Conference Program

This study investigates the role of visual organization in digital instructional materials within lower secondary education, focusing on its impact on comprehension, immediate and delayed retention, and perceived cognitive load. The research is grounded in Cognitive Load Theory (Sweller, 2010), Multimedia Learning theory (Mayer, 2002), and studies on graphicacy, assuming that spatial and hierarchical arrangement of visual elements may facilitate or hinder the integration of verbal and spatial representations.

The study examines whether specific design choices—namely text–image proximity and explicit visual hierarchy—affect cognitive processes involved in learning an introductory mathematical concept: the Cartesian plane.

A between-group experimental design was implemented with a sample of 52 students (age range: 11–12 years) from three first-year lower secondary classes. Each class attended a lesson on Cartesian axes delivered through one of three graphic templates derived from the combination of two independent variables: (a) text–image proximity and (b) presence of explicit visual hierarchy. Instructional content, teaching time, and delivery mode were held constant across conditions.

Dependent variables included prior knowledge (baseline test), immediate retention, comprehension, perceived cognitive load, and delayed retention one week later. Cognitive load was measured using a questionnaire based on the validated scale developed by Klepsch, Schmitz, and Seufert (2017), allowing differentiation among intrinsic, extraneous, and germane load components.

Results show statistically significant differences between experimental conditions only in the disaggregated analysis of cognitive load components. In particular, extraneous cognitive load varied systematically across templates, indicating that graphic design primarily influences cognitive processing rather than final performance outcomes. No statistically significant differences emerged in comprehension or retention tests.

The findings show that visual organization significantly affects extraneous cognitive load, even when measurable differences in learning outcomes do not emerge. While recent systematic reviews have questioned whether the spatial contiguity effect can be explained by reductions in extraneous load (Schroeder & Cenkci, 2020), the present study suggests a more articulated relationship between instructional design, cognitive load, and performance. Variations in extraneous load did not translate into immediate learning gains, yet they reconfigured the cognitive conditions under which spatial-quantitative reasoning unfolded.

This dissociation indicates that additional variables intervene in the construction of knowledge and, in line with representational models emphasizing the coordination of depictive and descriptive processing (Schnotz & Kürschner, 2007), points to a relationship between cognitive load and learning outcomes that is neither linear nor unidirectional. Current models linking instructional design, cognitive load, and performance may therefore oversimplify the dynamics of visuo-quantitative learning.

By showing that the arrangement of text and visual elements influences cognitive processing in the introduction of the Cartesian plane, the study contributes to the discussion on graphicacy as a foundational component of scientific literacy. In this perspective, the design of graphical materials becomes an epistemic and pedagogical issue: reducing unnecessary cognitive burden may support more equitable participation in visuo-quantitative practices central to mathematics and science education. The study therefore positions visual design as a potential mediator of inclusive access to scientific representations, contributing to the broader framework of democratic science education.

Mathematical knowledge is often associated with a certain dogmatic nature, detached from the world of human meanings and everyday contexts. This can result in a situation where a teacher presents mathematical theory in isolation from the individual needs and cognitive processes of individual students or groups (Zhou 2022).

The philosophy of mathematical practice perspective adopted in our research (Löwe and Müller 2010; Giardino 2018; Carter 2019; Van Kerkhove 2009) does not confer on mathematics the status of exceptional knowledge. Mathematics can be perceived not only as an abstract discipline, but also as a profoundly social phenomenon, as a construct born of interpersonal interactions, a language and tool embedded in social practices. One of the important aspects of this approach is research on the epistemology of visual thinking in cognition, reasoning, and even mathematical proof.

In (Tytko et al. 2023) mathematical practice is described from a psychological and social perspective. There is introduced the notion of Critical Concept Kinds. This notion is strongly related to the presence of visual representations in cognitive processes.

In this presentation, we will focus on a deeper analysis of the proposal of fostering Critical Concept Kinds in math and mathematical education (based on Critical Math Kinds defined by Mangraviti 2024, and Critical Gender Kinds defined by Dembroff 2019).

Df.: „a critical concept kind is a way of working with a concept which destabilizes mainstream ways of working with the “same” concept. Critical individual mental practices are those practices that generate critical concept kinds” (Tytko et al. 2023).

This proposal highlights the gap between the idealistic understanding of mathematical language and the process of creating, applying and learning mathematics, with particular emphasis on the processes and conditions that influence these processes. Especially multimodal aspects. Since the mental mechanisms and relationships observed in mathematical and math education practices can also be observed in other types of human practices, because of their psycho-experiental background, mathematics can also be a tool – especially in the educational context – for analyzing and shaping society.

It is fit to the need in the education model, for changing an absolutist epistemology to one based on critical thinking skills and team/project collaboration (Anna Nitecka-Walerych, 2021). There are two advantages to this:

1. More effective problem-solving at a technical level, based on practice, knowledge, and skills.

2. Contribution to a more inclusive (less harmful) society – complementing Bettcher’s (2014) proposal regarding the existence of multiple worlds of meaning, in which certain concepts have different meanings. Bettcher’s proposal primarily concerns practices that ultimately lead to a change in the meaning of concepts. Our proposal begins by fostering awareness and the need for critical thinking about mathematical concepts, aiming to foster the ability to reconsider dominant concept kinds in the social space, which may be harmful.

The gender gap in STEM disciplines is rooted in childhood, through the channels of visual culture. Gender stereotypes emerge as early as age 2 (Serbin et al., 2001) and by age 5 are consolidated into associations that children apply to themselves and others (Martin & Ruble, 2004). During this period of maximum brain plasticity (Santrock & Rolle, 2017), images are the primary means through which children construct their understanding of the world (Tizard & Hughes, 2002), conveying meanings without the need for verbal mediation (Berger, 2008; Gibson, 2004). When illustrations connote the scientific imaginary as masculine, they construct barriers (Biemmi, 2017) that girls internalize as limits, with documented effects on self-efficacy and future aspirations (Lopez, 2012), contributing to educational segregation (Bian et al., 2017) and the underrepresentation of women in the most remunerative sectors of the economy (Marone & Buccini, 2022). Data integrating INVALSI assessments and university enrolment records confirm that girls, despite showing high levels of competence, are less likely to enrol in STEM programs. The divergence in choices is therefore not explained by differences in academic performance (Falzetti, 2024). Empirical evidence from an experimental study supports this position [blinded for review]. The study involved 82 children aged 4–5 years (M = 41, F = 41; mean age = 4.73 years), exposed to graphically neutralized illustrations — through the elimination of gender markers such as clothing, hairstyles, and sexual traits; the reduction of paedomorphic features (Araujo et al., 2022; Karniol, 2011); and the adoption of a neutral colour palette through the desaturation of gender-attributable colours. Results show statistically significant differences such as reduced stereotype conformity (U = 234, p < .001), lower automatic recognition of activities as gender-associated (U = 292, p < .001), and greater hesitation in choice (U = 461, p < .001). Furthermore, qualitative observations support these findings, showing greater hesitation and verbalization of uncertainty — signals of a shift from automatic to more deliberative categorization. These results provide empirical evidence that gender biases are culturally induced and modifiable through visual design. This is relevant for STEM materials, as pathways toward these domains depend not only on performance, but on motivational variables such as self-efficacy and disciplinary identity, influenced by gender as early as preschool age (Hart & Ganley, 2019; Vincent-Ruz et al., 2018). If graphic neutralization suspends stereotype-driven automatisms, STEM materials designed with visual neutrality criteria can represent a concrete intervention. The criteria were applied to the redesign of two STEM books for the preschool age group, demonstrating the transferability of the method to real editorial contexts. Systematically adopting these criteria — by publishers, illustrators, and institutions — means intervening on one of the earliest mechanisms through which gender inequalities are reproduced. Visual representations mediate identity construction from early childhood, shaping attitudes toward scientific knowledge. Rethinking the visual design of STEM materials as an epistemic and political act is a necessary condition for a democratic scientific education, in which knowledge becomes a shared civic competence, accessible to all regardless of gender.

Visual representations in educational materials are not neutral tools: they organize, hierarchize, and filter knowledge through choices of typography, color, iconography, and layout that are simultaneously cognitive, cultural, and political decisions. Yet the literature on visual design in education has focused predominantly on two dimensions — graphicacy, understood as the reader’s competence in interpreting visual signs (Balchin & Coleman, 1965; Friel et al., 2001), and inclusive design, oriented toward reducing perceptual barriers (CAST, 2024; Rose & Meyer, 2002) — without systematically addressing the dimension of epistemic justice in representation.

This contribution introduces the theoretical and operational framework of Graphic Design Equity (GDE), which analyzes visual design choices in terms of the equitable distribution of the conditions for understanding a graphic artifact and the reader's ability to recognize themselves in that representation. The term “graphic” is used in its semiotic sense: it refers to any visual representation choice that mediates access to knowledge, regardless of the production context. GDE does not replace existing concepts but radicalizes them toward a third level not yet formally articulated: beyond perceptual accessibility and cognitive usability, toward epistemic and symbolic justice — that is, whether everyone can understand and recognize themselves in what is made visible.

The framework is structured around three interconnected dimensions: a design dimension, concerning technical choices of typography, color, layout, and visual hierarchy; a social dimension, addressing the symbolic and cultural inclusion or exclusion of identities and perspectives; and an educational-scientific dimension, aimed at educating teachers and designers as ethical agents of representation.

Studies on the gender gap in STEM show that differences in spatial reasoning emerge and grow with age rather than being present from birth (Lauer et al., 2019), and that women with cognitive profiles comparable to men's remain underrepresented in STEM fields, suggesting the role of social and structural barriers rather than cognitive ones (Sokolowski et al., 2026). The analysis of 1,255 textbooks across 34 countries documented the pervasiveness of visual bias in school materials (Crawfurd et al., 2024), while experimental research has demonstrated that counter-stereotypical images significantly improve female students’ performance in scientific domains (Good et al., 2010). D'Ignazio and Klein (2020) has shown how every visualization embeds a perspective and a hierarchy of values that structure the possibilities of recognition available to the learning subject.

GDE is proposed as an analytical tool to evaluate existing materials, a design framework to produce more equitable visualizations, and an ethical and educational device for the training of designers. In alignment with Goal 4 of the United Nations 2030 Agenda (2015), which aims to ensure inclusive and equitable quality education, GDE intervenes at the graphic level as a semiotic infrastructure that shapes access to understanding, translating the principles of educational equity into operational design criteria.

The proposed study articulates between two major issues in current educational discourse.

On the one hand it is the growing interest in Computational Thinking (CT) defined as “solving problems, designing systems and understanding of human behavior, using the fundamental concepts of computer science” (Wing, 2006). In particular, this study will focus on the CT subcomponent related to algorithms production and usage and its cross-disciplinary relevance into the educational domain (Popat & Starkey, 2019).

On the other hand this study relates to “Graphicacy”, namely the ability to produce and interpret graphics, as an underpinning skill for every educational background (Balchin & Coleman, 1966).

The relevance of both issues is confirmed by the attention paid by the internationally applied frameworks for K-12 computer science education (Association for Computing Machinery et al., 2016) and science education assessment (OECD, 2023).

This study will focus on algorithms defined as a finite number of steps to solve a problem (Louridas, 2020), their forms of representation (Merrill, 1980) and the flowchart in particular.

A narrative review of comparative studies will be included to assess the learning advantage granted by the algorithm's different forms of representation: flowchart, pseudocode, linear text, block-based tools (Giordano & Maiorana, 2015; Scanlan, 1989; Xinogalos, 2013).

This study will later review the evidence provided by recent neuroimaging studies in perception of the cartesian graphical space (Ciccione et al., 2023; Ciccione & Dehaene, 2026). The involvement of the visual system’s dorsal stream, known as the “how” pathway, and devoted to self spatial positioning and distance perception for purpose-oriented interactions is particularly relevant for this study, as well as the demonstrated long established graphicacy of the cartesian diagram (Perondi et al., 2025).

Calza & Perondi (2025) found that Reference Frame Systems have positive impact on understanding process compared to Rules of Composition System, e.g. flowchart. It could be hypothesized that a well established graphicacy of flowchart diagrams, leveraging on the discrete nature of the flowchart’s blocks, their function-based shape, their visuo-functional connection, may lead the flowchart to be reframed from mere rappresentative form and learning aid to operational device of epistemic relevance (De Toffoli, 2025) which activates “manipulative imagination" (De Toffoli & Giardino, 2014).

In this contribution, we develop a two-level discourse on scientific images.

In the first part, we focus on scientific images produced through human gestures. As Dondero (2015; see also 2024) argues, the scientific image has a Peircean diagrammatic function: it increases knowledge through the manipulations and experiments performed on it. We show how this function unfolds in a specific way of “reading images” based on two main features: the nonlinearity of reading and the dependencies between the whole and its parts (Bordron 1991; Goodman 1977). In this sense, the visual interpretation of images cannot be considered a mere duplication of accessing verbal texts.

We then argue that the meaning of the image must be understood through two levels of value circulation (Paolucci 2010). The first is the level of textual immanence, which allows us to conceive of the image as a signifying totality and to analyse its internal relationships and connections according to categories selected for the purposes of analysis: the topological, chromatic, and eidetic categories of Plastic Semiotics (Greimas 1984), as well as the texture as enunciative apparatus (Dondero 2020). The second is the level at which this value is exchanged with what lies outside the image-text. In the specific case of the scientific image, we often encounter a meta-textual foreground, in which the representation of an object, a method, or a scientific process is placed within another text belonging to a different semiotic system, such as verbal language. These interpretative circumstances require a work of translation that we define as intersemiotic translation (Jakobson 1959, D’Armenio et al. 2025; Dondero & Fontanille 2012; Eco 2012; Dusi 2000). A further external factor concerns the comparison between the value produced by the image and the knowledge brought by the observer’s competences regarding the represented phenomenon, as well as the relationship between that phenomenon and its terminus a quo (Eco 1997), which activates the semiosis of the graphic representation under examination.

In the second part, we compare scientific images produced through conventional scientific practices with scientific images generated by text-to-image systems (e.g., Midjourney). First, we consider the relevance of the connections among the data on which these systems depend. In the automatic analysis of large corpora, such connections can assume different configurations of meaning depending on the neural architectures and algorithms used to explore the datasets. Second, we examine how these different ways of identifying relations among data (attention mechanisms, cosine similarity, segmentation, and other procedures) have a substantial impact on the rendering of the scientific image.

In the context of scientific images for didactic purposes, the generation of synthetic images appears particularly useful for producing tailor-made educational resources and for personalising content. Our analysis therefore aims to define a perimeter of usability for artificial scientific images within educational practices (Mayer 2020). We showcase a corpus of scientific images drawn both from instructional materials and from multimodal generative AI systems, organised into three categories: objects (e.g. the cell), methods (e.g. how to use the microscope), and processes (e.g. the water cycle).

One major obstacle to the application of mathematical problem-solving strategies is the lack of an adequate mental representation of the problem. Children often struggle to imagine the domain of objects and transformations described in a problem situation and instead attempt to infer the required mathematical operation directly from the verbal formulation of the problem text (Nesher, 1980).

Cognitive psychology emphasizes the distinction between two phases in problem solving: representation (problem comprehension) and solution. This distinction suggests that many difficulties in mathematical problem solving may originate from an inadequate representation of the problem (Mayer, 1982).

Literature in mathematics education describes that visual and symbolic representations—e.g. number lines, coordinate systems, diagrams—play a crucial role in supporting students’ understanding of mathematical concepts, across different stages of education and cognitive profiles (Duval, 2017; Robotti et al., 2016; Hawes, 2020).

Mathematical problems expressed in textual form—defined as word problems or story problems—play a central role in primary mathematics education. They are a form of didactic transposition which supports students' understanding of quantities and transformations of quantities by embedding them into a narrative context (Zan, 2016; Verschaffel & De Corte, 1997). This pedagogical practice has been widely studied in mathematics education (Zan, 2011; Hickendorff et al., 2021; Fuchs et al., 2015; Kintsch, 1985) and represents a favourable domain for investigating numerical and computational reasoning in primary school children as well as in individuals with cognitive disabilities. Solving such problems typically requires the integration of linguistic comprehension, working memory and reasoning abilities (Hickendorff et al., 2021; Fuchs et al., 2015).

Literature on Augmentative and Alternative Communication (AAC) pays considerable attention on how the iconic nature of signs supports everyday communication and language comprehension (Drager et al., 2010; Beukelman et al., 2020). By contrast, limited attention has been paid to the composition and spatial articulation of symbol systems for representing logical or mathematical relations.

Symbol systems can be understood as epistemic tools that organize logical relationships between concepts from visual and spatial perspective (Meletis, 2025; Duval, 2017; Roth & Tobin, 1997). Visualizing such relationships may help learners avoid superficial problem solving strategies, such as performing arithmetic operations without verifying the plausibility of the result within the real-world context of the problem (Carotenuto et al., 2021).

This article presents examples of story problems represented through AAC symbols arranged on communication boards that spatially articulate mathematical relations involving quantities and quantitative changes. The aim is to verify whether representing the logical-mathematical structure of a problem facilitates the understanding of quantities and basic operations.

These symbol configurations may serve as a starting point for future empirical research on the teaching of mathematical concepts through AAC, where the meaning of the operation emerges from the interdependence between graphic signs and the visual space of the operation, rather than from isolated symbols.

More broadly, this perspective invites reconsidering AAC not only as a communication aid but also as a pedagogical tool for accessing mathematical knowledge, promoting personal autonomy particularly for learners with cognitive disabilities or visual-perceptual cognitive styles (Mottron, 2006; Soulières, 2009).

In today's society, which is inundated with information primarily conveyed visually, the ability to interpret graphical representations and images critically is essential for active and responsible citizenship. Teachers at all levels must, therefore, undertake the crucial challenge of developing this skill in their students by making instructional choices aimed at deconstructing clichés — such as the idea that "a picture is worth a thousand words" — and protecting students from the "tyranny of numbers," which are often presented as objective and indisputable data (Cairo, 2019; Weikmann & Lecheler, 2023; OECD, 2023).

To take a step forward in this direction, this article examines the role of visual models (such as graphs, diagrams, and simulations) in physics education, drawing on Roth and Tobin's (1997) findings regarding the difficulties students face in connecting observed phenomena to their mathematical formalisations. The two scholars argue that standard physics lessons involve multiple translations of an observed phenomenon into ontologically distinct representations, which teachers—having been enculturated into the scientific community—often treat as self-evident. For instance, a rolling ball may be represented by an experimental reconstruction, a table showing its positions at specific moments in time, a graph of the curve, or the equation of uniformly accelerated motion. However, their research—along with the misconceptions that emerged during lectures aimed at future primary school teachers regarding classic physics topics such as motion along an inclined plane, the oscillation of a pendulum, and the principle of inertia (Tombolato, 2020)—shows that these representations often remain a "black box" for most students (e.g., Shah & Hoeffner, 2002).

Against this backdrop, the aim of this study is to clarify the epistemological and learning constraints that must be satisfied for visual models to be used effectively in education, so that they can serve as mediators of knowledge (Bruner, 1974; Damiano, 2013; Kress, 2009) and help students reconcile their perceptual experience of the world with its scientific representation. Drawing on Ronald Giere’s (2010, 2013) intentional conception of representation, we argue that whether a physics novice can correctly decode the information associated with a visual model, draw inferences about its target system (e.g., the observed phenomenon), and connect it to mathematical formulas does not depend exclusively on the model’s intrinsic visual characteristics. Rather, it depends on understanding the modeller’s epistemic intentions, which are concretely realised through epistemic operations—such as abstraction and idealisation (Portides, 2007, 2021)—that require a proficiency in thinking in terms of variables (Arcà & Guidoni, 1987).

The work is divided into two parts. In the first part, after introducing the perspectival nature of scientific representation (Giere, 2010), we will classify visual models based on their underlying epistemic operations to account for the varying cognitive load (Sweller, 2023) associated with processing them. In the second part, we will provide teachers with guidance on combining different types of visual representation to optimise the management of the cognitive load associated with complex tasks such as the mathematisation of nature.