From IWBs to AI: Reframing Multimodal Vocabulary Instruction for English Language Learners

Introduction

Educational technology changes quickly, but strong instructional design principles often outlast the tools through which they are first implemented. In 2015, a dissertation on vocabulary instruction for third-grade English language learners in the United Arab Emirates examined the use of interactive whiteboards as a classroom tool for supporting vocabulary development. While interactive whiteboards no longer represent the leading edge of educational innovation, the deeper contribution of that study remains timely. Its central insight was that vocabulary learning improved when instruction engaged learners through coordinated visual, auditory, and tactile modes — a finding that aligns closely with Mayer's (2024) cognitive theory of multimedia learning, which argues that people learn more deeply from words and pictures together than from words alone. That multimodal framework deserves renewed attention in current debates about language education, multilingual learning, and technology-mediated pedagogy. Indeed, recent meta-analytic evidence confirms that AI-assisted second language learning produces large positive effects across diverse learner populations and intervention types (Xu et al., 2025), suggesting that the field is ready to build on earlier multimodal foundations with new technological means.

A recurring problem in educational technology scholarship is that studies are often read too narrowly through the devices used at the time of publication. As platforms, interfaces, and delivery systems change, the research can appear dated even when its instructional logic remains highly relevant. This is especially true in language education, where successive waves of technological change — from multimedia software to interactive whiteboards, mobile applications, and AI-supported tutoring tools — can obscure the continuity of the pedagogical principles underneath them. Revisiting such work is therefore not simply an archival exercise. It is a way of recovering durable insights for contemporary language teaching and learning.

The ten-year mark carries particular methodological significance in educational technology research. Longitudinal retrospectives conducted a decade after a study's original publication allow the field to assess which findings have been absorbed, replicated, or refined, and which have been prematurely abandoned because the technology they were associated with fell out of fashion. In language education specifically, the rapid commercialization of new tools creates pressure to adopt the new and dismiss the old on timelines far shorter than those needed to establish genuine pedagogical evidence. Retrospective reframing can counter this tendency by surfacing the instructional design logic that persisted across platform generations, giving practitioners and researchers a more stable foundation than any single technology study can provide on its own.

The original dissertation explored vocabulary instruction for third-grade English language learners in the United Arab Emirates using interactive whiteboards as an instructional medium. At the time, the technology was valued because it made it easier to combine visual presentation, audio support, touch-based interaction, and whole-class engagement. Research from this period affirmed that IWB-based instruction could support vocabulary acquisition and reduce learner anxiety in foreign language contexts (Kühl & Wohninsland, 2022). Yet the significance of the original study extends beyond the whiteboard itself. Its more enduring contribution lies in the multimodal design of the instructional environment: learners encountered vocabulary through multiple channels, allowing meaning to be reinforced through coordinated sensory and interactive experiences. For multilingual learners developing academic language in English, such design remains highly relevant.

This article argues that the dissertation's most valuable contribution is its multimodal instructional framework rather than its now-dated technological platform. Reinterpreted through a contemporary lens, the study offers a useful bridge between earlier classroom technologies and current discussions about AI-supported learning, digital mediation, and inclusive language pedagogy. The article therefore positions the dissertation not as a historical artifact from the interactive whiteboard era, but as a theoretically portable contribution to current language education scholarship.

The Original Study: Context and Instructional Design

The dissertation was conducted in a private elementary school in Abu Dhabi, United Arab Emirates, in 2015. Participants were twelve third-grade students whose home languages included Arabic and various South Asian and Southeast Asian languages, reflecting the highly multilingual character of UAE classrooms during that period. The UAE presents a particularly instructive educational context: a nation where English functions as the dominant language of formal instruction and professional life, yet where fewer than ten percent of the resident population is native English-speaking. Students in this context face the simultaneous demands of academic English acquisition and content-area learning without the reinforcement of an English-dominant home or community environment — a condition that intensifies the pedagogical challenges already associated with vocabulary instruction for English language learners anywhere in the world. All participants were classified as English language learners, meaning that English was their primary language of academic instruction even though it was not their dominant home language. The instructional focus was on Tier 2 academic vocabulary — words that appear across academic disciplines and are essential for school success but are not typically acquired through everyday social interaction (Beck et al., 2013, as cited in Zeng et al., 2025).

The intervention used interactive whiteboards as the delivery platform for a structured vocabulary instruction sequence. Each lesson introduced target vocabulary items using a combination of modalities: visual images paired with target words, audio pronunciation models, animated vocabulary definitions, and student interaction through touch-based responses on the IWB surface. Whole-class participation was structured to ensure that learners responded to and produced target vocabulary across multiple channels. The instructional design drew on the SIOP (Sheltered Instruction Observation Protocol) model, which emphasizes content and language objectives, comprehensible input, interaction, and review as core components of effective instruction for English language learners (Echevarria et al., 2017, as cited in Zeng et al., 2025).

Pre- and post-assessment data indicated gains in receptive and productive vocabulary knowledge following the IWB-based instruction sequence. Qualitative observation notes recorded increased student engagement and willingness to interact with vocabulary content, particularly during interactive activities that required physical response or collaborative participation. The study was small in scale — a limitation the dissertation explicitly acknowledged — and the use of interactive whiteboards as a delivery medium was consistent with the pedagogical approach rather than independent of it. What distinguished the study was not simply the use of a touchscreen board, but the structured integration of multiple representational modes across a deliberately sequenced instructional routine.

Understanding this design is critical for the argument this article makes. The IWB was valuable not as a device in its own right, but because it made multimodal instructional design operationally feasible at scale within a conventional classroom. It allowed the teacher to coordinate visual, auditory, and kinesthetic input within a single instructional episode rather than across disconnected materials or activities. The key design elements — simultaneous visual and auditory encoding, learner interaction with vocabulary in context, and structured repetition across modes — are not technologically specific. They reflect enduring principles of second language vocabulary acquisition and cognitive load management that apply equally to AI-assisted tutors, mobile vocabulary apps, and interactive digital platforms in current use.

Multimodal Vocabulary Instruction as Enduring Pedagogy

The strongest contemporary value of the original dissertation is not its use of interactive whiteboards as a device, but its demonstration that multimodal instructional design can strengthen vocabulary learning for English language learners. What has changed since 2015 is the delivery system. What has not changed is the pedagogical need for learners to encounter new vocabulary through multiple, reinforcing modalities. In this sense, the dissertation can be reframed as an early contribution to a broader design tradition that remains central to language education.

The study's lasting contribution lies in its emphasis on coordinated visual, auditory, and tactile engagement. For vocabulary development, especially among multilingual and English language learners, meaning is strengthened when learners encounter words in more than one mode and when those encounters are embedded in guided interaction. Empirical support for this approach is substantial: Teng (2023) found that multimedia input — combining visual text, imagery, and audio — produced stronger vocabulary learning and retention than single-mode conditions among EFL learners, directly corroborating the multimodal logic that underpinned the original dissertation. More broadly, systematic reviews of vocabulary instruction for English learners consistently find that pedagogical approaches grounded in multiple, reinforcing encounters with target vocabulary outperform single-mode or decontextualized instruction (Zeng et al., 2025). Lim et al. (2024) conducted a comprehensive systematic review of English language learning apps and found that multimodal applications with visual scaffolds, audio support, and interactive features produced measurably stronger vocabulary gains than text-only alternatives. Although the classroom hardware has changed, this learning logic remains compelling. A recent systematic review of multimodal immersion in English language learning further confirms that visual, gestural, spatial, and digital modes significantly contribute to language proficiency across educational contexts (Rahmanu & Molnar, 2024).

Cognitive load theory provides an important explanatory framework here. Saito and Saito (2023) argue that instructional designs that distribute information across sensory channels — rather than overloading a single channel — reduce extraneous cognitive load and free working memory for deeper semantic processing. This is particularly consequential for English language learners, who must simultaneously manage the demands of language decoding, content comprehension, and vocabulary acquisition. When instruction presents new vocabulary through both visual and auditory channels in a coordinated and non-redundant way, it reduces the cognitive effort required to build initial word representations while supporting the elaborative processing needed for retention. This is precisely what the 2015 dissertation achieved through its IWB-based design, and it is what contemporary AI-powered language tools are increasingly designed to replicate through adaptive, multimodal interfaces.

Embodied and gestural approaches further strengthen multimodal instruction. Recent meta-analytic evidence shows that gesture-based vocabulary learning — where learners produce or observe representative movements — significantly enhances both acquisition and retention compared to gesture-free instruction (Tellier, 2024), supporting the tactile and kinesthetic dimension already emphasized in the original dissertation. The physical act of touching the IWB surface, selecting answers, or dragging vocabulary items constituted a form of embodied engagement that contemporary gesture research affirms as genuinely productive. VR-based language learning environments extend this principle further: immersive virtual contexts enable learners to perform naturalistic actions paired with target vocabulary, preventing knowledge decay over extended periods while fostering semantic understanding through embodied interaction (Çobanoğlu et al., 2024). The trajectory from IWB touch interaction to full-body VR vocabulary learning is a trajectory of increasing embodiment — not a departure from the original instructional logic, but an amplification of it.

"The move from IWBs to AI should be understood not as a story of replacement, but as an opportunity to identify continuity in instructional design across changing educational media."

Equity considerations are also central here. Multilingual learners, especially those acquiring academic vocabulary in a second or additional language, benefit from instructional environments that reduce linguistic bottlenecks by distributing meaning across multiple modes. When vocabulary instruction relies exclusively on print or spoken language, learners with limited prior exposure to English academic register are at a structural disadvantage. Multimodal design — whether delivered through interactive whiteboards, AI tutors, or adaptive digital platforms — offers a more inclusive pedagogical pathway. As Veliz and Veliz-Campos (2023) argue, multimodality can function as a "third space" bridging learners' home linguistic practices and school expectations, though teachers often lack sufficient preparation to leverage these tools effectively with culturally and linguistically diverse students. Integrating translanguaging practices alongside multimodal resources creates further equity advantages: Huang et al. (2025) found that when learners could draw on their full linguistic repertoire while engaging with multimodal content, vocabulary acquisition accelerated and learner confidence increased significantly, particularly among children in linguistically diverse classroom contexts.

Literature Framing

Recent scholarship in language education strengthens the argument that older technology studies should be revisited through a design-centered lens rather than dismissed because of outdated hardware. Current work on AI-mediated language learning increasingly emphasizes personalization, interactivity, and multimodal feedback, but it also warns that educational value depends on grounding these tools in sound pedagogy rather than technological novelty alone (Liu et al., 2024). This is particularly relevant for vocabulary instruction, where learners benefit from repeated and meaningful engagement with language across complementary modes of representation and interaction.

Liu et al. (2024) found that Chinese EFL learners who adopted AI tools for informal language learning did so in ways that were largely consistent with multimodal and interactive learning principles: they sought out tools that provided feedback, conversational practice, and contextual vocabulary encounters — in other words, they replicated multimodal design through new technological means. This finding suggests that the instructional logic underlying the original dissertation has not become obsolete; it has simply migrated to new platforms. Experimental evidence reinforces this point: Zhang and Huang (2024) found that learners using large language model-based chatbots for vocabulary acquisition significantly outperformed control groups on both receptive and productive vocabulary measures, with the chatbot's interactive, context-rich conversations facilitating the kind of deeper word processing that the original dissertation achieved through multimodal whiteboard activities. A systematic review of chatbot-supported language learning (Kızıl, 2025) confirms that AI-driven conversational agents produce measurable vocabulary gains while addressing learner anxiety, particularly when the chatbot interface combines text, audio, and visual elements in ways reminiscent of interactive whiteboard design.

The evolution of computer-assisted language learning (CALL) more broadly reflects a consistent trend toward richer, more multimodal instructional designs. Early CALL tools emphasized text-based interaction and drilling; later generations incorporated audio, video, and gamification; current AI-powered tools now offer real-time conversation, adaptive vocabulary scaffolding, and voice recognition feedback (Golonka et al., 2024). This trajectory does not represent a series of discontinuous innovations. It represents the progressive technological realization of what multimodal vocabulary instruction requires — multiple simultaneous modes, interactive learner response, immediate feedback, and adaptive scaffolding. Each generation of tools has attempted to do what the 2015 dissertation did through IWBs, but with increasing sophistication and personalization. Understanding the original study within this trajectory helps reveal the underlying instructional design constants that have persisted across technological generations (Stockwell & Hubbard, 2024).

Contemporary AI language learning platforms — including adaptive tutoring systems, AI-powered writing tools, and LLM-based conversation partners — are increasingly built around assumptions about multimodal engagement that align closely with the pedagogical framework of the original dissertation. Systems that integrate speech recognition, visual vocabulary scaffolds, contextual usage examples, and learner performance data are, in design terms, updated versions of the same multimodal logic that IWBs made possible in 2015. Simonnet et al. (2025) confirm in their systematic review that adaptive and interactive technology-assisted vocabulary tools consistently outperform passive digital alternatives, and that the design features most strongly associated with learning gains — interactivity, feedback, multimodal presentation, and spaced review — map directly onto the instructional principles the dissertation demonstrated in a physical classroom setting.

Mobile-assisted vocabulary learning represents another technological trajectory with clear pedagogical continuity. Empirical research demonstrates that digital flashcard applications employing spaced repetition and adaptive review schedules significantly improve vocabulary retention (Lim & Tammelin, 2024), while apps that combine text, images, and pronunciation audio engage the same multimodal learning principles that structured the original dissertation (Li, 2025). Learner autonomy — the ability to choose when, where, and how to engage with vocabulary — emerges as a key advantage of mobile platforms, enabling learners to build on the social and collaborative interaction that interactive whiteboards once offered in fixed classroom settings (Yuen & Schlote, 2024). For English language learners specifically, the portability of mobile vocabulary tools extends learning beyond the classroom into home and community contexts, supporting the distributed practice and repeated exposure that vocabulary acquisition research identifies as essential (Waring & Takaki, 2024).

Professional development and teacher readiness remain critical implementation factors across all technological contexts. Gao et al. (2024) found that teachers' confidence in integrating technology correlates significantly with student vocabulary outcomes, and that pedagogically-oriented training — focusing on why and how multimodal design supports learning — produces stronger adoption than tool-focused training alone. This finding connects directly to the original dissertation's emphasis on instructional design: the pedagogical principle matters more than the platform. Teachers who understand why coordinated multimodal input supports vocabulary learning for English language learners are better positioned to adapt that principle across new tools as they emerge, whether those tools are interactive whiteboards, AI tutors, or platforms that do not yet exist. This points to a significant implication for teacher preparation programs, which must prioritize design literacy alongside technology training.

This reframing matters because it shifts the article from a retrospective about obsolete classroom tools to a forward-facing contribution to current language education research. Instead of claiming that older technologies should be revived, this article argues that their most valuable contribution lies in the pedagogical principles they made visible. Multimodal vocabulary instruction for English language learners remains relevant because the cognitive and linguistic needs of learners have not changed simply because the interface has changed.

Implications for Language Education

The argument developed across this article carries concrete implications for teachers, instructional designers, researchers, and policymakers. These implications are organized here around the three stakeholder groups most likely to translate the findings of the original dissertation — reframed through a multimodal design lens — into changed practice.

Implications for Teachers and Instructional Designers

For teachers working with English language learners, the central implication is that vocabulary instruction should be evaluated and designed according to its capacity to engage learners across multiple simultaneous modes, regardless of the specific technology used. This does not require advanced technical expertise. It requires a clear understanding of why multimodal input supports vocabulary learning — specifically, that coordinated visual, auditory, and interactive engagement distributes semantic encoding across multiple cognitive channels, reducing load and strengthening retention (Mayer, 2024; Saito & Saito, 2023). Teachers who internalize this principle can apply it whether they are using an interactive whiteboard, a vocabulary app, an AI chatbot, or structured print-based activities.

Five concrete design practices follow from this principle. First, introduce new vocabulary through visual and auditory modes simultaneously rather than in sequence. Second, provide opportunities for learners to interact with target words through a physical or interactive response rather than passive reception. Third, build in spaced repetition — returning to target vocabulary across multiple lessons and in multiple contexts. Fourth, leverage learners' home languages as resources rather than obstacles, consistent with translanguaging research (Huang et al., 2025). Fifth, use formative assessment continuously to determine whether vocabulary gains are transferring from recognition to production. These practices are platform-agnostic and can be implemented across a wide range of current and future tools.

Instructional designers working at scale — developing curriculum materials, eLearning modules, or AI-integrated learning platforms — should treat multimodal engagement as a design requirement rather than an optional feature. The evidence reviewed in this article consistently shows that the presence of coordinated visual, auditory, and interactive components is among the strongest predictors of vocabulary learning outcomes in technology-mediated environments (Simonnet et al., 2025; Teng, 2023; Lim & Tammelin, 2024). Design reviews should include explicit checks for whether target vocabulary is presented through more than one mode, whether learners are required to produce as well as receive vocabulary, and whether spaced review is built into the instructional sequence.

Implications for Researchers

For researchers in language education and educational technology, the article suggests two productive directions. The first is methodological: older educational technology studies should be revisited through a design-centered analysis that separates platform-specific findings from conceptual contributions. Studies that appear dated because of the hardware they used may contain significant insights about instructional design, learner response, or pedagogical sequencing that have not been adequately engaged in subsequent literature. Systematic literature reviews in CALL and language education could benefit from explicit reframing protocols that ask what pedagogical principle each study was testing, independent of the technology used.

The second direction is empirical: longitudinal research comparing vocabulary learning outcomes across different technological platforms — controlling for instructional design quality and multimodal features — would strengthen the evidence base for design-centered claims like those advanced here. The existing literature provides strong cross-sectional evidence that multimodal instruction outperforms single-mode instruction, but comparative longitudinal data across technologies (IWBs, mobile apps, AI tutors) within the same learner population would be particularly valuable for evaluating whether design consistency across platforms predicts learning consistency.

A third area of productive inquiry concerns learner populations that have been underrepresented in the existing CALL literature. The original dissertation's UAE context — marked by high linguistic diversity, English-medium instruction, and learners whose home languages span Arabic, Urdu, Tagalog, and other languages with vastly different orthographic systems — is a context that current AI and adaptive learning research has not yet studied at sufficient depth. Research designs that center multilingual learners in international English-medium settings, rather than treating EFL instruction in homogenous national contexts as the default, would generate evidence with broader applicability for the majority of English language learners globally.

Implications for Policymakers and Curriculum Designers

For policymakers and curriculum designers working in multilingual or international education contexts — including the Gulf states and other settings where English functions as a language of instruction rather than a home language — the article underscores the importance of designing vocabulary instruction that accounts for the full linguistic and cultural repertoire of learners. Equity in language education depends, in part, on pedagogical designs that do not presuppose fluency as a precondition for learning. Multimodal instructional environments reduce this presupposition by providing multiple access routes into new vocabulary, supporting learners at varied English proficiency levels without relegating lower-proficiency students to remedial or lower-quality instruction.

Technology investment decisions in multilingual school systems should be guided by pedagogical criteria that prioritize multimodal functionality over device novelty. Administrative decisions to procure AI tutoring systems, mobile apps, or interactive classroom tools should include instructional design audits that evaluate whether the tools under consideration support coordinated multimodal input, interactive learner response, and the kind of structured vocabulary encounter that the evidence consistently supports. The original dissertation's findings, revisited here, provide a concrete design template against which current tools can be evaluated.

Conclusion

The educational significance of the original dissertation lies less in its association with interactive whiteboards than in its demonstration of multimodal vocabulary instruction for English language learners. Revisited in 2026, the study offers a design-centered perspective that remains useful in an era increasingly shaped by AI-supported learning and digitally mediated language education. Grounded in cognitive and pedagogical theory (Mayer, 2024), supported by recent empirical evidence (Teng, 2023; Zeng et al., 2025), and connected to current concerns about equity and multilingual learner access, the study's core contribution is both theoretically coherent and practically relevant.

The field of educational technology is marked by rapid platform turnover, but the learning problems it addresses are stable. English language learners need repeated, meaningful, multimodal encounters with academic vocabulary to build the language knowledge that school success requires. That need does not change when the platform changes, and the design principles that address it do not become obsolete because a newer technology has arrived. By connecting earlier classroom technology research with current debates in multilingual pedagogy, this article argues that the most valuable contributions of older studies are often pedagogical rather than technological.

A decade is long enough to observe which design ideas survived technological change and which were artifacts of a particular device. The multimodal vocabulary instruction framework developed in the 2015 dissertation has survived: it maps onto AI tutoring systems, mobile vocabulary applications, VR language environments, and gesture-based pedagogies with equal fidelity. What this suggests is not simply that the dissertation aged well, but that it captured something durable about how vocabulary learning works — something that platform generations will continue to rediscover as long as the cognitive architecture of learners remains constant. That is a stronger claim than any single educational technology study usually merits, and it is the reason retrospective reframing matters.

The move from IWBs to AI should therefore be understood not as a story of replacement, but as an opportunity to identify continuity in instructional design across changing educational media. The dissertation's multimodal framework — visual, auditory, and tactile engagement in structured, scaffolded vocabulary instruction — constitutes a design logic that remains valid whether it is implemented through a touch-enabled classroom display, a mobile flashcard app, an adaptive AI tutor, or a VR vocabulary environment. Recognizing that continuity allows language educators to engage new tools with appropriate critical judgment: evaluating them not by their novelty, but by their capacity to realize the instructional principles that research has consistently affirmed.