–240.
Nassaji H (2020) Good qualitative research. Language Teaching Research 24(4): 427–431.
Salovey P and Mayer JD (1990) Emotional Intelligence. Imagination, Cognition and Personality 9(3): 185–211.
Small ML (2021) What is “Qualitative” in Qualitative Research? Why the Answer Does not Matter but the Question is Important. Qualitative Sociology 44(4). Springer: 567–574.
Tanweer A, Gade EK, Krafft PM, et al. (2021) Why the Data Revolution Needs Qualitative Methods. Harvard Data Science Review. Epub ahead of print 17 June 2021.
Taylor L and Schroeder R (2015) Is bigger better? The emergence of big data as a tool for international development policy. GeoJournal 80(4). Springer Netherlands: 503–518.
Tenenbaum G, Gershgoren L and Schinke RJ (2011) Non-numerical data as data: a positivistic perspective. Qualitative Research in Sport, Exercise and Health 3(3): 349–361.
Waldner D (2015) Process Tracing and Qualitative Causal Inference. Security Studies 24(2). Routledge: 239–250.