{"id":"f539c3c6-9043-4d51-92df-37b3c0f97ecd","arxiv_id":"2504.12593","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"VR education should be designed around learner agency and legitimate participation rather than immersive fidelity or content retention alone.","lead":"This position paper argues that virtual reality's main educational value is the sense of agency it gives learners, not just immersion or interactivity. It proposes designing VR learning around legitimate participation so students can act as professionals, supporting situated learning in classrooms.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper equates designer-scripted control in VR with Lave and Wenger's legitimate peripheral participation, but omits the community-of-practice condition that the cited theory requires; the central inference from agency to situated learning is therefore unsupported.","rationale":"After reading the manuscript and the reader's verdict, I find the same load-bearing weakness the reader identified: the paper's central claim depends on treating VR-scripted task control as legitimate peripheral participation. The concern is not 'outside consensus' but an internal conceptual mismatch: the paper cites Lave and Wenger (1991) for the claim that situated learning requires LPP, and the design example lacks the social and productive conditions of LPP. I checked where this enters: Section 2.1, after correctly noting that LPP 'goes beyond using your body' and requires 'meaningful actions that progressively move you into an expert role.' The author's own example—control over a day in an engineer's life—grants no membership, no community, no mutual engagement, and no authentic product. The paper does not argue that a simulated community can confer legitimacy; it simply asserts that task control is LPP. Because this step is the only mechanism connecting agency to situated learning, the central argument is unsupported as written. I agree with the reader's assessment and would not change the CONDITIONAL verdict: the paper is a coherent position piece but must either expand its theoretical frame to defend VR-based legitimacy or soften the claim to 'task engagement' rather than LPP. My proposed check—applying Lave and Wenger's criteria to the described scenario—would settle whether the equivocation is real. No ad hominem; no fabricated issue. Credit where due: the paper openly cites its own null or weak findings on agency and learning gains, which is a fair honesty signal.","tokens_in":4494,"tokens_out":6217,"duration_ms":65909,"concrete_test":"Apply Lave and Wenger's criteria from reference [13] to the automotive VR scenario described in Section 2.1: (i) is there a community of practice that grants the learner legitimacy, (ii) is the task a genuine productive activity whose outcome matters beyond the simulation, and (iii) are there other practitioners with whom the learner interacts? If any answer is no, the claim that the design gives legitimate peripheral participation is unsupported. A second, empirical variant: add a virtual mentor and a collaborative task with a shared product to the same VR scenario and compare situated-learning outcomes to the solo-control version; if outcomes change, the missing social dimension is the operative factor.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim (Section 2.1) is that 'VR's potential to make transformative contributions to education rests on how well it is designed for agency,' and that designing for agency means 'giving learners legitimate participation in real-world tasks,' illustrated by 'control over a day in the life of an automotive engineer.' This step silently redefines Lave and Wenger's legitimate peripheral participation (LPP; reference [13]) from a social, community-grounded process into an individual, scripted simulation. In LPP, 'legitimate' means access is granted by a community of practice, 'peripheral' means the learner participates in the community's genuine productive activity with gradual centripetal movement, and 'participation' involves mutual engagement with other members and a shared enterprise. The proposed VR scenario has no other members, no community grant of legitimacy, no authentic product, and no consequences. The paper itself says learning is 'embedded in the context ... with community' and that situated learning 'explicitly requires legitimate participation'; yet the design example drops the community dimension. Thus the bridge from 'heightened agency' to 'situated learning' rests on an equivocation: a learner controlling a simulated day is not a peripheral participant in an engineering community. If this equivocation is real, the central claim is not established; the paper would need either a worked-out argument that virtual communities can confer legitimacy, or a weaker claim about simulation-based task engagement rather than LPP.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This position paper argues that virtual reality (VR) can enable situated learning by heightening learner agency, where agency is defined not as mere interactivity but as legitimate participation in authentic tasks. The author distinguishes this view from models like CAMIL that equate agency with interaction, draws on her prior empirical work on VR field trips, and proposes a design example involving a VR physics lesson set in automotive manufacturing. The paper concludes that VR's transformative educational potential hinges on designing for agency and legitimate participation rather than for immersion or content-retention efficiency alone.","tokens_in":4929,"tokens_out":3324,"duration_ms":36635,"significance":"If the proposed link between agency and situated learning were established, the paper would provide a useful design principle for VR learning environments and would challenge the dominant focus on presence and interactivity. The author deserves credit for honestly reporting counterevidence from her own studies, including that increased sense of agency did not improve content learning gains, and for attempting to reframe the value of VR around motivation and identity. However, the central theoretical argument is not yet supported, because it relies on a contested redefinition of legitimate peripheral participation and does not reconcile the reported null results with the strong claim about 'transformative contributions.'","major_comments":[{"comment":"The argument rests on identifying designer-scripted VR scenarios with Lave and Wenger's legitimate peripheral participation (LPP, Ref. [13]), but this identification drops the community-of-practice condition that is central to the original theory. In LPP, 'legitimate' means access granted by a community, 'peripheral' means participation in the community's genuine productive activity, and 'participation' entails mutual engagement with other members. The automotive engineer scenario described in Section 2.1 has no other community members, no community-granted legitimacy, and no authentic shared enterprise; it is an individual, simulated task. The paper's paraphrase ('meaningful actions that progressively move you into an expert role') strips LPP of its social and relational content. Because the claim that agency enables 'legitimate participation' is the stated mechanism connecting agency to situated learning, this equivocation is load-bearing. The author should either stop invoking LPP and argue instead for 'authentic task participation' as a distinct construct, or provide a worked-out account of how a virtual community can confer legitimacy and enable peripheral participation in the full Lave-and-Wenger sense.","section":"Section 2.1"},{"comment":"The paper's central claim that VR's educational potential 'rests on how well it is designed for agency' is contradicted, or at least heavily qualified, by the author's own empirical result, stated in Section 2.1: 'My own research finds that increased sense agency does not lead to increased content learning gains [18].' The paper then pivots to self-efficacy, intrinsic motivation, awe, and wonder as the valued outcomes. That pivot may be defensible, but it is not reconciled with the vocabulary of 'situated learning,' which is conventionally tied to learning outcomes and identity transformation. As written, the strong claim that transformative educational potential rests on agency is not established by the cited evidence, which shows only affective or motivational effects. The paper should either explicitly limit the claim to motivational and identity outcomes, or provide a theoretical account of why affective gains should be privileged in a situated-learning framework.","section":"Section 2.1 (and the cited Ref. [18])"},{"comment":"The paper defines 'agency' inconsistently. It opens Section 2 with 'Agency is one's capacity for acting and exerting control,' then contrasts this with 'taking meaningful actions that progressively move you into an expert role,' and later invokes self-determination theory (Ref. [22], 'autonomy to set goals and make decisions'). These are three different constructs (control, expert-role progression, and autonomy). The paper needs to specify which sense of agency is supposed to produce situated learning, because the design principles that follow depend on that choice. If the intended sense is autonomy, then the automotive-engineering scenario, with its scripted 'day in the life,' may not actually provide that kind of agency.","section":"Sections 2 and 2.1"}],"minor_comments":[{"comment":"There are several typos: 'agenct' for 'agency', 'authenticlaly' for 'authentically', and 'Massachuetts' for 'Massachusetts' in the author affiliation. The sentence 'embedded in the context in which occurs' is missing 'it'.","section":"Abstract and Section 1"},{"comment":"The phrase 'a day in the life of an automotive engineers' has a subject-verb agreement error; it should be 'an automotive engineer'.","section":"Section 2.1"},{"comment":"Several references lack full venue information: Ref. [17] is a dissertation and is cited without page details, and Ref. [22] appears to be a journal article without its full bibliographic entry. The reference list formatting should be completed according to ACM style.","section":"References"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick read on McGivney's arXiv:2504.12593. It's a 3-page workshop position paper, and it does one thing well: it pushes back against the default VR-education framing that immersion and interactivity are what matter, and argues that agency—the learner's capacity to make meaningful choices—is the design axis worth attending to. That distinction is worth keeping, and the author's own empirical work (cited as [16] and [18]) showing that interactivity and agency don't always move together gives it some teeth. She's also admirably honest that her own data didn't find agency improving content learning; she doesn't hide that.\n\nWhere it gets wobbly is the bridge to situated learning theory. She states that situated learning requires legitimate participation, then gives an example: giving learners 'control over a day in the life of an automotive engineer' in VR transfers them into 'legitimate participation in real-world tasks.' That's not what Lave and Wenger's legitimate peripheral participation means. In the theory, legitimacy is granted by a community of practice, and participation is peripheral but real engagement with that community's shared enterprise. A scripted VR scenario, with no other members, no practitioner-granted access, and no consequences for the simulated work, doesn't have the social structure the theory requires. I think the stress-test note is right: the central inference from agency to situated learning via LPP rests on that equivocation.\n\nThat doesn't sink the paper as a design proposal. A weaker claim would work: designing VR for agentic role-play in authentic task contexts might support motivated, transferable learning, and the community dimension could be simulated or scaffolded—but that needs to be argued, and citing LPP as the mechanism doesn't get there.\n\nAlso note: this is a position paper, not a study. No new data, no new derivation. That's fine for a workshop, but the transformative claim in Section 2.1 is a prediction, not an established result. The paper itself is honest about that.\n\nBottom line: it's a clear, useful position piece for the VR-learning community, with a real conceptual slip in how it uses LPP. Should a referee spend time on it? For a workshop, yes. For a full journal, I'd want the author to either defend virtual communities as conferring legitimacy or soften the claim to simulated participation. I'd engage with it. I'd cite the interactivity/agency distinction. I'd bring it to a reading group if we were discussing VR learning theory.","headline":"A short, honest position paper that makes a useful point about agency over interactivity in VR learning, but overreaches when it equates scripted VR role-play with Lave and Wenger's legitimate peripheral participation.","tokens_in":5254,"tokens_out":2692,"would_cite":true,"duration_ms":25376,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Virtual reality's educational value depends on designing for learner agency, not just interactivity or immersion.","keywords":["virtual reality","education","agency","situated learning","legitimate peripheral participation","interactivity","learning outcomes","instructional design"],"falsifier":"A controlled study comparing two VR field trips matched for content and interactivity — one giving learners a defined role with consequential decisions, the other allowing free exploration without a role — would falsify the central claim if the role-based version produced no greater self-efficacy, motivation, or transfer to real-world tasks than the free-exploration version.","tokens_in":4313,"feed_emoji":"🥽","tokens_out":7276,"duration_ms":66403,"temperature":0.7,"pith_summary":"This position paper argues that virtual reality's most important contribution to education is not its ability to transport students or let them handle objects, but its capacity to give learners genuine agency: legitimate participation in authentic tasks that situates learning in real-world practice. It claims that the field's focus on presence and interactivity has led designers to equate agency with the freedom to move and manipulate objects, when in fact the two are only loosely connected. The paper warns that if VR is treated as a highly controlled, didactic medium, it will simply reproduce the passive schooling that situated learning theory criticizes. Instead, VR should be designed so learners take meaningful actions that move them from novice toward expert roles, which could bring field-trip-like learning into classrooms at scale.","feed_headline":"Design VR for learner agency, not just interactivity","feed_subtitle":"Position paper: authentic tasks and meaningful choices, more than hand tracking, make VR learning work.","key_machinery":"The conceptual machinery is the distinction between interactivity and agency, with agency understood as 'legitimate peripheral participation' from situated learning theory: taking part in authentic, meaningful tasks that progressively move a novice toward expert status. This redefinition redirects the design goal of VR learning environments from maximizing immersion and hand-level interaction to giving learners consequential roles and decisions within a practice. The empirical anchor is the observation that agency and interactivity are not tightly coupled in VR, so more interactive systems do not automatically produce the kind of agency that supports situated learning.","core_discovery":"The central claim is that VR's potential to make transformative contributions to education rests on how well it is designed for agency, where agency means a learner's capacity to take meaningful actions that progressively move them into an expert role — the kind of legitimate participation that situated learning theory requires. The paper distinguishes this from interactivity, the ability to move through and manipulate a virtual environment, and cites evidence that interactivity does not reliably increase a sense of agency, nor does an increased sense of agency reliably improve content recall. The proposed remedy is to embed learning in authentic tasks and narratives: for example, a VR field trip that situates physics learning in automotive manufacturing, where students run crash simulations and also control a day in the life of an automotive engineer. The paper warns that responding to VR's cognitive load by making environments more restrictive would turn VR into a tool of traditional passive schooling and forfeit its transformative benefits, which include self-efficacy, intrinsic motivation, and a sense of awe.","pith_inferences":["A practical extension: the strongest test of the paper's claim would be a comparison of two matched VR environments, one with a defined role and consequential choices and one with free exploration, measuring whether role-based agency produces greater self-efficacy and real-world transfer.","The paper's emphasis on legitimate participation suggests that collaborative VR, where learners interact with simulated or real expert characters, might satisfy the social dimension of situated learning more fully than the single-user scenarios described.","If the argument holds, VR learning environments should be designed as spaces for judgment and consequence rather than as high-tech lecture halls; that would be a detectable shift in how educational VR is built and assessed."],"forward_implications":["VR learning environments designed for agency should increase learners' self-efficacy and intrinsic motivation even when they show no better content recall than slideshows or video.","Authentic, role-based VR field trips can make situated learning opportunities available to classrooms that lack access to field trips and apprenticeships.","Designers should blend guided instruction with unguided discovery and use narratives that make the consequences of learners' actions visible, rather than simply adding more interactivity.","Evaluations of VR learning should measure motivational, identity, and transfer outcomes alongside content-knowledge gains."],"supporting_citations":[{"why":"supplies the theory of legitimate peripheral participation that the paper's concept of agency is built on.","marker":"[13]"},{"why":"a model equating VR control factors with agency that the paper argues is too narrow.","marker":"[14]"},{"why":"shows that learners can feel agency in less-interactive media like 360-degree video, decoupling agency from interactivity.","marker":"[16]"},{"why":"shows that increased interactivity does not necessarily enhance sense of agency, and agency does not raise content learning gains.","marker":"[18]"},{"why":"an empirical study that links interactivity to increased agency, serving as the contrasting baseline the paper complicates.","marker":"[19]"},{"why":"argues immersive interfaces can bring situated learning to classrooms, supporting the paper's premise that VR can scale authentic experiences.","marker":"[4]"}],"fun_headline_variants":["VR learning needs agency, not just interactivity","Agency, not interactivity, is VR's learning key","Design VR for learner agency to enable situated learning","Situated VR learning hinges on agency, not interactivity","VR's real power: learner agency for authentic tasks"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a designer-created virtual scenario can provide legitimate peripheral participation to a solitary learner, even though the original theory of legitimate peripheral participation involves membership in a real community of practice with social interaction.","fun_headline_variants_meta":{"raw":{"variants":["VR learning needs agency, not just interactivity","Agency, not interactivity, is VR's learning key","Design VR for learner agency to enable situated learning","Situated VR learning hinges on agency, not interactivity","VR's real power: learner agency for authentic tasks"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000151,"raw_usage":{"total_tokens":1139,"prompt_tokens":820,"completion_tokens":319,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":436,"completion_tokens_details":{"reasoning_tokens":242}},"tokens_in":436,"tokens_out":319,"duration_ms":4205,"temperature":1.0,"reasoning_tokens":242,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T12:27:07.526844+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A controlled study comparing two VR field trips matched for content and interactivity — one giving learners a defined role with consequential decisions, the other allowing free exploration without a role — would falsify the central claim if the role-based version produced no greater self-efficacy, motivation, or transfer to real-world tasks than the free-exploration version.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the theory of legitimate peripheral participation that the paper's concept of agency is built on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"shows that learners can feel agency in less-interactive media like 360-degree video, decoupling agency from interactivity."},{"cited_title":"Proceedings of the Purposeful XR Workshop for CHI 2025","cited_arxiv_id":"2504.07475","evidence_quote":"an empirical study that links interactivity to increased agency, serving as the contrasting baseline the paper complicates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"argues immersive interfaces can bring situated learning to classrooms, supporting the paper's premise that VR can scale authentic experiences."}],"review_version":1}