{"id":"c8e7f1f1-1c83-47d9-967b-ab27d6de5f7e","arxiv_id":"2608.01688","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"VR neck rehab games that reward movement with environmental or musical feedback were preferred and rated as more relaxing than score-based games, without hurting exercise accuracy.","lead":"This study compares three virtual reality game designs for neck exercises: a score-based game, an evolving island environment, and a musical interaction. Non-patient users found the experience-based designs more relaxing and engaging, with similar exercise accuracy.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Score-based baseline adds an upbeat music track not present in the experience conditions, so anxiety/UX benefits cannot be attributed to reward mechanism; H1 and H3 rest on a confounded comparison.","rationale":"The reader's weakest assumption—that the three conditions differ only in gamification mechanic—is the most load-bearing concern. Section 3.5.1 explicitly adds 'Funky Dude' to the score-based condition, while the two experience conditions use calming/nature audio, so the comparison confounds reward type with audio style. This directly undermines the abstract's attribution of anxiety/UX benefits to 'immersive environmental and musical feedback' rather than to the score mechanism. I agree with the CONDITIONAL verdict: the paper is exploratory and the authors are transparent about limitations (non-patient sample, no family-wise correction, H1 only partially supported), so it is not worthless; but the specific comparison behind H1/H3 needs a control that isolates the scoring mechanic from the audio. The other candidate concern—generalizing from non-patients to chronic neck pain patients—is real but secondary and already acknowledged in §7.2; it would affect external validity even if the mechanism were cleanly isolated. H2's performance-equivalence claim is more robust because it uses objective kinematic data and remains plausible even with the audio confound. I would not move the verdict to accept or reject; the existing conditional status is appropriate pending a matched-audio replication.","tokens_in":15939,"tokens_out":5565,"duration_ms":63596,"concrete_test":"Run a follow-up within-subjects study crossing reward mechanism (score displayed vs. hidden) with audio style (upbeat 'Funky Dude' vs. relaxing alpha-wave/nature ambience), keeping the virtual island environment otherwise constant. If the score+relaxing-audio condition performs as well as the experience conditions on SUDS shifts, anxiety-reduction preference, and UX items Q2/Q6/Q10, while the score+upbeat-audio condition reproduces the original score-based results, then the reported anxiety/UX differences are attributable to music rather than to the gamification paradigm. A simpler version: add a fourth arm identical to the score-based baseline but with the musical-interaction condition's calm audio; if it becomes indistinguishable from the experience-based conditions on the preference rankings, the mechanism attribution fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central claim is that experience-based environmental/musical feedback, as opposed to score-based gamification, reduces anxiety and improves UX. To support that attribution, the three conditions must differ only in the reward/feedback mechanism. They do not: the score-based baseline uses a distinct upbeat background track, 'Funky Dude', blended with the natural island soundscape (§3.5.1), while the spatiotemporal condition uses the island ambience without that added track (§3.5.2) and the musical condition uses AI-generated 'zen, deep focus, meditation, alpha wave' ambient music (§3.5.3). The score condition therefore differs from the experience conditions in both the scoring mechanic and musical arousal/valence. The significant UX items (Q2 immersion, Q6 perceived muscular benefit, Q10 theme liking) and the preference rankings could all be driven by music preference rather than by implicit vs. explicit rewards. This is not merely hypothetical: 2/15 freeform comments described the score-condition audio as unpleasant, and the musical condition was most often called calming (§6.4). The objective anxiety measures (SUDS, HR, tonic EDA) showed no significant differences, so the anxiety claim rests on preference rankings and non-significant trends—precisely the measures most sensitive to music. The authors also acknowledge not correcting for multiple comparisons across UX items (§7.2), further weakening the few significant differences. H2 (performance equivalence) is less affected by this confound, but H1 and H3 are not cleanly tested.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This exploratory user study compares two experience-oriented VR neck range-of-motion (ROM) exercise designs—Spatiotemporal Progression and Musical Interaction—against a conventional score-based gamification baseline. Twenty non-patient participants completed all three game conditions in randomized order. The authors measure subjective distress (SUDS), physiological arousal (heart rate, tonic EDA), kinematic accuracy, SUS/usability, eleven custom UX Likert items, preference rankings, and freeform comments. They report no statistically significant differences in the primary anxiety outcomes (SUDS, HR, tonic EDA) across conditions; three UX items favored the spatiotemporal condition over the score baseline; preference rankings strongly favored the experience-based designs; and overall exercise performance scores were comparable. The paper concludes that immersive environmental and musical feedback may reduce anxiety and improve user experience without compromising exercise performance, and positions this as an alternative to performance-centric gamification for VR rehabilitation.","tokens_in":16285,"tokens_out":4828,"duration_ms":54681,"significance":"If the central claim were supported, the study would make a useful contribution to VR rehabilitation design by showing that implicit, experience-based rewards can maintain movement quality while improving emotional engagement. The work is novel in applying spatiotemporal environmental progression and generative-AI-driven music to a therapist-designed neck ROM exercise, and it benefits from a randomized within-subjects design, personalized calibration, multi-modal measurement, and transparent reporting of limitations. However, the evidence supporting the anxiety/UX claim is currently weakened by a confound in the baseline audio condition and by uncorrected multiple comparisons. The abstract's statement that the experience-based designs 'show better potential to reduce anxiety and improve user experience' overstates what the objective measures and the corrected analysis support. The performance-equivalence result (H2) is the most defensible contribution as reported.","major_comments":[{"comment":"The baseline score-based condition is not matched to the experience-based conditions in auditory content. Section 3.5.1 states that 'to keep the user engaged in the score-based paradigm, we used the royalty-free track Funky Dude from a free Unity store asset as background music, which is blended with the natural sounds of the environment.' In contrast, the spatiotemporal condition uses the island ambience without this added track, and the musical condition uses AI-generated 'zen, deep focus, meditation, alpha wave' ambient music. Because H1 and H3 compare anxiety and UX against the score baseline, this confound means that the significant UX differences and preference rankings cannot be attributed to the reward mechanism (score vs. implicit) rather than to musical style/valence. The freeform results strengthen this concern: 2/15 participants described the score-condition audio as unpleasa","section":"§3.5.1 vs. §3.5.2–3.5.3"},{"comment":"The UX analysis performs 11 Friedman tests without family-wise correction, and the authors acknowledge this in §7.2. With 11 tests and α=0.05, the expected number of false positives under the global null is about 0.55, so three significant results at p<0.05 is not strong evidence of a systematic effect. The significant post-hoc contrasts (Q2, Q6, Q10) all compare Spatiotemporal Progression to Score-based gamification; Musical Interaction does not reach significance on any of these items. Since these uncorrected UX items are a primary pillar of the H3 claim, the abstract and conclusion should either report family-wise corrected values, apply a false-discovery-rate correction, or explicitly label these as exploratory trends rather than as support for the broad claim that experience-based designs improve UX.","section":"§6.1, §7.2"},{"comment":"The primary anxiety outcomes—SUDS shifts, heart-rate shifts, and tonic EDA shifts—show no statistically significant differences across conditions (p=0.11, p=0.830, p=0.132 respectively). The anxiety-reduction claim therefore rests on preference rankings and non-significant descriptive trends, the very measures most susceptible to the audio confound noted above. The abstract's phrase 'show better potential to reduce anxiety' should be materially weakened, for example to 'non-significant trends and user preferences suggest a possible advantage, but objective measures did not distinguish the paradigms.' As written, the abstract implies a stronger empirical basis than the data provide.","section":"§6.2, Table 1, Fig. 4C"},{"comment":"While final performance scores did not differ significantly, Table 3 shows a statistically significant difference in fixation-phase angular error (χ²=10.30, p=0.006), with Musical Interaction exhibiting a larger median error (0.50°) than Score-based (0.37°). The paper's claim of 'little to no impact on successful performance' should acknowledge this significant error increase, even though final aggregate scores were similar. The discussion does mention the effect and attributes it to drowsiness, but the abstract's blanket claim is stronger than the data warrant.","section":"§6.3, Table 3"}],"minor_comments":[{"comment":"The text refers to 'Table 5.2' when reporting physiological results; this should be 'Table 1'.","section":"§7.1.1"},{"comment":"The text says 'see Fig. 1B' for the spatiotemporal progression paradigm, but Fig. 1B is the score-based design; the correct reference is Fig. 1C.","section":"§3.5.2"},{"comment":"The methods state that medians of pre–post differences are evaluated, but Table 1 reports mean±SD. Please clarify which statistics are reported or reconcile the description.","section":"§5.2"},{"comment":"Minor typo: 'emtion regulation' should be 'emotion regulation'.","section":"§3.5.3"},{"comment":"The scoring exponent of 8 in Eqs. (1)–(2) is described as empirically determined; a brief note on how this value was selected and whether the main results are robust to reasonable alternatives would improve reproducibility.","section":"§3.4.2"},{"comment":"The figure caption and the text are inconsistent about the number of subpanels: the caption lists B–E, but the text refers to 'Fig. 3B-3C' when discussing preferences. Please align the citation with the actual panels.","section":"Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The confound in the baseline audio is the central issue: it directly undermines the attribution of anxiety/UX differences to the reward mechanism. The authors could address this by substantially revising their claims and adding a sensitivity analysis or a control condition, but the latter would require new data collection. Given the exploratory framing and the acknowledged limitations, I do not think rejection is warranted; however, the abstract and conclusions need to be toned down, and the empirical claims should be restricted to what the design can support. I would also check whether the journal's scope welcomes a purely empirical design-comparison study with this sample size; if so, the revised paper could be a reasonable fit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a solid, well-described exploratory user study with two genuinely new interaction paradigms for VR neck rehab, and the performance-equivalence result is credible. But the anxiety/UX claim rests on a confounded baseline and mostly non-significant objective measures, so the abstract's \"better potential\" should be read as a hypothesis worth testing, not a demonstrated benefit.\n\nWhat's new: the spatiotemporal progression and musical interaction designs aren't just score/token variations; they tie movement to environment and music in ways prior neck rehab work hasn't. The theory grounding (SRT, embodied cognition, music emotion) is sensible, and the implementation is detailed enough to replicate. The paper is transparent about limitations, including the lack of multiple-comparison correction and the non-patient sample. H2—equivalent performance—is the strongest result, well supported by both final scores and angular errors.\n\nSoft spots: the biggest is the audio confound. The score-based baseline adds an upbeat \"Funky Dude\" track, while the experience conditions use either island ambience or alpha-wave music. Any difference in anxiety, immersion, theme liking, or preference could be driven by music rather than by the reward mechanism. The paper frames H1/H3 as comparing score-based vs experience-based designs, so one could argue the comparison is at the holistic design level, but the discussion repeatedly attributes effects to the score display. That attribution isn't clean. Second, the objective anxiety measures (SUDS, HR, tonic EDA) are all non-significant; the significant UX differences are three Likert items out of 11 with no family-wise correction, though this is acknowledged. Preference rankings are subjective and susceptible to demand characteristics. None of this makes the paper bad—it's exploratory—but it means the central claim is over-stated relative to the evidence.\n\nWho it's for: people designing VR rehabilitation or studying gamification and affect will get value from the design ideas and the warning about score display. It deserves a serious referee, and with a matched-audio control and a patient sample it could become a solid contribution. As it stands, I'd read it as a useful design exploration, not a demonstration.","headline":"A competent exploratory study with genuinely new VR rehab designs, but the anxiety/UX claim is undermined by an audio confound and non-significant objective measures; read it as a design exploration, not a demonstrated benefit.","tokens_in":16749,"tokens_out":2844,"would_cite":false,"duration_ms":34029,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper reports that replacing score-based rewards with immersive environmental or musical feedback in VR neck rehabilitation exercises can lower anxiety and improve user experience without losing movement accuracy.","keywords":["virtual reality rehabilitation","chronic neck pain","range-of-motion exercise","gamification","anxiety reduction","multisensory feedback","user experience","implicit rewards"],"falsifier":"A replication that holds the audio track identical across all three conditions—or swaps the upbeat track into the experience-based conditions—and re-measures SUDS, heart rate, and tonic electrodermal activity would settle whether the reward mechanism or the music caused the observed differences.","tokens_in":15883,"feed_emoji":"🧘","tokens_out":6676,"duration_ms":71243,"temperature":0.7,"pith_summary":"This paper argues that the usual way of motivating rehabilitation exercises in virtual reality—a running score with encouragement messages—may add evaluative stress that works against therapy. It proposes two alternatives: Spatiotemporal Progression, where completing each neck-movement segment changes the virtual island's time of day, weather, and viewpoint, and Musical Interaction, where neck movements trigger bell-like notes over AI-generated ambient music. In a within-subject study of 20 healthy adults, both experience-based designs were preferred, judged more relaxing, and rated as more likely to be used long-term than the score-based baseline, with no statistically significant loss in movement accuracy. The authors conclude that implicit, experience-based rewards are a viable alternative to performance-centric gamification in VR neck rehabilitation and should be tested in chronic neck pain patients.","feed_headline":"No-score VR neck rehab relaxes users, keeps movement accuracy","feed_subtitle":"In 20 users, scenery- and music-based rewards beat a point score on anxiety and preference, with equal accuracy.","key_machinery":"The load-bearing mechanism is the implicit reward loop: the user's head-tracked movement itself produces experiential feedback—environmental/spatiotemporal transformation or musical notes—rather than awarding points after the fact. All three conditions share the same personalized range-of-motion protocol, the same tropical island environment, and the same target-tracking visual guidance; only the feedback channel differs. The score-based baseline converts angular tracking error into an explicit accumulating score with encouragement text, while the experience-based conditions suppress the live score and replace it with movement-triggered scenic or auditory events, with the final score shown o","core_discovery":"The paper's central claim is that therapeutic neck movements can be coupled to multisensory experiential rewards—changing the virtual environment or layering musical notes on ambient sound—and that these implicit reward systems perform as well as explicit score feedback on movement quality while creating a calmer, more preferred experience. The authors designed a range-of-motion exercise in a tropical island VR scene with personalized calibration of six neck positions. The score-based condition displayed segment scores and encouragement text over an upbeat track; the spatiotemporal condition altered time of day, weather, and location after each segment; the musical condition played bell-like","pith_inferences":["The comparison is partly confounded by audio style: the score-based baseline uses an upbeat track while the experience-based conditions use ambient nature sound or alpha-wave music, so the cleanest test of the paper's claim would hold music constant across conditions.","If the anxiety-reduction effect is driven mainly by the absence of an evaluative score rather than by the multisensory rewards themselves, a neutral no-score condition with a matched soundscape should produce similar benefits.","For patients with kinesiophobia, the spatiotemporal progression could double as graded exposure—the environment visibly moves forward as the patient holds a movement—which the paper does not develop as a therapeutic direction.","The small but significant higher fixation error in musical interaction hints that relaxation can tip into drowsiness; adding an alertness measure, such as EEG or reaction-time probes, would test whether the relaxing reward needs a minimum-arousal floor to preserve precision."],"forward_implications":["VR rehabilitation designers can drop persistent score displays without sacrificing measured movement accuracy, freeing visual and auditory channels for emotion regulation.","Anxiety reduction and engagement become first-class design goals alongside biomechanical performance, which may improve adherence in chronic neck pain therapy.","The success of the two experience-based prototypes supports a follow-up trial with chronic neck pain patients; if replicated, experience-based rewards could be used in home-based VR rehabilitation.","Delaying performance feedback until the end of the exercise (as in the spatiotemporal condition) appears viable, suggesting continuous score display is not necessary for accurate movement.","Combining spatiotemporal and musical feedback may add anxiety-reduction benefits, but the possible drowsiness and cognitive load would need dedicated testing."],"supporting_citations":[{"why":"Supplies evidence that score-based gamification can induce evaluative stress, motivating the anxiety-reduction comparison.","marker":"[38]"},{"why":"Provides empirical basis that nature experience improves mental well-being, grounding the tropical-island environment.","marker":"[5]"},{"why":"Stress Recovery Theory, the stated theoretical basis for the natural-environment spatiotemporal progression.","marker":"[32]"},{"why":"The attentional-gate model invoked to justify accelerated time-of-day transitions as stress-reducing.","marker":"[41]"},{"why":"Embodied cognition, cited to justify movement-triggered environmental progression as an implicit reward.","marker":"[4]"},{"why":"Auditory-motor integration, grounding the movement-triggered musical notes design.","marker":"[11]"},{"why":"Music and emotion theory, grounding the relaxing effect of the musical stimuli.","marker":"[18]"},{"why":"Systematic review of VR relaxation interventions, supporting the choice of anxiety outcome measures and expected effects.","marker":"[27]"},{"why":"Validation of the EmotiBit wearable sensor used for heart rate and electrodermal activity measurements.","marker":"[24]"}],"fun_headline_variants":["VR neck rehab trades scores for scenery, cuts anxiety","Music and virtual weather beat points in neck rehab","No-score VR rehab eases anxiety, keeps motion accuracy","Sensory rewards calm VR neck rehab without losing accuracy"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The three conditions are assumed to differ only in the gamification mechanic, but the score-based baseline used an upbeat music track while the experience conditions used ambient nature or alpha-wave audio, so music style, not reward type, could account for the reported anxiety and preference differences.","fun_headline_variants_meta":{"raw":{"variants":["VR neck rehab trades scores for scenery, cuts anxiety","Music and virtual weather beat points in neck rehab","No-score VR rehab eases anxiety, keeps motion accuracy","Sensory rewards calm VR neck rehab without losing accuracy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000351,"raw_usage":{"total_tokens":1766,"prompt_tokens":773,"completion_tokens":993,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":517,"completion_tokens_details":{"reasoning_tokens":930}},"tokens_in":517,"tokens_out":993,"duration_ms":9103,"temperature":1.0,"reasoning_tokens":930,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T22:50:34.661582+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A replication that holds the audio track identical across all three conditions—or swaps the upbeat track into the experience-based conditions—and re-measures SUDS, heart rate, and tonic electrodermal activity would settle whether the reward mechanism or the music caused the observed differences.","supporting_citations":[{"cited_title":"Yang and D","cited_arxiv_id":null,"evidence_quote":"Supplies evidence that score-based gamification can induce evaluative stress, motivating the anxiety-reduction comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides empirical basis that nature experience improves mental well-being, grounding the tropical-island environment."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Stress Recovery Theory, the stated theoretical basis for the natural-environment spatiotemporal progression."},{"cited_title":"Zakay, R","cited_arxiv_id":null,"evidence_quote":"The attentional-gate model invoked to justify accelerated time-of-day transitions as stress-reducing."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Embodied cognition, cited to justify movement-triggered environmental progression as an implicit reward."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Auditory-motor integration, grounding the movement-triggered musical notes design."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Music and emotion theory, grounding the relaxing effect of the musical stimuli."},{"cited_title":"Riches, P","cited_arxiv_id":null,"evidence_quote":"Systematic review of VR relaxation interventions, supporting the choice of anxiety outcome measures and expected effects."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Validation of the EmotiBit wearable sensor used for heart rate and electrodermal activity measurements."}],"review_version":1}