It is widely accepted within the academic community that a fundamental step towards the realization of an equitable and fair education system entails the implementation of teaching strategies that take the diverse range of learners’ abilities and needs into account (Lindner & Schwab, 2020; Smale-Jacobse et al., 2019).

The concept of differentiated instruction (DI) is frequently discussed as a potential strategy for creating inclusive classrooms and promoting equity in education (Gheyssens et al., 2020; Lindner & Schwab, 2020; Maia & Freire, 2023). While the implementation of differentiated instruction is widely recognized as a significant challenge for teachers (Lindner & Schwab, 2020; Smale-Jacobse et al., 2019), digital tools are discussed as potential resources to support teachers in dealing with pupils’ heterogeneity (Dumont & Ready, 2023).

Yet relatively little is known about students’ perceptions of these educational settings (Lindner & Schwab, 2020; Scarparolo & MacKinnon, 2024). Moreover, research on the perspectives of pupils typically includes students from secondary schools (e.g. House, 2020). The scarce existing evidence indicates that pupils in settings of differentiated instruction develop a sense of agency and control over their learning (House, 2020; Netcoh, 2017). This is also consistent with the perception of self-regulation being crucial in these learning contexts – for example, represented in the elements of personal responsibility and motivation (Netcoh, 2017) or self-efficacy and self-control (House, 2020).

In order to address the aforementioned gap this contribution aims to identify what is significant from primary school students’ perspective regarding differentiated instruction supported by digital tools.

We present preliminary results of a project that combines ethnographic and participatory research methods. The project is being implemented in six classrooms, with five to six pupils participating as co-researchers in each class representing a diverse group of pupils (e.g. regarding their academic performance, gender, socio-economic background, and special educational needs).

Overall, the project employs a combination of methodological approaches. Firstly, academic researchers engage in classroom participant observation and document their findings using field notes (Beuchling, 2015). Secondly, pupils engage in collaborative research with academic researchers as co-researchers, with the aim of documenting their perception of learning in settings of differentiated instruction supported by digital tools (Von Unger, 2014; Zahnd & Oberholzer, 2022, 2024). Prior to assuming the role of research participants, the pupils engage in workshops designed to familiarize them with the specific research questions and methodologies that will be employed. All pupils are provided with research diaries and smartphones, which facilitate the utilization of photovoice as a participatory method (Wöhrer et al., 2017). The collected data is discussed in mutual meetings during which the pupils collaborate with the academic researchers to analyze and systematize their findings.

The preliminary findings contribute to the existing knowledge concerning primary students’ perception of differentiated instruction and give some insight on the perceived role of digital tools in this context. In line with the findings at the secondary school level, pupils have identified strategies for self-regulation, such as the ability to decide which tasks they wish to work on or to plan their own learning, as being especially beneficial.

Regarding digital tools, pupils have indicated that the used digital media is sometimes perceived as confusing and boring. Nevertheless, e.g. educational videos are sometimes also rated positively by pupils. Yet, in comparison to this form of “digital” learning support by videos, students have rated learning support from teachers and classmates as even more benefical. In their experience, teachers can answer questions immediately and classmates can provide support in a more immediate and personalised manner.

Adaptivity enables teachers to address the diverse learning needs of their students. This concept underscores the importance of tailoring teaching methods to accommodate varied learning styles, abilities, and interests within a classroom. However, research indicates a gap between teachers' theoretical understanding of differentiation and its practical implementation in everyday teaching practices (Gheyssens et al., 2022; Smale-Jacobse et al., 2019).

Technological tools have emerged as potential resources that can support teachers in implementing differentiation effectively. While these tools create novel opportunities to meet individual learning needs, it is essential to recognize that teachers, like their students, form a heterogeneous group. They possess varying attitudes, experiences, and skill levels regarding both digitization and differentiation. These differences can greatly affect how – and whether – teachers choose to integrate digital tools and differentiated teaching methods into their classrooms (Ostermann et al., 2022; Maia & Freire, 2023).

Moreover, if digital tools begin to take on essential tasks within the teaching profession – such as diagnosing learning outcomes to differentiate between students – there may be profound implications for the professional role of teachers. This shift may affect key aspects of the profession, including the teachers’ autonomy in decision-making, professional expertise, and accountability. It may also reinforce current trends, such as the growing expectation for teachers to document student assessments (Selwyn et al., 2017).

Against this backdrop, this contribution aims to analyze teachers' perceptions of their professional roles to better understand how these perceptions could shape their differentiated teaching practices. The contribution adds to the broader discussion on the role of professional expertise in light of recent technological advancements (Susskind & Susskind, 2017) and offers insights into the conditions under which teachers are willing to integrate digital tools into their professional practice (Petko, 2012).

To address the wide array of differentiated instruction practices currently observable in Swiss primary schools, the project examines teachers using different types of learning material with varying degrees of adaptivity. It focuses on math teachers in primary education in grades 4 to 6 in the German-speaking part of Switzerland. The project employs qualitative methods, specifically focus groups and expert interviews. The focus groups are designed to foster collective discussion among teachers, illuminating the formation of professional judgments, while the expert interviews provide deeper insights into individual teachers' reasoning and experiences regarding differentiation.

In total six focus groups with approximately 30-40 teachers will be conducted, alongside individual interviews with all participating teachers. Together, these focus groups and expert interviews aim to provide a more nuanced understanding of differentiation in Swiss primary schools, by systematically comparing different schools, tools, and teachers. The qualitative data analysis follows a Grounded theory approach (Corbin & Strauss, 2008), emphasizing a constant comparison across cases and educational resources.

Adapting learning materials entails shifting the focus from a one-size-fits-all approach to one that recognizes and encourages the diversity of learners. Focusing on pupils' needs in the scientific discourse is a starting point for educational equity and fair educational systems (Lindner & Schwab, 2020).

Prior knowledge can be seen as a foundation for adaptivity, as it is crucial in students' ability to process new information and tackle learning tasks (e.g., Hattie, 2009). One way to address differences in prior knowledge is through pre-training. Xie et al. (2017) found that activating prior knowledge is an appropriate pre-training strategy for learners (i.e. explaining concepts priorly) with limited prior knowledge. However, pre-training interventions should be tailored to learner's characteristics to avoid an expertise reversal effect (Sweller et al, 2011).

Another way is designing materials accordingly (i.e. instructional support) which should also be tailored to the learner's expertise level. As learners progress and gain more knowledge, the scaffolding should be gradually faded.

Our study investigates the interaction between learners' initial prior knowledge states and the design of suitable learning interventions. Specifically, it examines how tailoring learning materials that align with individual cognitive characteristics can facilitate learning performance addressing a crucial question in personalized learning: Where should instructional adaptation begin to support diverse learning needs?

The study involved 139 Swiss primary school students from both inclusive and regular education settings (63 female, 72 male, 4 not specified, M_age = 10.51 years). Using a 2x2 between-subjects design in mathematics, two factors were manipulated. The first was prior knowledge induction: in one condition, students received contextual information to activate prior knowledge (video duration: 6.52 minutes), while in the other, they learned about the origin and standardization of the meter without prior knowledge induction (video duration: 7.02 minutes). The second factor was instructional support, offered at two levels: high (step-by-step guidance) and low (minimal guidance), with video durations of approximately 4.5 minutes.

Four instructional videos tailored to these conditions covered the topic of “Converting Units of Measure.” Each video was delivered individually to participants via tablets for consistency. Initial prior knowledge was assessed through four single-choice items with varying difficulty levels. After the videos, a learning test (10 items, α = .68) was administered, and participants rated the utility of the initial video in solving the test.

Results showed that prior knowledge induction significantly improved learning outcomes (p < .001, R² = .130). Students who received prior knowledge induction (M = 7.38, SD = 2.01) outperformed those who did not (M = 5.76, SD = 2.50). High instructional support (M = 6.80, SD = 2.33) also led to better results compared to low support (M = 6.34, SD = 2.47).

Crucially, the effects of prior knowledge induction and instructional support were moderated by learners’ initial knowledge levels. For learners with high prior knowledge, the positive effects of these interventions diminished or turned neutral to negative (p < .001, R²_change = .19 for prior knowledge induction; p = .005, R²_change = .09 for instructional support). A three-way interaction revealed that the combined effects of prior knowledge, instructional support, and learning outcomes are highly dependent on the design of the learning materials.

These findings emphasize the importance of tailoring instructional support and pre-training strategies to learners’ initial knowledge levels. While instructional support generally improves learning outcomes, its effectiveness depends on aligning it with learners’ cognitive characteristics. The moderation effects highlight that interventions must be carefully calibrated to prevent overloading advanced learners while supporting those with less prior knowledge.