REVIEW 4 major objections 4 minor 20 references
Hands vs. Controllers: Comparing User Interactions in Virtual Reality Shopping Environments
T0 review · 4 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash
Pith's one-line read This paper claims that in a virtual supermarket, hand-tracking and controllers score similarly on usability, yet 75% of 40 participants prefer hand-based interaction for its natural, immersive feel, while controllers win on precision and…
desk verdict The precision claim is asserted, not measured; a clean but statistically incomplete VR shopping study with reference-list red flags. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central machinery is a within-subjects experimental design in which each participant performs two cereal-selection tasks in both interaction conditions, with virtual hands rendered identically for hand tracking and controller input to isolate the input modality. The study measures outcomes through the System Usability Scale, the User Experience Questionnaire Short, the Affinity for Technology Interaction scale, and custom Likert items on preference and ease of use, then compares these scores and preference distributions to map the trade-off between immersion and precision.
What would settle it
Run a within-subjects VR shopping study with a larger product set, aisles to navigate, and a time limit, and measure task completion time, error rate, and grasping attempts; if controllers show a clear performance advantage while hand tracking remains preferred, the trade-off stands, but if precision differences disappear or reverse under realistic load, the paper's central contrast would be falsified.
Extended reading notes
Core claim
In a within-subjects experiment of 40 participants selecting cereal packages in a simulated supermarket, the paper reports that System Usability Scale scores were comparable (hand-based M = 71.38, controller M = 68.88) and pragmatic quality ratings were nearly identical, while hedonic quality slightly favored hand tracking (M = 2.00 vs. 1.94). Despite this statistical parity, 75% of participants preferred hand-based interaction because it felt natural and immersive, even though it required repeated grasping attempts and lacked haptic feedback; controllers provided smoother, more accurate manipulation. The paper concludes that neither method dominates, and that VR shopping platforms should support customizable interaction modes rather than assume one input method fits all tasks.
Load-bearing premise
The load-bearing premise is that selecting three cereal packages from one shelf in a laboratory is a representative virtual shopping task, so the observed balance between immersion and precision generalizes to real shopping with many products, navigation, and time pressure.
Editorial extensions
If this is right
- VR shopping applications should offer both hand-tracking and controller modes, defaulting to hands for exploratory browsing but allowing controllers for tasks like checking product details or precise placement.
- Because hand tracking's main weakness is lack of haptic feedback, adding even simple vibrotactile or proxy-based haptics could close the usability gap without sacrificing immersion.
- The even split in perceived ease of use (48% each) suggests that usability judgments depend on prior device familiarity, so training or onboarding may matter more than the input method itself.
- The gender difference observed (female participants found hands easier, male participants preferred controllers) implies that user populations with different gaming or controller experience will respond differently to VR shopping interfaces.
- If both input methods perform comparably on standardized usability measures, then preference, not performance, becomes the deciding factor, pushing designers to focus on user choice and adjustable settings.
Reading between the lines
- The following inferences extend beyond the paper's explicit claims and are the editor's own. First
- the paper's minimal task (selecting three cereal packages) likely suppresses the precision advantage of controllers
- with more items
- navigation
- or time pressure
- the performance gap would probably widen
- so the reported parity may be upper-bound for hand tracking. Second
- since controllers in this study also had no haptic feedback
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports a within-subjects VR shopping experiment with 40 participants comparing hand-tracking and controller-based interaction for selecting cereal packages from a virtual shelf. It collects demographic data, ATI, SUS, UEQ-S, and custom Likert/preference questions, and reports that 75% of participants preferred hand-based interaction for its immersive feel, while usability scores were similar between conditions. The paper concludes that hand interaction is more natural and immersive but that controllers offer greater precision and reliability, making them more suitable for fine-motor tasks.
Significance. If the precision/reliability trade-off were empirically supported, the finding would offer practical guidance for VR shopping and similar manipulation tasks. The study uses a sound within-subjects design, randomized condition order, identical virtual hand rendering across conditions, and validated questionnaires (SUS, UEQ-S, ATI), which are strengths. However, the paper reports no objective task-performance data and no inferential statistics, so the core claim that controllers 'offered greater precision and reliability' is not evidenced by the reported experiment. The descriptive preference result is clear, but the central trade-off conclusion is ungrounded as presented.
major comments (4)
- [Section 3 (Results and Discussion)] The statement that UEQ and SUS differences 'were not statistically significant' is unsupported: no test statistic, p-value, confidence interval, or standard deviation is reported anywhere in the manuscript. Without inferential statistics, the reader cannot determine whether the near-identical means indicate real similarity or merely underpowered measurement. This is load-bearing because the paper's argument about comparable usability between hands and controllers rests on these numbers.
- [Section 2 (Methods) and Section 3 (Results and Discussion)] The paper claims that controller-based interaction 'offered greater precision and reliability' and that hand-based interaction 'often resulted in multiple attempts to grasp and manipulate objects, causing frustration and inefficiencies,' yet no objective task-performance variables (e.g., grasp attempts, completion time, error count, placement accuracy) are defined in Section 2 or reported in Section 3. The precision claim is supported only by qualitative assertions and questionnaire means, which are not measures of actual task performance. This is a central, load-bearing part of the conclusion and is empirically ungrounded.
- [Section 3 (Results and Discussion)] The 75% preference result and the 48%/48% ease-of-use split are presented without any statistical test (e.g., a binomial test against a 50% baseline) and without confidence intervals. Additionally, the gender-related observation ('Female participants found hand-based interaction easier to use, while male participants preferred controllers') is based on unspecified subgroup counts and no interaction test. Given the small total sample (40) and missing responses (39 and 38 for the custom questions), these claims should at least be accompanied by exact subgroup sizes and inferential tests or be explicitly labeled as exploratory.
- [Section 2 (Methods) and Section 4 (Conclusion)] The task design—two trials of selecting three cereal packages from a single shelf—does not appear to require the 'fine motor skills' invoked in the abstract and conclusion. The paper generalizes from this minimal task to a broad precision/immersion trade-off in VR shopping, but the chosen task lacks the navigation, product variety, time pressure, and fine-grained manipulation demands of realistic shopping. The authors themselves call for 'more complex tasks' in future work, which is appropriate, but this limits the external validity of the current central claim.
minor comments (4)
- [Section 2 (Methods)] The experimental instructions for the two tasks ('healthy' versus 'tasty' cereal selection) are not specified beyond these category labels; future replication would benefit from a description of how 'healthy' and 'tasty' were operationalized (e.g., package labels, participant judgment) and whether participants received any training or practice trials.
- [Figure 1] The figure caption and axes are incomplete: it does not state what the proportions represent, how missing responses are handled, or the number of participants per gender group. The reader cannot assess the stability of the 75% preference or the gender split without this information.
- [References] Reference [17] appears to duplicate reference [5] (same title, same journal and page), and reference [8] shares the same journal/issue/page as [17] but with a different title; reference [7] has a DOI that may not correspond to a real article. The reference list needs to be checked and corrected for these apparent errors.
- [Section 3 (Results and Discussion)] 'Ease of use was evenly split, with 48% of participants favoring each method' is confusing because only 38 respondents answered this question; 48% of 38 is approximately 18, so a 48/48 split leaves 4% unaccounted for. Please clarify whether there was a 'no preference' option or how missing data were treated.
Circularity Check
No circularity: preferences and usability scores are measured directly, not derived from the conclusions.
full rationale
This paper is an empirical within-subjects usability study with no model fitting, no derived equations, and no prediction constructed from inputs. The central claims—75% preference for hand-based interaction, comparable SUS and UEQ-S scores, and an even split on ease of use—are direct questionnaire and self-report outcomes, not quantities engineered to match a target result. The statement that controller-based interaction 'offered greater precision and reliability' is an interpretive summary of qualitative observations and prior literature, not a variable defined in terms of the conclusion; whether it is adequately measured is a validity/correctness concern, not circularity. The reference list contains no works by the present authors, so there is no self-citation chain, and no uniqueness theorem or imported ansatz is used. The paper also explicitly acknowledges its scope limitation ('Future studies should explore more complex tasks'), which further indicates the conclusions are not being protected by definitional circularity. No specific circular step can be quoted, so the appropriate score is 0.
Assumptions & free parameters
assumptions (4)
- domain assumption The virtual supermarket environment is a high-fidelity simulation that adequately mimics real-world shopping tasks.
- domain assumption Within-subjects design and identical virtual hand representations isolate interaction modality as the main variable.
- domain assumption Self-report measures (SUS, UEQ-S, custom Likert items) validly capture usability and user experience in this VR context.
- domain assumption The sample of 40 university students is sufficiently representative for the preference and gender-related claims.
Cite this review
Pith. "Pith review of Hands vs. Controllers: Comparing User Interactions in Virtual Reality Shopping Environments." pith.science (2026). https://pith.science/paper/P2IN3S7L
@misc{pith2026250205100,
author = {Pith},
title = {Pith review of: Hands vs. Controllers: Comparing User Interactions in Virtual Reality Shopping Environments},
year = {2026},
howpublished = {\url{https://pith.science/paper/P2IN3S7L}},
note = {Machine review of arXiv:2502.05100}
}
read the original abstract
Virtual reality enables users to experience real-life situations in immersive environments. Interaction methods significantly shape user experience, particularly in high fidelity simulations mimicking real world tasks. This study evaluates two primary VR interaction techniques, hand based and controller based, through virtual shopping tasks in a simulated supermarket with 40 participants. Hand-based interaction was preferred for its natural, immersive qualities and alignment with real-world gestures but faced usability challenges, including limited haptic feedback and grasping inefficiencies. In contrast, controller-based interaction offered greater precision and reliability, making it more suitable for tasks requiring fine motor skills.
Figures
Reference graph
Works this paper leans on
-
[17]
Khundam, C., V orachart, V ., Preeyawongsakul, P., Hosap, W., & No¨el, F. (2021). A comparative study of interaction time and usability of us- ing controllers and hand tracking in virtual reality training. *Informat- ics, 8*(3), 60. 1
work page 2021
-
[7]
Johnson, T. M., Howard, A., & Davidson, L. (2023). Evaluating the accuracy of hand tracking in virtual reality object manipulation tasks. *Journal of Human-Computer Interaction, 39*(5), 385-397. https://doi.org/10.1080/07370024.2023.1123456
-
[8]
Johnson, C. I. (2021). Exploring hand tracking and controller-based interactions in a VR object manipulation task. *Informatics, 8*(3),
work page 2021
- [1]
-
[2]
Branca, G., Resciniti, R., & Loureiro, S. M. C. (2023). Vir- tual is so real! Consumers’ evaluation of product packaging in virtual reality. *Psychology & Marketing, 40*(3), 596-609. https://doi.org/10.1002/mar.217
-
[3]
Franke, T., Attig, C., & Wessel, D. (2019). A personal re- source for technology interaction: Development and validation of the affinity for technology interaction (ATI) scale. *Interna- tional Journal of Human-Computer Interaction, 35*(6), 456-467. https://doi.org/10.1080/10447318.2018.1456150 1
- [4]
-
[6]
Brooke, J. (2013). SUS: A retrospective. *Journal of Usability Studies, 8*(2), 29-40. 1
work page 2013
Show all 20 references
-
[9]
Neamoniti, S., & Kasapakis, V . (2022). Hand tracking vs motion con- trollers: The effects on immersive virtual reality game experience. In *2022 IEEE International Symposium on Multimedia (ISM)* (pp. 206-207). https://doi.org/10.1109/ISM55400.2022.00045 1
2022
-
[10]
C., Zenner, A., & Simeone, A
Nilsson, N. C., Zenner, A., & Simeone, A. L. (2021). Propping up virtual reality with haptic proxies. *IEEE Computer Graphics and Applications, 41*(5), 104-112. https://doi.org/10.1109/MCG.2021.3097671 1
2021
-
[11]
F., M ¨obus, M., Fender, A., & Holz, C
Luong, T., Cheng, Y . F., M ¨obus, M., Fender, A., & Holz, C. (2023). Controllers or bare hands? A controlled evaluation of input techniques on interaction performance and exertion in virtual reality. *IEEE Transactions on Visualization and Computer Graphics, 29*(11), 4633-
2023
-
[12]
H., & Suk, H
Laine, T. H., & Suk, H. J. (2024). Investigating user expe- rience of an immersive virtual reality simulation based on a gesture-based user interface. *Applied Sciences, 14*(11), 4935. https://doi.org/10.3390/app14114935 1
2024 doi
-
[13]
Lataifeh, M., & Ahmed, N. (2019). 3D user interface design and usability for immersive VR. *Applied Sciences, 9*(22), 4861. https://doi.org/10.3390/app9224861
2019 doi
-
[14]
M., & Marks, S
Lee, Y ., Connor, A. M., & Marks, S. (2024). Mixed interaction: Evalu- ating user interactions for object manipulations in virtual space. *Jour- nal of Multimodal User Interfaces*. https://doi.org/10.1007/s12193- 024-00431-2 1, 2
2024 doi
-
[15]
Insko, B. E. (2001). *Passive haptics significantly enhances virtual environments* (Doctoral dissertation, University of North Carolina at Chapel Hill). 1, 2
2001
-
[16]
Masurovsky, A., Chojecki, P., Runde, D., Lafci, M., Przewozny, D., & Gaebler, M. (2020). Controller-free hand tracking for grab-and- place tasks in immersive virtual reality: Design elements and their empirical study. *Multimodal Technologies and Interaction, 4*(4), 91. https:...
2020 doi
-
[18]
Schrepp, M., Hinderks, A., & Thomaschewski, J. (2017). Design and evaluation of a short version of the user ex- perience questionnaire (UEQ-S). *International Journal of In- teractive Multimedia and Artificial Intelligence, 4*(6), 103. https://doi.org/10.9781/ijimai.2017.09.001 1
2017 doi
-
[19]
Siegrist, M., Ung, C.-Y ., Zank, M., Marinello, M., Kunz, A., Hart- mann, C., & Menozzi, M. (2019). Consumers’ food selection behav- iors in three-dimensional (3D) virtual reality. *Food Research Inter- national, 117*, 50-59. https://doi.org/10.1016/j.foodres.2018.01.016
2019 doi
-
[60]
https://doi.org/10.3390/informatics8030060
-
[4643]
https://doi.org/10.1109/TVCG.2023.3320211 1
2023
Reviewed August 8, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.