1. Introduction
The promotion of linguistic diversity within plurilingual learning processes, which simultaneously embodies linguistic richness and supports the preservation and nurturing of cultural roots, has been a practical demand for several decades, serving as an optimal condition for inclusive educational environments (
Buchs & Maradan, 2021;
Muscarà, 2010;
Targamadzė, 2018). This study, by reviewing the results of a previously conducted plurilingual experimental series, aims to demonstrate how the discourse of cooperative learning (
Davidson, 2021;
Jacobs et al., 2022;
Johnson & Johnson, 2021;
Gillies, 2007) provides an effective plurilingual learning environment for integrated international students. These practices seek to foster intercultural dialog and to promote acceptance and mutual understanding of each other’s cultures through language.
This inquiry is situated within the specific context of the internationalization efforts of higher education institutions. International students, those from outside the host country, are increasingly present at universities across Europe, including in Hungary. The central question is to what extent do they experience a welcoming and inclusive environment. Research in the field of student voice indicates that international students often feel a certain degree of isolation at most universities, as their academic programs are largely segregated, offering limited opportunities to participate in shared courses with the domestic majority (
Cruickshank et al., 2012;
Matthews & Dollinger, 2023;
Ravi et al., 2024;
Resch, 2023).
The aim of this paper is to highlight that plurilingual joint courses involving both international and domestic students, including those who do not speak each other’s languages, are not only feasible but have already been implemented in practice. These implementations have been especially successful when grounded in cooperative structures and principles aligned with the cooperative learning paradigm (
Arató, 2023b).
1.1. The Cooperative Structural Framework for Plurilingual Cooperative Classrooms
The methodological paradigm of the course, initially designed to be bilingual (English and Hungarian), was grounded in cooperative learning. The experimental process built on the paradigmatic structural framework of the cooperative learning discourse (
Arató, 2017,
2023b). This framework provides structural principles for restructuring the traditional learning process. These are practical principles educators can follow easily (
Table 1).
Following the cooperative paradigm is recommended in any case, as it leads to more effective learning than individual or competitive instructional models (
Johnson & Johnson, 2009,
2021). Within this framework, the core CL principles are contextualized specifically for plurilingual settings. Since these are structural principles, designed to create learning structures independent of specific content or disciplines (
Kagan, 1992,
2021), they are applicable across all subject areas and at any educational level.
Critiques related to multicultural or intercultural education, which highlight the need to move beyond both conservative and liberal multiculturalism, typically emerge in the context of Critical Multicultural Education (
Gorski, 2009;
Gorski & Parekh, 2020;
May, 1999;
Sleeter & McLaren, 1995;
Jenks et al., 2001). These critiques draw attention to the necessity of structural transformation, placing equity at the center of the restructuring process. This involves the structural guarantee that local cultural and social contexts are represented, even at the level of individual learners, within a critically friendly and transparent atmosphere in which equality and equal rights of cultures create a partnership-based environment.
What hinders, or even unconsciously makes discriminatory, the formation of such an environment are the inherited, learned, and internalized behaviors and social interaction patterns that, in both interpersonal and school contexts, reproduce stereotypical, discriminatory, and antidemocratic modes of interaction and cooperation (
Cohen, 1994;
Cohen & Lotan, 2014).
The paradigmatic principles of cooperative learning provide structural guarantees for dismantling hierarchical status relationships in the interest of joint learning and the successful development of all students.
Arató (
2023b) refers to this critical restructuring as deconstruction, since it dismantles the ingrained structures that exclude some, while simultaneously constructing and offering cooperative structures to enable more effective learning for all. In this way, hierarchical fixations are gradually eroded, and learners become increasingly autonomous, mobilizing their own cultural-linguistic and social dimensions to enhance their learning outcomes.
In summary, the framework presented below (
Table 1) offers structural guarantees that support the practical implementation of plurilingual education across a wide range of educational contexts.
The author sought to follow the above structural principles both in the single-language control groups and in the multilingual experimental groups.
While this is not the place to elaborate on the basic principles of CL, it is worth reflecting on the necessary technical conditions from the aspect of plurilingualism. These foundational issues have been explored in detail by
Arató (
2023b), particularly in relation to online education (
Arató, 2023a).
1.2. Technical Conditions for Cooperative and Plurilingual Learning
During the COVID-19 pandemic, no Erasmus students arrived at our university. However, one of our English-language training programs was launched, and according to university policy, Erasmus students may be integrated into the international (English-language) programs, but not into the domestic, Hungarian-language courses. Nevertheless, as a result of the two-year period of emergency remote teaching, it became possible to map out the optimal conditions required for the digital infrastructure of an effective cooperative learning environment (
Arató, 2023a).
With the widespread accessibility of LLM systems, new opportunities have emerged for implementing more inclusive classroom practices. This is especially true in plurilingual language environments. In live language communication, it is now possible, via digital environments or tools, to ensure immediate access to content in any language used in the classroom. This means that real-time utterances can be instantly translated into all the languages spoken in the class, whether in written or audio format. Thus, multilingualism is no longer limited to written materials but extends to spoken and chat-based interactions as well.
Table 2 summarizes the framework for an optimal online classroom learning environment from the perspective of plurilingual education.
A future study could examine to what extent, alongside the use of the Cooperative Structural Framework, the above technical framework contributes to enhancing the efficiency and effectiveness of plurilingual education.
2. Materials and Methods
At the time of the experiment (between 2013 and 2019), the author resumed teaching at the University of Pécs, where international teacher students arriving through the Erasmus program of the European Union were also able to attend his classes. While most instructors offered separate English-language courses for their small groups of 3–5 Erasmus students, the author decided to integrate them into his regular, Hungarian-language teacher training courses, as he had announced identical course topics in both programs (Learning-centered communication in the classroom: Cooperative Learning). This most popular course proved to be the most suitable setting for the experiment, as continuous control groups were available, and he also had a process-assessment test that he had been using for years as a feedback tool to measure the depth of students’ understanding of the course content.
2.1. Plurilingual Heterogeneity on Micro-Group Level
Based on the two instructional languages (English and Hungarian), four linguistic categories had been established according to the equal access and participation basic principle of CL (
Arató, 2017,
2023a,
2023b):
A: students who spoke only English,
AM: students proficient in both English and Hungarian,
MA: students fluent in Hungarian but beginners in English, and
M: students who spoke only Hungarian.
The author formed cooperative micro-groups that, wherever possible, included at least one participant from each category, but at minimum one AM student was always present if an A student was part of the group.
The course itself focused on cooperative learning and was designed for teacher education students. It followed Kolb’s experiential learning cycle (
D. A. Kolb, 1984;
A. Y. Kolb & Kolb, 2013). The author built the experiential learning cycle on the four phases of Kolb’s original model: Concrete Experience, Reflective Observation, Abstract Conceptualization., Active Experimentation. Beginning with an experience-based phase in which students were introduced to cooperative structures and communication techniques in an engaging, low-stress manner through activities that also allowed them to get to know each other. The next phase of the cycle derived theoretical principles, specifically, the four most well-known core principles of the cooperative learning paradigm, from these shared experiences. This stage employed an Expert Jigsaw structure to guide collaborative learning processes (
Aronson et al., 1978;
Aronson, 2021), helping students identify and internalize concepts and principles through reflection on their prior activities.
At the end of the second phase, the author administered a test to monitor the level of conceptual understanding and provided feedback to clarify and consolidate students’ grasp of the core principles and their practical applications. This test served as the basis for the statistical analysis of results for this study.
After reviewing the test results together with the participants, the course moved into the third phase of the Kolb cycle, in which students designed 25 min cooperative learning sessions in their micro-groups for their peers. In the fourth phase, they implemented these sessions, receiving feedback not only from the instructor but also from their peers, acting as developers, participants, and observers. Giving peer feedback from these three perspectives was also an integral part of their learning process.
The course concluded with the submission of a portfolio in which students designed a similar 25 min cooperative learning session for a selected age group, based on a topic from their own teaching subject.
2.2. Statistical Analyses of Effectiveness of the Cooperative Structural Framework, Confirmed Across Diverse Target Groups
The experiment was conducted over six semesters between 2013 and 2019. Due to privacy considerations, the specific semesters are not disclosed, as the guest students could be easily identified. Each experimental group consisted of maximum 16 participants (n = 91), including 1 to 4 Erasmus teacher students per course. The Hungarian-language control group consistently included maximum of 20 teacher students each semester (101).
To analyze the data, a series of statistical tests have been performed to examine potential differences between the groups. In the first case, the analysis examined whether there was a significant difference between the Experimental groups and the Control groups. For this purpose, a Two-Sample t-Test for Unequal Variances was applied. In the next step, the performance of the individual language groups was examined using the ANOVA method, and then the Tukey HSD post hoc test was applied to determine between which groups significant differences in performance could be observed.
3. Results
The statistical analyses led to three key findings. First, there was no significant difference between the control groups and the experimental groups, whether comparing groups from the same academic year or the results across the six examined cohorts. The second finding was that within the experimental group, there were significant differences among students classified into different language groups, even though the performance of each group was high. The third finding was that this difference was primarily observed among students using the language of the host country.
3.1. Plurilingual Education Does Not Reduce Performance
The author performed a two-sample
t-test for unequal variances to compare two specific groups’ means (
Table 3).
t-statistic: 0.841;
Degrees of Freedom (df): 192;
p-value (one-tailed): 0.2007;
Critical t-value (one-tailed): 1.6528;
p-value (two-tailed): 0.4014;
Critical t-value (two-tailed): 1.9724.
The t-statistic was 0.841, which is less than the critical t-value of 1.9724 for a two-tailed test at a 0.05 significance level. The corresponding p-value of 0.4014 also exceeds the typical significance threshold of 0.05, indicating that there is no significant difference between the means of the two groups (Experimental and Control).
The second part of the analysis examined whether there were substantial differences among students in the experimental groups, based on their classification according to linguistic categories. The analysis involved a one-way ANOVA analysis. Both tests were essential in assessing the central tendencies of the groups and determining whether observed differences were statistically significant.
3.2. Differences May Occur Among Students with Different Linguistic Backgrounds
The author first conducted a one-way analysis of variance (ANOVA) to determine if there were significant differences in means across the four groups (A, AM, MA, and M). The results of the ANOVA indicated a significant difference in group means, as the p-value was found to be 0.02396, which is less than the significance threshold of 0.05. This suggests that there is a statistically significant difference between the groups.
Further, the F-statistic was 3.45, which exceeds the critical F-value of 2.81. Given these values, we can reject the null hypothesis, confirming that at least one pair of groups differ significantly from each other.
3.3. Different Outcomes, but Not to the Detriment of International Students
To further explore the group differences, the author applied a Tukey HSD post hoc test. This test revealed (
Table 4) a significant difference between groups AM and M (
p = 0.015), while no other pairwise comparisons reached statistical significance. This finding highlight that the AM group differs notably from the M group, but other groups do not show significant variations in their means.
3.4. Limitations of Statistical Analysis
The data was developed not for statistical outcome calculation, but for tracing individual responsibility. With the help of the test, the experiment aims to demonstrate that using the micro-group structure, it is possible to precisely identify how the distributed responsibilities in the course topic (the interrelations of cooperative principles and structures) can be made transparent, even within an individual test question set.
The results of the 14-question test, expressed as percentages, magnify the differences, while the results range from 0 to 4 points of error, meaning that grades of 5 and 4 are typical, based on the test. The test is not overly easy, as only one 100% result was achieved out of the 12 groups (tester n = 192). The test evaluates the cognitive domains of understanding, application, analysis, and synthesis from Bloom’s taxonomy in the context of the basic principles of CL and its application in cooperatively structured learning processes.
In fact, the test serves as a process-evaluation tool, which allows participants to assess how well they can illustrate their understanding with examples, apply principles to learning structures, analyze structures in terms of principles, and synthesize the lessons of their analyses.
4. Discussion
In the discussion of the results, the author points out that the test used for the statistical analyses was not part of a course-end assessment, but rather a formative evaluation tool that did not count toward the final semester results. This made it more likely to map the participants’ current knowledge, acquired specifically during the course, as no additional home preparation time was involved. The author also highlights that although students achieved high scores on the test, indicating successful knowledge acquisition, differences in performance could still be identified. Through joint analysis of the processes, three typical situations influencing performance emerged. In the fourth part of the discussion, the author describes how the experimental practice, initially designed to be bilingual, organically transformed into a plurilingual practice, open to the linguistic resources of the participants, both for the instructor and for the students.
4.1. Testing the Comprehension in the Middle of a Learning Process
The group progressed through the above-mentioned cognitive domains in a multiple stages process. Through a cooperatively structured process, since the pedagogical aim was to prove that they understood a completely new paradigm (cooperative learning) with 78–100% accuracy, for which the lecturer did not even need to give a 5 min lecture.
Building upon this level of understanding, the task of planning was introduced according to the Kolb-cycle, further refining students’ comprehension. The test included in the statistical analysis was administered after the second stage of the Kolb cycle, when participants captured their experientially lived processes using the concepts of cooperative principles and structures. The test was designed as a tool embedded in the block-taught course, meaning that no prior home preparation preceded the testing. This implies that it measured the students’ actual knowledge acquisition, specifically to ensure that they were adequately prepared to move on to the next stage of the learning process: the phase of practicing the design of their own teaching module based on cooperative principles and structures.
Planning is also a learning process, meaning that they do not know everything perfectly in advance and then plan, but rather, they learn through the planning process itself. During the planning phase, several cooperative learning principles and structures are considered from a practical standpoint, thus sharpening their understanding.
Finally, in the fourth phase of the Kolb-cycle, students highlighted implementation as the most significant learning phase in their anonymous evaluations. The trial and reflection stage are preceded by the engagement and group-building phases, followed by conceptual learning and connection with the presented experiences, and finally, planning based on the knowledge gained from the previous two stages.
The results of the study should be re-examined in further controlled experimental environments, where a crucial element is the conscious and practiced application of cooperative structures and linguistic heterogeneity.
4.2. Slightly but Significatn Differences in High-Level Achievement
The 14-question test is, in fact, a process-evaluation tool and was not designed for grading purposes. However, the statistical analysis, despite showing only a few points of difference in the tests, was able to identify a significant trend. There was no significant difference between the bilingual and monolingual groups, meaning that bilingualism did not negatively affect performance. It seems that everyone performs at a 4 or 5 grade level when graded, regardless of whether they speak only English, both languages, and only Hungarian.
This indicates that the performance of international students integrated through English language instruction did not differ significantly from the average performance of other students (their performance ranges from 78% to 93%, with the overall group variance lying between 71% and 100%). The AM students, who worked most in both languages, performed similarly to the foreign students, with their performance falling in the same 78–93% range. The performance of the other groups (like MA or M students) did not differ significantly from one another either. A significant difference was noted between the Hungarian-only speakers (M) and those who used both English and Hungarian (AM). This suggests that bilingual participation was more advantageous, even with a slight difference (1–2 points in the test), in terms of success.
In the further development of the course, the groups placed greater emphasis on the role of the Encourager, assigning the AM students in each group the responsibility to ensure that all information, including personal conversations within the group, was communicated in both languages.
4.3. Performance Differences in Plurilingual Micro-Groups
The cooperative micro-group structure made it possible to track the level of individual accountability down to each participant. The test was specifically designed for this purpose. Based on the test, the responsibilities associated with each group member’s role in the Jigsaw structure can be clearly identified, while five of the test questions also make the success of areas falling under the facilitator-instructor’s responsibility transparent.
Over the course of the six experimental semesters, several typical patterns were observed:
The “excluded Hungarian speaker”: In a few groups, the M (monolingual Hungarian-speaking) student achieved the lowest test score. While this did not reflect a significant difference compared to the entire cohort, statistical analysis revealed a notable trend when compared specifically to bilingual participants.
The “distracted interpreter”: In some groups, the AM (bilingual English and Hungarian) student scored lowest on the test. This was typically because they were so focused on supporting others’ understanding that they may have had less time to organize their own thoughts. It is important to note that these students still only lagged by one or two points out of 14, which corresponds to a solid “B” grade level.
The “team that clicked”: Some teams had multiple members scoring in the lower range of the overall data set (around 78%, though not at the lowest end of 71%). In these groups, many off-topic and social interactions were observed, even after the half-day dedicated to becoming acquainted. In such cases, students were more focused on bonding and enjoying each other’s company, an outcome that, from the instructor’s perspective, was also a valuable pedagogical goal.
After analyzing the responsibility structures revealed by the test results, the groups themselves made informed decisions. For example, reinforcing the responsibility of AM students to ensure equal participation. In cases where the group’s overall performance was lower compared to others, strengthening the role of the Tracer, whose task is to keep the group focused on the objective. In the case of the distracted interpreter, some groups decided to slightly increase their own participation in group communication at times, allowing the AM student to step back and focus on listening. Additionally, the MA (Hungarian-dominant bilingual) students were encouraged to take on more of the translation duties, sharing the cognitive load more evenly.
4.4. From Bilingual to Plurilingual Education: The Benefits of the Cooperative Learning Framework
By the second iteration of the course (in experimental semester 2), the bilingual setting evolved into a plurilingual one, as four Erasmus students from the same country and with the same first language joined the group. This allowed us to switch to a three-language model, at least to the extent that all written materials in each group had to be produced in three languages: the Erasmus students’ first language, English, and Hungarian. From that point on, it became natural to incorporate additional languages into later groups. The highest number of actively used languages in a single course was nine.
Later on, the author intentionally assembled groups so that each one included at least three languages shared by pairs of group members. This always included Hungarian and English as a baseline. Such a plurilingual linguistic performance proved particularly effective for fostering connection and building teams, and it also enabled the local identification and contextualization of a scientific-practical field from the perspective of the languages and cultures of the participants. This resulted in a kind of multi-perspective resource richness that not only encouraged critical engagement with mainstream trends but also facilitated critical insight into participants’ own more familiar contexts.
The most noteworthy development from a pedagogical and intercultural perspective was that, by the time international groups progressed to designing their own practice-based activities, it had become a natural and self-initiated standard to produce session materials not only in English and Hungarian, but also in the participants’ respective first languages. This multilingual approach ensured that all group members, regardless of linguistic background, had access to key content in at least two languages, thus promoting equitable participation and deepening mutual understanding within a plurilingual learning environment.
5. Conclusions
The results of the study conducted over six semesters clearly demonstrate that university students learning in plurilingual environments and processes aligned with the Cooperative Structural Framework perform at a comparable quality level to those in monolingual groups. It can be stated that a plurilingual educational environment does not disadvantage either local or integrated international students, as both are able to achieve the same level of performance in both versions. Based on the findings, it may also be concluded that when significant differences do appear, even if they do not imply a substantial decline in learning outcomes, they are primarily observed among the performances of local students. In other words, within integrated, plurilingual, cooperatively structured processes, the performance of international students will certainly not be lower than that of local students. Performance differences among local students can be addressed through formative assessment, provided that, alongside efforts aimed at fostering inclusion, equal emphasis is given to applying the cooperative principle that ensures equal access and participation for monolingual learners.
The most important outcome of the experiment does not lie in the statistical data but in the change in attitude that makes it natural for pre-service teachers to consider their students’ preferred languages and to integrate them into the teaching processes, thereby also bringing into dialog the cultural contexts represented by those languages.
To support the intentional development of plurilingual learning environments, the author presented elements of a framework that can be applied not only in higher education but also in public education, with appropriate adaptations to the learners’ developmental stage. The Cooperative Structural Framework serves as a structural guide by offering the principles of the cooperative learning paradigm for organizing plurilingual learning processes.
Today, the technical conditions for implementing a plurilingual approach are already in place. Thanks to the paradigmatically new structures of cooperative learning, the structurally guaranteed inclusion of learners’ own languages becomes the key element. In a plurilingual environment, through language, and by consciously following a cooperative strategy that seeks to harness the richness of linguistic diversity, the specific local, familial, social, and linguistic-cultural contexts of each learner are smoothly and, eventually, naturally integrated into shared learning processes.