Theoretical background: Evidence-based educational decision-making draws on three types of information: internal, social and external evidence (Blumenthal & Mahlau, 2015; Kuhr & Kulawiak, 2018). External evidence, based on scientifically obtained information, can be divided into three dimensions: efficacy, effectiveness, and efficiency (Beushausen & Grötzbach, 2018; Streiner, 2002). Efficacy studies (or explanatory trials) are conducted under controlled experimental conditions and are characterised by high internal validity, meaning the probability of measuring the benefits of an intervention is high (Eisele-Metzger & Breuer, 2021; Singal et al., 2014). Effectiveness studies (or pragmatic trials) are conducted under practical conditions and therefore have high external validity (ibid.). In the context of language support in schools, they provide information about whether an intervention is effective in school practice. Thus, both study designs fulfil different functions within external evidence. For educational decision-making processes, it is essential to distinguish between the two to consider the dimension of evidence on which a methodology has been tested.

Research question: A two-stage process is employed to validate PROSAIOS and evaluate its usability. First, the optimisations that can be derived from expert assessments are examined. Secondly, the assessments of different user groups regarding the usefulness, comprehensibility, and reliability of the revised instrument are reviewed. 

Methodology: The need for optimisation was examined in a pilot study. To this end, ten expert evaluations were collected using PROSAIOS for a given study (Smith-Lock et al., 2013), and PROSAIOS revision assessments were compiled. Based on this feedback, PROSAIOS was revised and will undergo a final review in the second validation stage. The usefulness and comprehensibility of PROSAIOS will be assessed, as will the ratings (k = 6, four-point scales) from three user groups (scientists, students, and practitioners, with approximately n = 24 in each group) for four different studies, using digital survey tools (soscisurvey.de). The efficacy and effectiveness studies to be evaluated describe interventions for vocabulary improvement. 

Results: The pilot study revealed high inter-rater reliability (n = 9; intra-class correlation coefficient r = .944*** [CI .824 – .993]). The second stage of validation will take place in spring 2026. This paper presents the full set of results on inter-rater reliability, including intra-class correlations and internal consistency, as measured by Cronbach's alpha. Additionally, users' assessments of the instrument's usefulness and comprehensibility are evaluated and presented in a comparative manner.

Discussion: The high level of consistency in the ratings provided by different raters in the pilot study indicates that the PROSAIOS instrument allows for accurate differentiation between efficacy and effectiveness. The second validation stage, which is currently in preparation, will examine whether the consistency of assessments conducted by scientists, practitioners, and students is confirmed. 

Educational Significance of the research: The PROSAIOS validation aims to address a methodological gap that has previously existed in educational discourse. Although instruments for differentiating between efficacy and effectiveness are well-established in medicine, there has been a lack of validated instruments in education. This currently prevents practitioners from making informed decisions regarding evidence-based educational practice and the scientific classification of the evidence dimension of educational methods in language support and beyond.

Introduction

Spatial thinking, the ability to represent, manipulate, and reason about spatial relationships, is fundamental to geometry learning and broader STEM achievement. For young children, Augmented reality (AR) technologies offer possibilities for supporting challengeable spatial thinking tasks such as transforming between 2D and 3D representations by enabling children to manipulate and observe virtual 3D objects embedded in physical space.

While previous studies have documented positive learning outcomes associated with AR-based geometry activities, the field lacks a fine-grained understanding of the micro-level interaction processes through which children engage with AR environments and how these interactions contribute to spatial thinking. Understanding these interaction processes can provide empirical grounding for designing AR learning environments and activities as well as promote further theoretical understanding of how spatial thinking emerges through embodied activity.

Accordingly, this study explores: how do children’s interactions support spatial thinking during the construction of 3D objects from 2D prompts in an AR environment?

Methods

Twenty-one students from Grades 2–5 (ages 6–10) across six London primary schools participated in AR-based geometry construction tasks. Using a mobile AR application, children built 3D objects from 2D pictorial prompts. Each session was recorded from two perspectives: an overhead camera and a screen recording, synchronised for analysis.

First, we conducted a reflexive thematic analysis (Braun & Clarke, 2006, 2019) of the videos based on the tangible interaction framework (Hornecker & Buur, 2006), generating eight types of interactions. Second, we performed a micro-ethnographic analysis (Streeck & Mehus, 2004) according to the constrained co-occurrence framework for spatial thinking (Khan et al., 2015) to explore how these interactions unfolded moment by moment and formed events that supported spatial thinking.

Results and Discussion

Eight types of interactions were promoted by AR affordances: changing component number, changing spatial relations, changing component properties, changing one's own position, changing posture, keeping static to observe, referencing physical objects, and talking with an instructor. These interactions were dynamically combined throughout the construction process.

Micro-ethnographic analysis synthesised these interactions into nine spatial thinking events, which clustered into four underlying mechanisms. First, as one of the underlying mechanisms, children improved the accuracy and sophistication of mental representations through visualising and modifying mental models, building interconnectedness between components, and recognising shapes and their characteristics. Second, they perceived spatial information intentionally by highlighting specific object information and perceiving multi-perspective representations. Third, they built spatial information by connecting real and virtual objects by measuring virtual objects using physical references and constructing positional relations. Fourth, they built spatial information by connecting self with objects through constructing relations by viewing from specific perspectives and using body position and posture to generate new spatial understanding.

Three insights emerge from these findings. First, building on prior AR research that conceptualises virtual objects as externalisations of learners’ mental models, this study further shows that the relationship between the objects and mental models is dynamically reciprocal. In the activity, children reported that their mental models became clearer and more detailed as they adjusted the size and direction of virtual shapes and moved to combine them during construction in AR. Second, this study suggests that children’s movements and postural shifts are associated with changes not only in perspective-taking but attentional allocation. For example, children employed fine-motor actions, such as shifting weight between feet and rotating the wrist, to gain new visual access and selectively enlarge specific elements of the object. Third, children actively incorporated themselves into the environment. They positioned themselves at specific locations and moved along circular paths to create linear, symmetrical, or centrosymmetric relations among objects.

Educational Significance

This study provides an empirical account of how AR support children’s spatial thinking through their interactions. Theoretically, the findings promote understanding of spatial thinking as dynamic and enacted process: children simultaneously refine mental representations through iterative interaction and actively create spatial relations by situating themselves within the problem space. Practically, the findings show the need to design AR environments that encourage physical movement, postural adjustment, and integration of physical and virtual objects, attending not only to digital content but to the configuration of physical space that enables embodied engagement. For assessment, the identified interactions, events, and mechanisms provide a basis for formative monitoring and evaluation of children's spatial thinking during AR activities.

Across Europe, the “digital transformation of schooling” is increasingly presented as necessary, urgent, and largely self-evident. Policy documents promise modernisation, preparedness, and future-proofing, yet their underlying educational assumptions often remain implicit. What, precisely, is being transformed when “school” is reframed through the horizon of digitalisation? Which images of education, agency, and “progress” are stabilised by contemporary strategies—and which educational questions become harder to pose once certain keywords and imaginaries harden into policy common sense?

This paper offers an exploratory theoretical contribution that treats digital education policy as a philosophical problem of educational meaning. Rather than approaching policy texts with a predetermined diagnosis (e.g., that they are primarily “instrumental” or “technocratic”), it adopts a disciplined interpretive openness: it reconstructs what the documents themselves present as problems, solutions, and normative horizons, and then thinks those reconstructions through the Bildung tradition (Humboldt, Dilthey, Klafki, Benner). Bildung is mobilised not as a single doctrine but as a family of educational philosophies concerned with formation (self-relation and world-relation), judgement, responsibility, and the integrity of education beyond functional adaptation. From this perspective, digitalisation is not merely a technical shift; it is a normative re-description of what counts as “good” schooling, “educational improvement”, and legitimate educational agency under digital conditions.

Empirically, the paper is anchored in a small, transparent corpus of policy “cornerstones” across European and national levels: the European Commission’s Digital Education Action Plan 2021–2027, Slovenia’s national Digital Education Action Plan (ANDI) 2021–2027, and Austria’s digital schooling policy architecture as articulated in the Masterplan for the Digitalisation of Education and the subsequent 8-Point Plan for a Digital School (Digitale Schule). The goal is not to evaluate these policies as effective or ineffective, but to use them as exemplary sites for clarifying the normative grammar of contemporary digital education discourse and its internal tensions.

Methodologically, the paper proceeds in three steps. First, it performs a close reading that extracts each document’s (i) problem framing (what is presented as requiring transformation), (ii) model of agency (who is expected to act and how), and (iii) horizon of value (what counts as improvement). Second, it organises these findings into a comparative matrix of policy vocabularies, mapping recurrent terms, metaphors, and temporalities (e.g., inevitability, urgency, preparedness, innovation, resilience, responsibility), and tracing how “the digital” is positioned (as tool, environment, or quasi-inevitable condition). Third, it conducts a philosophical re-reading through Bildung by applying a set of Bildung-questions that structure interpretation without pre-judging outcomes: What conception of autonomy is implied (readiness versus judgement)? What kind of world-relation is presupposed (information environment versus shared world)? How is educational responsibility distributed among institutions, teachers, and learners? Which forms of educational experience are foregrounded or backgrounded (study, attention, transmission, critique, participation)? In this step, the Bildung tradition is treated as internally differentiated, enabling more than one line of rethinking: classical formation-oriented accounts, critical-constructive interpretations of Bildung, and contemporary philosophies of education that emphasise judgement and responsibility in mediated worlds.

The paper makes three contributions. First, it clarifies “digital transformation” as a normative category that does not simply describe technological change but redefines educational aims, temporalities, and expectations of agency. Second, it offers a comparative map of convergences and divergences across EU, Slovenian, and Austrian strategies (including the shift from framing-level articulation to operational steering in the Austrian case), thereby enabling a more precise philosophical discussion than generic claims about “policy instrumentality”. Third, it develops a Bildung-based interpretive framework for re-describing digital education policy in educational terms: not as a set of prescriptions, but as a vocabulary for articulating what might be at stake if digital transformation were oriented toward formation—cultivating judgement in digitally mediated environments, sustaining meaningful attention and study, and shaping responsible agency in relation to infrastructures that increasingly structure knowledge, interaction, and public life.