Action Research in Learning and Classroom Contexts

IJAR – International Journal of Action Research 3-2022: Retrospective Action Research on Facilitating Equitable Learning Outcomes in a Diverse Class

Retrospective Action Research on Facilitating Equitable Learning Outcomes in a Diverse Class

Ariane Janse van Rensburg

IJAR – International Journal of Action Research, Issue 3-2022, pp. 247-267


Abstract: In a South African class with complex diversity, certain student groupings were not performing equitably in relation to their potential. An Educational Action Research (AR) process of designing multiple, integrated practice changes over three years successfully redressed disparities, but the full impact of interventions could only be analysed in retrospect. Combining empirical observations with subsequent data collection to produce a theorised model, a transferable methodology using quantitative triangulation was designed to overcome the challenges of a rigorous retrospective AR study. This article discusses the integrated teaching interventions and the application of retrospective AR methodology.

Keywords: action research, retrospective study, equitable learning outcomes, diversity


La Investigación Acción restrospectiva en la facilitación de resultados equitativos del aprendizaje en una clase diversa

Resumen: En una clase de Sudáfrica con diversidad compleja, algunos grupos de estudiantes no estaban teniendo logros que se correspondían equitativamente con su potencial. Un proceso de investigación acción educativa (IA) en el que se diseñaron cambios múltiples e integrados en la práctica a lo largo de tres años ajustó estas disparidades con éxito, pero el conjunto del impacto de las intervenciones sólo pudo ser analizado en retrospectiva. Mediante la combinación de observaciones empíricas y la subsiguiente recolección de datos para producir un modelo teorizado, se ha diseñado una metodología transferible que utiliza una triangulación cuantitativa para superar los retos a los que se enfrenta un estudio de IA retrospectivo riguroso. Este artículo debate las intervenciones educativas y la aplicación de la metodología retrospectiva de IA que se produjeron de forma integrada.

Palabras clave: Investigación acción, estudio retrospectivo, resultados de aprendizaje equitativos, diversidad


1 The problem context

This project did not start as an educational research study – it was simply an urgent teaching problem to be solved. I had started teaching Architectural Design (AD) to a first-year class in a Bachelor of Architectural Studies (BAS) degree at a South African university, using the preexisting course, and found that in a class of students coming from diverse lived experiences, students with the same potential were not achieving the same academic outcomes. The learning outcomes were non-negotiable for the accreditation of the degree, the syllabus was set and the variables that I could redesign were the approach, format and content of actual lectures and tutorials, assignments, student support and ways of teaching. In the context of a previously racially segregated society, epistemological access to university degrees was a social justice issue. My immediate challenges were: How should I change my teaching to give all students equitable access to successful academic outcomes? Equally importantly, how could I equip future architects with a broader social understanding that would enable them to be relevant designers in a diverse society? These problems had to be solved in action, without the time to test them. We tried various, simultaneous, potential solutions, making it difficult to track which teaching changes produced which results. The other key question was whether first-year teaching interventions could create a foundation for ongoing success in future learning.

2 Introduction

The process of developing improved ways of teaching AD happened over the three years in which I led the first-year AD course. By the end of this time there was an ever-improving passrate in the course, and it also seemed that students who had done the revised course continued to perform successfully afterwards. This merited a formal study, which was accepted as a PhD proposal, and is described in detail in the dissertation ( Janse van Rensburg, 2015). The objective of the retrospective study, conducted after the changes in teaching had been completed, was to confirm whether these changes had indeed improved learning outcomes, whether these outcomes were sustained after completing the course, whether equal opportunities could be created by using this model, and to produce a theoretical model that could be applied in similar contexts.

AR principles had been followed from the beginning using a practice mode with a strong secondary emancipating mode (Kemmis & McTaggart, 2003), but the retrospective study required a shift to a knowledge-generating mode which presented additional methodological challenges. This paper also aims to describe the methods that were used to meet the criteria of dependability/reliability and process validity in a situation where student feedback and certain types of evidence were not consistently collected when the interventions started, and where continued outcomes could only be assessed after some time had elapsed. I believe that this can provide a precedent for similar studies where some data can only be collected after the actual interventions have been completed.

The contribution of this paper to the field of AR is therefore two-fold: It addresses the question of how equitable academic success can be facilitated in diverse learning communities, as well as how to produce valid AR in a context where one has to resort to retrospective data collection.

The action outcome of this study was that students’ overall academic outcomes improved, and there was a much more equitable distribution of marks in the class, as we developed strategically sequenced educational tasks to build a foundation of social and academic skills while teaching the formal syllabus. This pattern of improved performance continued as these students progressed into subsequent years of study. This research is relevant in many diverse learning communities where disparities in prior experience can polarise or enrich learning, particularly in the escalating context of global migration.

The retrospective AR study triangulated study data with quantitative data from other sources, producing clear correlations and confirming the validity of this method. It produced valuable longer-term insights and conclusions that could not have been obtained within the time frame of the interventions. This methodology is particularly appropriate for complex situations where the effects of different interventions can only become clear to participants over time.

AR literature tends to focus on defined problems which can be addressed and assessed within the scope of a clear, short-term study. AR is however an effective tool for addressing complex, “wicked” challenges, and these studies are seldom described in textbook educational AR. I believe it is important to expand the methodology to explore this field.

This paper is structured to cover the choice of methodology with its attendant challenges in section 3, the socio-political and pedagogic context in which the problems developed and had to be addressed in section 4, and moves to the action component of the study in section 5: the problems, causes, theoretical base for interventions and the interventions themselves. In section 6 it discusses the research component, outlining the study parameters, how data was collected, processed and analysed, and the model construction. In section 7 the outcomes of the interventions are discussed in relation to the different types of data analysis and research validity is confirmed. Section 8 concludes with the learning from this study, both on pedagogy and on AR methodology.

3 Methodology

3.1 Choice of methodology

This complex course design challenge was very similar to an architectural design project, which as an architect, I initially approached in the same way. Every architectural design is a research project. One has to collect data on the context, the limitations, the challenges, the ideal functioning of the proposed facility, the available resources and technologies and successful precedents. One is in a constant dialogue with the client to understand his/her needs and wants, and whether the design is addressing them. One has to understand the theory, and from an analysis of this complexity produce a single, integrated proposal that addresses all these issues. After testing the proposed solution against these criteria, one evaluates and adjusts it, until one reaches a satisfactory balance of outcomes and then resubmits this to the client for critique. It is a cyclical dialogue of problem definition, design, testing, observation and improvement, very equivalent to an Action Research (AR) process, but it has the advantage of being able to test multiple iterations before building the final one.

The closest recognised research methodology was Educational AR, as in addition to comprising cycles of improvement, I was applying “informed, committed and intentional educational action (McNiff, 2010, p. 16)” that problematised existing forms of educational practice (Newton & Burgess, 2008). It was participatory in that the process was constantly being informed by conversations with students and colleagues, all changes were immediately tested by students and their learning outcomes were assessed throughout this process, informing the process. It conformed to McNiff’s definitions, as it centred around my own learning in order to bring about social and educational change in my studio, hoping to establish more equitable education praxis (social justice) in a broader academic and professional context (McNiff, 2010). This resonates with Kemmis and McTaggart’s (2003) emphasis on aiming to bring about changes in people’s learning, actions, values, interactions and interpretations. The people in question were my co-teachers, my students and myself as we interacted to create a better integration between our roles and the course content.

Although the action component progressed with some momentum in this direction, it was only two years into the process that it became a research study, producing some challenges that had to be specifically addressed.

3.2 Methodological challenges and limitations

Retrospective student feedback: The only formal feedback from students on our teaching at the time of the course changes was the standard annual university course evaluation, in which the questions did not specifically address our interventions. Because ethical approval for the study had to be obtained before students could be formally questioned on their experience, this only happened after students had completed the course and had to rely on their recollections. As they were no longer primarily focused on this course and some were no longer on campus, there was a low response rate. The academic staff being questioned had to recollect specific observations on a repetitive timeline of similar events, and in all these situations the validity of memories had to be confirmed.

Complexity: Many educational AR projects set out to address and improve a single issue with  a single intervention, making it simple to design a study and attribute causality. This study contained many integrated variables, while both the course and the students had changed every year, making it more difficult to make comparisons and ascribe causality.

Limited documentation of student learning: The consistent data from the study period was a full and meticulous recorded of every student’s marks per assignment, but this is a compressed quantitative indicator of the combination of skills, understanding, learning, time constraints, personal challenges and available resources. The formative discussions en route to those marks had been oral and unrecorded. I had informal notes on particular students’ work and random projects that were documented for administrative reasons1, but it was a concern whether this data would be sufficiently representative to achieve saturation in a qualitative research context.

Undocumented observations and changes: AD is primarily taught in a studio format, where students present their work for critique and there is the opportunity for teacher interaction with every student. My role as the studio leader was to design the detailed course content and give the lectures, while I was assisted by three equally qualified part time co-teachers (“staff tutors”) in studio critiques. As a teaching team we gained considerable insight into every student’s engagement, effort and understanding and the common learning problems that emerged in the class, but our observations and discussions at the time were not documented. The rationale for changes to teaching formats, although carefully considered and agreed on as a team, had not been documented per se, and there was no consistent reflexive journaling.

1 Digital submissions were not yet the norm during the study period.

* * *

Would you like to continue reading? This article was published in issue 3-2022 of IJAR – International Journal of Action Research.

© Unsplash 2023 / image: Sam Balye