The research team undertook a document analysis of selected middle school science textbooks from the two countries, coding an extraordinary 1,797 visual representations spread across 620 pages. Every image was classified with a multidimensional framework that captured formal characteristics, such as whether a visual was a photograph, a diagram, a graph or a composite; relational characteristics, such as how the image connected to the verbal text; pedagogical characteristics, such as whether the visual merely decorated the page or actively supported conceptual understanding; and sociocultural dimensions, including the contexts in which physics phenomena were portrayed. The coding combined qualitative interpretation with descriptive and inferential statistics, and intercoder reliability was established using kappa statistics, a standard guard against subjective drift when thousands of items must be judged consistently.

The headline finding is a study in contrasts. Turkish science textbooks contained more visual representations overall and a higher visual density per analysed page, meaning that a student turning the pages of a Turkish book encounters images at a noticeably faster rate. English science textbooks, by contrast, showed stronger text–image integration. This integration was achieved through indexical references, explicit textual signals such as phrases directing the reader to a specific figure, and through descriptive captions that explain what an image shows and why it matters. In other words, the English books may show fewer pictures, but they do more work connecting each picture to the argument of the text.

Why does this matter pedagogically? Decades of research in multimedia learning, most prominently the cognitive theory of multimedia learning developed by Richard Mayer, hold that well-designed combinations of words and images can dramatically improve comprehension of abstract scientific ideas, provided the two channels are coordinated rather than left to compete for the learner’s attention. Physics is arguably the discipline where this matters most, because its central concepts, from electric fields to energy transfer, are invisible and must be constructed mentally from models, diagrams and analogies. Studies of multiple representations in physics education have repeatedly shown that students who can translate fluently between words, equations and pictures develop deeper and more flexible understanding, while students who cannot often revert to memorising formulas without meaning.

Seen through that lens, the Turkish books’ visual abundance looks like a strength: variety gives students more entry points into difficult concepts and more opportunities to practise representational translation. Yet abundance without integration carries a risk documented in the cognitive load literature. If images are not explicitly linked to the relevant text, learners must expend mental effort deciding how the pieces fit together, and that extraneous load can crowd out the very reasoning the visuals were meant to support. The English books’ strategy of explicit referencing and descriptive captioning is precisely the kind of scaffolding that multimedia research recommends, because it tells the learner where to look and what to conclude.

The statistical analysis confirmed that these were not chance impressions. Cross-national differences were significant across most of the representational dimensions the framework measured, although the authors are careful to note that effect sizes ranged from small to large. That nuance matters. Some aspects of textbook design differ modestly between the two systems, while others, such as the density of visuals and the consistency of text–image linkage, diverge substantially. The authors interpret the pattern not as evidence that one country’s books are instructionally superior, but as evidence of different textbook design orientations, each shaped by its own curriculum traditions, publishing practices and educational culture.

Those orientations have identifiable roots. England’s science programmes of study, set out in the national curriculum, provide a tightly specified framework within which publishers design books that must serve a market of schools with diverse approaches, encouraging clear, self-contained text–image units. Türkiye’s system, guided by the Ministry of National Education and most recently by the Türkiye Century Education Model, has undergone successive curriculum reforms, and previous analyses of Turkish textbooks from 2002 to 2017 had already documented a rich but uneven visual landscape. The new study adds a further observation: Turkish textbooks show greater visual variety and limited forms of digital interactivity, while English textbooks more consistently provide the explicit textual references and descriptive captions that bind visuals to prose.

The comparative approach itself is part of the study’s contribution. Earlier work had compared textbook visuals between Australia and Taiwan, and between China and Singapore in the specific domain of optics, establishing that national context leaves a measurable fingerprint on how science is pictured for children. By extending this line to Türkiye and England, and by applying an integrated framework spanning formal, relational, pedagogical and sociocultural dimensions, the researchers offer other scholars a reusable instrument. Textbook analysts can now ask not only how many images a book contains, but what kinds they are, how they relate to the text, what pedagogical functions they serve, and whose world they depict.

That last question is becoming more urgent as science education grapples with equity and relevance. The sociocultural dimension of the coding framework acknowledges that visuals do ideological work: they signal who does science, in what settings, and with what technologies. A photograph of a child experimenting at a kitchen bench and a diagram of an industrial power station both teach physics, but they also teach implicit lessons about where physics belongs. Cross-national comparisons make such implicit choices visible, because practices that seem natural within one publishing culture appear as deliberate decisions when set against another.

For textbook designers and curriculum developers, the practical implications are refreshingly concrete. The study suggests a synthesis rather than a verdict: the visual richness and variety characteristic of the Turkish books, combined with the explicit referencing and descriptive captioning characteristic of the English books, would align closely with what the multimedia learning literature recommends. As digital and interactive textbooks proliferate, the authors note that forms of digital interactivity remain limited in the analysed materials, leaving open a rich agenda for future research into how dynamic, manipulable representations change the calculus of visual design. Meanwhile, the study stands as a reminder that the most ordinary classroom artefacts, the textbooks millions of children open every day, encode deep and nationally distinctive assumptions about how abstract science should be made visible, and that those assumptions can be measured, compared and improved.

Subject of Research: Cross-national analysis of physics-related visual representations in middle school science textbooks from Türkiye and England

Article Title: Cross-national analysis of visual representations related to physics in middle school science textbooks: pedagogical insights from Türkiye and England

Article References: Elmacı, H., Özmen, K., & Körhasan, N. D. (2026). Cross-national analysis of visual representations related to physics in middle school science textbooks: pedagogical insights from Türkiye and England. SN Social Sciences, 6(10), Article 513. https://doi.org/10.1007/s43545-026-01792-z

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01792-z

Keywords: visual representations, physics education, science textbooks, Türkiye, England, multimedia learning, document analysis, textbook design, science education, multiple representations, cognitive load, comparative education