{"id":"33d05829-15f5-46a8-9dd2-7aae04159d71","arxiv_id":"2411.10230","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Customizable AR table decor and lighting in a simulated restaurant improved user experience, psychological ownership, and perceived waiting time relative to fixed defaults.","lead":"In a VR simulation of a restaurant, letting people customize augmented table decorations and lighting improved their mood, sense of control, ownership, and perceived waiting time compared with fixed defaults. The study is an early concept test for AR glasses in restaurants, though the small sample and confounded design limit how strongly the effects can be attributed to customization.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The customizable condition adds an active interaction task and moderator prompting that the default condition lacks (Section II.D); the reported benefits may therefore reflect activity, not customization.","rationale":"The reader's weakest_assumption correctly identifies the activity confound as the key load-bearing threat. This concern does not move the verdict because CONDITIONAL already reflects that the causal claim requires an active-control condition (or equivalent) before acceptance. I see no independent support that would overturn the reader's assessment; the empirical measures are real, but the causal interpretation is conditional on addressing this confound. The concrete test above would settle whether the customization option itself, rather than the act of interacting, drives the observed benefits.","tokens_in":8892,"tokens_out":5636,"duration_ms":55726,"concrete_test":"Run a three-condition within-subjects comparison in the same simulated restaurant: (1) passive default, as in the paper; (2) customizable, as in the paper; and (3) active-default, in which participants are instructed to select and confirm the default options (e.g., red roses, candle, medium white lighting) using the same UI and moderator prompting. Compare PWT, SAM, UEQ-S, POQ, and SAQ between (2) and (3). If (3) produces the same improvements over (1) as (2) does, then the customization-specific attribution is falsified; if (2) significantly beats (3), the customization effect is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central causal claim—that offering customizable AR table setup and lighting improves UX, perceived waiting time, psychological ownership, and acceptability—is not identified because the two conditions differ on more than customization. Per Section II.D, in customizable conditions participants are asked after 3 minutes to 'set up the table or lighting as they liked best' and to tell the moderator when finished, requiring active hand-tracking manipulation and social interaction. In default conditions, no such instruction is given; participants wait passively for the full 4 minutes. Thus the paired comparisons in Tables I and II contrast customizable conditions (customization + interaction + task + moderator prompting) against default conditions (passive viewing). Any effect could be driven by the opportunity to act, the sense of agency, demand characteristics, or the fact that the final minute is filled with a task, which directly biases perceived waiting time. The Discussion itself attributes the PWT effect to 'engaging in AR activities', which is exactly the confound. Without an active control that holds interaction and prompting constant while removing choice, the claim that customization per se is beneficial is unsupported. Secondary issues (unadjusted multiple t-tests, unreported order or counterbalancing) are present but secondary to this systematic difference in the manipulation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a controlled laboratory study (N=19) using a within-subject design in which participants experienced four conditions in a simulated restaurant rendered via a 360-degree VR scene: default table setup, customizable table setup, default lighting, and customizable lighting. The authors measured perceived waiting time, SAM (valence, arousal, dominance), UEQ-S (pragmatic, hedonic, overall), psychological ownership (POQ), and social acceptability (SAQ), and used paired samples t-tests to compare default versus customizable conditions. They report broad statistically significant benefits of customization for UX, emotional state, psychological ownership, and acceptability, and a reduction in perceived waiting time for customizable table setup. The central claim is that offering customizable augmented table setup and lighting improves the restaurant user experience.","tokens_in":9095,"tokens_out":2946,"duration_ms":29498,"significance":"The concept is timely, the use of VR to simulate AR is practical and resource-efficient, and the measurement suite is broad and well-established. If the causal interpretation were valid, the findings would be directly actionable for hospitality AR design and would motivate further field studies. The paper also reports effect sizes for all tests, which is a strength. However, the central causal claim is currently undermined by a systematic confound in the procedure: the customizable conditions differ from the default conditions in more than customization alone. Consequently, the significance of the paper depends on whether the authors can disentangle customization from the active interaction task, which is a fixable but load-bearing issue.","major_comments":[{"comment":"The customizable and default conditions differ not only in customization but also in whether participants performed an active interaction task during the final portion of the wait and whether the moderator prompted them. In customizable conditions, after 3 minutes the participant was asked to 'set up the table or lighting as they liked best' and to tell the moderator when finished, whereas in default conditions no such instruction was given and participants waited passively for the full 4 minutes. Therefore, the paired t-tests in Tables I and II contrast 'customization plus activity plus moderator prompting' against 'passive waiting'. This confound is load-bearing for the central claim that customization itself improves UX, perceived waiting time, and psychological ownership. The Discussion in Section IV explicitly attributes the perceived-waiting-time benefit to 'engaging in AR activities', which is exactly the confounded factor, and the limitations paragraph does not acknowledge this confound. An active control condition that holds interaction and prompting constant while removing choice is necessary to support the causal interpretation.","section":"Section II.D, Tables I and II"},{"comment":"The manuscript reports 12 significant paired t-tests for table setup and 8 significant tests for lighting, all judged against a p<0.05 threshold, without any correction for multiple comparisons or a pre-registered analysis plan. Because the dependent variables are correlated constructs drawn from overlapping questionnaires (UEQ-S subscales, SAM dimensions, POQ subscales, SAQ dimensions), the familywise error rate is high, and borderline effects (e.g., SAQ Interaction p=.049 and SAQ Safety p=.031 in Table I) may arise from chance. The authors should report adjusted p-values or explicitly frame the analysis as exploratory.","section":"Section III.A-III.B"},{"comment":"The procedure does not state whether the order of the four within-subject conditions was counterbalanced or randomized, nor whether the moderator was blind to the research questions. With N=19 and repeated measures, order and carryover effects can materially change results, and the paper reports no check or model term for order. This omission is load-bearing for the claim that the observed differences are attributable to the manipulation rather than to presentation order, fatigue, or learning effects.","section":"Section II.D"}],"minor_comments":[{"comment":"The word 'Valence' appears as 'V alence' in the abstract; please remove the stray space.","section":"Abstract"},{"comment":"The column header 't d f' appears to be a typographical error for 'df'; also consider labeling the effect-size column explicitly as Cohen's d.","section":"Table I"},{"comment":"The exact wording of the moderator's instruction in the customizable conditions is not reported; given the confound discussed above, including the verbatim script in a supplementary appendix would improve transparency.","section":"Section II.D"},{"comment":"The limitations paragraph mentions fatigue and the need to imagine other people, but it does not mention the lack of an active control condition; adding this as an explicit limitation would be helpful.","section":"Section IV"}],"recommendation":"major_revision","confidential_remarks":"The systematic confound in Section II.D is serious enough that the current manuscript cannot support its central causal claim, but it is fixable with an active control condition. The multiple-comparison issue and the missing counterbalancing details are also important and should be addressed. I believe the paper is within scope for an HCI venue and the simulation approach is valuable; with a revised design and more cautious causal language it could become a useful contribution. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this is a clearly written, small lab study of AR-based table setup and lighting customization in a simulated restaurant, and the headline results are plausible but the causal claim about customization is not supported by the design. The customizable conditions include an active interaction task (and moderator prompting) that the default conditions lack, so the comparison is customization-plus-activity versus passive viewing.\n\nWhat it does well: the concept model is sensible, the questionnaires are standard (SAM, UEQ-S, POQ, SAQ), and the authors report effect sizes and transparent procedures. Using VR to simulate AR is a legitimate method with prior support. The data are internally consistent—the significant effects are large, in the expected direction, and spread across varied measures, which argues against a pure measurement artifact.\n\nThe main soft spot is the confound in Section II.D. In the customizable conditions, after 3 minutes participants are asked to set up the table or lighting as they like and tell the moderator when done. In the default conditions they just wait the full 4 minutes. So the comparisons in Tables I and II contrast customization plus active manipulation plus social prompting against passive waiting. The perceived-waiting-time result is especially likely to be driven by having a task to fill the final minute, and the paper's own Discussion attributes the effect to 'engaging in AR activities'—which is exactly the confound. An active control that holds interaction constant while removing choice would be needed to say anything about customization per se.\n\nSecondary issues: about a dozen dependent variables tested with no multiple-comparison correction, though several p-values are so small they'd survive most corrections. The order of conditions is not described, so carryover effects are unknown. And n=19 is small, though acceptable for an exploratory within-subjects study.\n\nNone of this makes the paper worthless. It is a reasonable initial exploration of a worthwhile research question, and the pattern of results is suggestive. But the central interpretation needs to be softened substantially or the design needs an active control. For peer review, I'd send it to a serious referee, with the expectation of heavy revision.\n\nWho's it for? Researchers in AR/VR and hospitality UX who want a starting point for designing interactive customization experiences. I probably wouldn't cite it as evidence for the benefits of customization, but it could be cited as a pilot.\n\nBest.","headline":"A clean, small VR/AR restaurant study with a real confound between customization and active interaction; the causal claim about customization is unsupported, but it is a legitimate exploratory pilot that deserves review with revision.","tokens_in":9616,"tokens_out":2504,"would_cite":false,"duration_ms":24613,"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":"Customizing augmented table decor and lighting in a simulated restaurant improves self-reported experience.","keywords":["augmented reality","AR glasses","restaurant experience","customization","user experience","psychological ownership","social acceptability","perceived waiting time"],"falsifier":"Give a third group the same final-minute task of touching, selecting, and moving the virtual objects through the same hand-tracked interface, but without any effect on the appearance (a sham customization control); if their perceived waiting time, UX, ownership, and acceptability scores match the customizable conditions, the effects are due to activity rather than customization, and if they match the default condition the customization explanation is supported.","tokens_in":8699,"feed_emoji":"🍽️","tokens_out":7453,"duration_ms":70509,"temperature":0.7,"pith_summary":"This paper tries to show that letting restaurant guests customize augmented-reality elements of their table and lighting improves the experience, compared with passively viewing a fixed AR setting. The authors built a restaurant scenario inside virtual reality and asked 19 participants to wait for an appetizer under default and customizable conditions, measuring perceived waiting time, emotional state, user experience quality, psychological ownership, and social acceptability. Statistically significant improvements appeared for both customizable table setup and customizable lighting on pragmatic and hedonic UX, pleasure, dominance, psychological ownership, and affect; customizable table setup additionally reduced perceived waiting time and increased arousal, while table-setup customization also lowered perceived safety. If the effect is real, restaurateurs and AR designers have a concrete reason to offer diners control over their immediate environment, and a method for testing such concepts before deployment.","feed_headline":"Custom AR table decor and lighting improve the dining experience","feed_subtitle":"A VR restaurant experiment finds customizable settings boost emotional quality, ownership, and acceptability.","key_machinery":"The central mechanism is the within-subjects pairwise comparison between a default and a customizable variant of each of two manipulable dimensions, table setup and lighting, all presented through an AR-in-VR simulation of a restaurant. Customization was implemented as an interactive tabletop UI operated by hand tracking, letting participants choose flower type and color, table-light type, and lighting intensity and color. The default conditions presented a fixed arrangement (three red roses and a candle; medium white lighting). Statistical significance was assessed with paired samples t-tests on the questionnaire scores.","core_discovery":"The central claim is that customizing an augmented table setup and lighting in a simulated restaurant yields broad self-reported benefits over default presentation. The authors report that customizable table setup significantly reduced perceived waiting time and improved SAM valence, arousal, and dominance; UEQ-S pragmatic, hedonic, and overall quality; POQ psychological ownership and affect; and SAQ interaction and isolation acceptability, while SAQ safety was rated significantly lower. Customizable lighting significantly improved SAM valence and dominance; UEQ-S pragmatic, hedonic, and overall quality; POQ psychological ownership and affect; and SAQ interaction acceptability, with no significant effect on perceived waiting time, arousal, isolation, or safety. These results are offered as evidence that these variables are worth considering for AR applications in restaurants, especially when offering customizable augmented table setup and lighting.","pith_inferences":["Inference: The results do not separate customization from the act of actively manipulating the interface, because only customizable conditions asked participants to interact with the objects during the final minute; an active-control condition with the same hand gestures but without choosing would isolate the customization contribution.","Inference: The large standardized effects on psychological ownership, especially for lighting, suggest that ownership may be a mediator of the UX and acceptability gains, but the study does not test mediation; a path analysis or larger replication could.","Inference: The lowered safety rating for table-setup customization points to a design tension for real restaurants: interactive AR decor may be engaging but could distract diners from physical hazards; future work could test gaze-guided or voice-activated customization that requires less motor attention.","Inference: Because the simulated environment used a 360-degree VR image rather than a real physical restaurant, transferring these results to actual AR glasses requires field testing; the paper's AR-in-VR methodology is a reasonable first filter but not a guarantee of real-world equivalence."],"forward_implications":["If the effect is real, a restaurant offering AR glasses with editable table decor could shorten the subjectively experienced wait for food, because the table-setup customization condition had significantly lower perceived waiting time.","Customizable augmented lighting is a broad UX lever, improving pragmatic and hedonic quality, emotional valence and dominance, and psychological ownership, while leaving perceived waiting time and arousal statistically unchanged.","Psychological ownership and emotional affect toward restaurant elements rise when guests choose their own augmented decor or lighting, which the authors connect to prior work linking ownership with customer identification and willingness to pay more.","Active AR customization is rated more socially acceptable than passive AR viewing on interaction and, for table setup, isolation dimensions, but table-setup customization lowered perceived safety, so interaction design must address distraction.","The paper's stated next steps are to test the combined table-setup-plus-lighting condition, search for a threshold of choice overload, and study shared and social dining contexts; all remain necessary before generalizing to real restaurants."],"supporting_citations":[{"why":"Supplies the method of evaluating AR usability and user experience inside a VR simulation, which this study uses to set up the simulated restaurant.","marker":"[9]"},{"why":"Pilot work showing wearable AR interaction can be prototyped and tested in a virtual environment, supporting the AR-in-VR apparatus.","marker":"[10]"},{"why":"Establishes that waiting in a VR restaurant and environmental factors affect customer emotions and approach-avoidance, motivating the perceived-waiting-time measure.","marker":"[15]"},{"why":"Provides the social acceptability dimensions and SAQ instrument that the study adapts for AR interaction, isolation, and safety ratings.","marker":"[20]"},{"why":"Defines psychological ownership in shared augmented reality and supplies the adapted POQ scale used to measure ownership and affect.","marker":"[28]"},{"why":"A systematic review of perceived waiting time that frames the construct and supports its relevance to service evaluation.","marker":"[29]"},{"why":"Supplies the Self-Assessment Manikin used to measure valence, arousal, and dominance.","marker":"[36]"},{"why":"Supplies the UEQ-S short questionnaire used to measure pragmatic, hedonic, and overall user experience quality.","marker":"[37]"}],"fun_headline_variants":["AR table and light choices boost dining UX","Custom AR settings win in restaurant study","Personalized AR ambience improves dining experience","Augmented table and lighting customization enhances UX"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the benefits come from offering customization rather than from the extra activity required in the customizable conditions, because participants in those conditions actively selected and adjusted virtual objects during the final minute of the wait while default-condition participants simply waited; if activity or task novelty drives the improvements, the customization-specific conclusion is not supported.","fun_headline_variants_meta":{"raw":{"variants":["AR table and light choices boost dining UX","Custom AR settings win in restaurant study","Personalized AR ambience improves dining experience","Augmented table and lighting customization enhances UX"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000209,"raw_usage":{"total_tokens":1389,"prompt_tokens":906,"completion_tokens":483,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":429}},"tokens_in":522,"tokens_out":483,"duration_ms":5815,"temperature":1.0,"reasoning_tokens":429,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T19:49:22.721070+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Give a third group the same final-minute task of touching, selecting, and moving the virtual objects through the same hand-tracked interface, but without any effect on the appearance (a sham customization control); if their perceived waiting time, UX, ownership, and acceptability scores match the customizable conditions, the effects are due to activity rather than customization, and if they match the default condition the customization explanation is supported.","supporting_citations":[{"cited_title":"Evaluating usability and user experience of AR applications in VR simulation,","cited_arxiv_id":null,"evidence_quote":"Supplies the method of evaluating AR usability and user experience inside a VR simulation, which this study uses to set up the simulated restaurant."},{"cited_title":"A prototyping method to simulate wearable augmented reality interaction in a virtual environment - a pilot study,","cited_arxiv_id":null,"evidence_quote":"Pilot work showing wearable AR interaction can be prototyped and tested in a virtual environment, supporting the AR-in-VR apparatus."},{"cited_title":"Desired privacy and the impact of crowding on customer emotions and approach-avoidance responses: Waiting in a virtual reality restaurant,","cited_arxiv_id":null,"evidence_quote":"Establishes that waiting in a VR restaurant and environmental factors affect customer emotions and approach-avoidance, motivating the perceived-waiting-time measure."},{"cited_title":"Influence of interactivity and social environments on user experience and social acceptability in virtual reality,","cited_arxiv_id":null,"evidence_quote":"Provides the social acceptability dimensions and SAQ instrument that the study adapts for AR interaction, isolation, and safety ratings."},{"cited_title":"What’s mine is a hologram? How shared augmented reality augments psychological ownership,","cited_arxiv_id":null,"evidence_quote":"Defines psychological ownership in shared augmented reality and supplies the adapted POQ scale used to measure ownership and affect."},{"cited_title":"Perceived waiting time and waiting satisfaction: a systematic literature review,","cited_arxiv_id":null,"evidence_quote":"A systematic review of perceived waiting time that frames the construct and supports its relevance to service evaluation."},{"cited_title":"Measuring emotion: the self-assessment manikin and the semantic differential,","cited_arxiv_id":null,"evidence_quote":"Supplies the Self-Assessment Manikin used to measure valence, arousal, and dominance."},{"cited_title":"Schrepp, User Experience Questionnaire Handbook , 2015","cited_arxiv_id":null,"evidence_quote":"Supplies the UEQ-S short questionnaire used to measure pragmatic, hedonic, and overall user experience quality."}],"review_version":1}