Addressing Injustices in and Through Research as a
Methodological Rich Point
Featured article by Adrian Lundberg, Malmö University
What are the Alternatives to Traditional Research Methodologies?
During a time of urgent need for a more equitable and just society, we as a research community should look for alternative ways to listen to participants’ voices without imposing our own categories upon them. This contribution’s purpose is therefore to elaborate on a methodological rich point (Hornberger, 2013) which can help to establish unprejudiced dialogues.
Selecting appropriate research methodologies and methods can be challenging, in particular for early career scholars with comparatively narrow experience and, hence, a potentially limited methodological toolkit (Hult, 2010). Researchers usually choose to investigate what interests them and what they have strong views about (Ricento, 2015). Particularly those who actively engage with participant subjectivity may often struggle with accessing and representing participants’ own viewpoints in a neutral and just manner. To stimulate your thought process regarding your researcher positionality (Lin, 2015), I invite you to respond to the following questions:
Who are you?
Who are your participants?
How do you make sure that, in your research, you represent their views and not your own?
My Methodological Rich Point
As for myself, a Sweden-based language teacher and researcher from Switzerland, I quickly realized after relocating to Scandinavia that some central phenomena surrounding educational language policy and practice, such as multilingualism or multilingual students, are not conceptualized the same way in these seemingly similar countries in the Global North.*** Hence, I needed to find an approach to my research inquiries that allowed me to investigate the complexity of teachers’ own viewpoints regarding their theoretical understanding(s) of multilingualism and multilingual students, and how those viewpoints might influence their pedagogical actions as key policy arbiters (Menken & García, 2010). In the following sections, I present my self-reflexive account of this methodological rich point, which resulted in my choice of using Q methodology (Stephenson, 1935) for my doctoral dissertation on viewpoints about educational language policies (Lundberg, 2020).
Q Methodology to Study Subjectivity
Humans are often unaware of the reasons for their actions. The assumption that features of teacher cognition might neither be observable nor entirely consciously accessible (Rokeach, 1968) adds a further layer to the methodological challenge. Purely asking teachers about their views in an interview situation will not help us reach the most interesting, often underlying thought patterns. Moreover, interview situations bear other pitfalls, such as social desirability issues or the exclusion of non-verbal participants.
In Q methodology (a suggested Q methodological research process is visually represented in Figure 1), participants are presented with a set of items to rank-order in a Q card sorting activity. In this process, there is no need for them to actively produce written or spoken language. As Q researchers, we thoughtfully and carefully construct that data collection instrument. We collect items from a range of sources, preferably including statements by members of the same community we intend to investigate (step 1). Items are then culled to present a balanced and representative sample of all that could be said about the topic in question (step 2). Multiple rounds of pilot studies and evaluations by external experts help avoid any researcher bias in the final Q set. In my doctoral research, I started out with more than 100 potential items and ended up with 39 single-barreled and diverse responses to the question What does “multilingualism” and “multilingual students” mean according to you?
Figure 1: Suggested process of Q methodological research (Lundberg, de Leeuw, & Aliani, 2020)
During the card sorting activity, participants make numerous decisions to present the items in a constellation that best represents the entirety of their subjective viewpoints (step 3). To somewhat reduce the number of participants’ decisions and receive data in a convenient and easily processable form, participants are asked to place all items on a distribution grid (Figure 2) and thereby create their personal Q sort. The grid represents a continuum, for example, from “least agreeable” to “most agreeable,” and assigning every item a value on such a grid points out the item’s inherent relativity to the other statements, made visible through the sorting process in Q methodology. The vertical distribution of items is irrelevant in the illustrated grid.
Q sorting, in combination with the described Q factor analysis further down, allows to investigate participants’ viewpoints about a single question at hand in a much more complex manner than more mainstream approaches, such as interview- or survey-based research. Despite the option in these more traditional approaches to group questions in a Likert scale survey study design, they are answered separately and independently from each other. In Q methodology, respondents provide their viewpoint regarding one question—this, however, in a highly detailed and multi-facetted manner.
Figure 2: Empty distribution grid for 39 items. Numbers in parentheses represent
the requested number of items to be assigned per value.
This highly engaging and thought-provoking procedure reduces participants’ ability to respond in a socially desirable way (Fluckinger, 2014). Often, there are follow-up questions to the card sorting activity (step 4). In these so-called post-sorting activities, more qualitative information, for example, about the reasons for placing particular items at the extreme positions, can be collected in written or oral form to eventually support data interpretation.
Q Factor Analysis and Interpretation
The type of analysis in Q methodology provides further opportunities to study marginalized voices. Purely quantitative analysis often results in the concealment of less mainstream responses due to the reporting of either averages across different demographic variables and/or the majority’s view. In Q methodology, on the other hand, researchers are interested in illustrating the range and diversity of viewpoints on a particular topic, guided by the principle that there is no right or wrong subjectivity (Brown, 1980).
To find groups of like-minded people within the purposefully selected sample of participants, the researcher inserts all completed distribution grids into software that correlates individual Q sorts and applies inverted factor analysis (step 5). This data condensation technique results in a number of newly created Q sorts. These new Q sorts form the basis for the subsequent interpretation, an iterative procedure guided by an abductive logic that aims to discover likely causes of participants’ sorting through informed guessing. The result is the emergence of participants’ own shared viewpoints (step 6).
As a concrete example of Q methodological results, I describe three statistically diverse viewpoints about multilingualism and multilingual students among Swedish primary school teachers (Lundberg, 2020). They are represented by a summarizing label for mnemonic reasons:
Let us help you exercise your right to be multilingual.
You’re tolerated. Now adapt.
Why make such a fuss? Everyone will be fine.
Combining Q methodology with a participatory research approach is especially intriguing. An increased sense of accountability and transparency can be achieved by inviting participants to contribute to various stages of the research process, which seems to be of particular importance during policy formation. In addition, this allows researchers to further minimize their own influence. To illustrate that, I allow myself to add an example from another small research project about bullying in primary schools (Hellström & Lundberg, 2020).
The most common definitions of bullying are generally based on adult-imposed categories. However, if we aim to account for students’ needs in school, we have to invite young people to express their perspectives, and research should aim to include their unbiased voices more often. My colleague and I have used Q methodology to elicit students’ own definitions of bullying. We first involved them in an informal discussion about their usage of various social media, online communication and online games. Their statements, in combination with literature on offline bullying, were used as the basis for constructing bullying scenarios that were as authentic and close to the participants’ everyday lives as possible. The Q methodological results from that study are expected to contribute to the development of anti-bullying strategies tailored to students’ needs. Drawing parallels between school bullying and JPB Gerald’s featured article in the fall 2019 AAALGrads Newsletter on choosing the right name for racial discrimination in language studies, it becomes obvious that we as researchers need to be highly sensitive to and aware of the categories and methodological approaches we apply in our research.
Q Methodology as a Promising Tool in Applied Linguistics
This brings me to the main conclusion for this featured article that aimed to present Q methodology as a contribution to more equitable, diverse, and racially just spaces. I am convinced that Q methodology is a promising tool in any researcher's repertoire, be it non-BIPOC scholars conducting research on and with BIPOC individuals, or academics who aim to engage with communities posing that systemic racism is not real. It is the built-in features of Q methodology that help us see things from our research participants’ point of view and eventually adapt policy and practice towards more social justice.
*** While the predominant Swedish understanding of multilingualism is based on language acquisition due to biographic reasons (e.g. migration), Swiss respondents tend to focus on curricular aspects of language learning (Lundberg, 2020).
Brown, S. R. (1980). Political subjectivity. Yale University Press.
Fluckinger, C. D. (2014). Big five measurement via Q-sort: An alternative method for constraining socially desirable responding. SAGE Open, 4(3), 1-8. https://doi.org/10.1177/2158244014547196
Gerald, J. (2019). Choosing the right name for racial discrimination in language studies. AAALGrads Newsletter, 4(1), 7-9.
Hellström, L., & Lundberg, A. (2020). Understanding bullying from young people’s perspectives: An exploratory study. Educational Research, 62(4), 414-433. doi:10.1080/00131881.2020.1821388
Hornberger, N. H. (2013). Negotiating methodological rich points in the ethnography of language policy. International Journal of the Sociology of Language, 219, 101-122. doi:10.1515/ijsl-2013-0006
Hult, F. M. (2010). Theme-based research in the transdisciplinary field of educational linguistics. In F. M. Hult (Ed.), Directions and prospects for educational linguistics (Vol. 11, pp. 19-32). Springer.
Lin, A. M. Y. (2015). Researcher positionality. In F. M. Hult & D. C. Johnson (Eds.), Research methods in language policy and planning: A practical guide (pp. 21-32). Wiley-Blackwell.
Lundberg, A. (2020). Viewpoints about educational language policies: Multilingualism in Sweden and Switzerland. Malmö University.
Lundberg, A., de Leeuw, R. R., & Aliani, R. (2020). Using Q methodology: Sorting out subjectivity in educational research. Educational Research Review, 31, 100361. doi:10.1016/j.edurev.2020.100361
Menken, K., & García, O. (2010). Negotiating language policies in schools: Educators as policymakers. Routledge.
Ricento, T. (2015). Foreword. In F. M. Hult & D. C. Johnson (Eds.), Research methods in language policy and planning: A practical guide (pp. xii-xiv). Wiley-Blackwell.
Rokeach, M. (1968). Beliefs, attitudes, and values: A theory of organization and change. Jossey-Bass Publishers.
Stephenson, W. (1935). Technique of factor analysis. Nature, 136, 297. https://doi.org/10.1038/136297b0