Review Reports - Ecological and Embodied Assessment of Inhibitory Control Using a VR Stroop Task in Cognitively Healthy Older Adults: A Cross-Sectional Study

Round 1

Reviewer 1 Report (Previous Reviewer 1)

Comments and Suggestions for Authors

The authors have already addressed all my suggestions. I accept their resubmission. Please continue the journal process for this manuscript.

Author Response

1. The authors have already addressed all my suggestions. I accept their resubmission. Please continue the journal process for this manuscript.

RESPONSE#1: We are sincerely grateful to the reviewer for the thorough and constructive feedback provided during the previous rounds of review. We appreciate the positive evaluation and are pleased to hear that our revisions have successfully addressed all of your concerns. Thank you for recommending this manuscript for publication.

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

Although the previously revised and resubmitted paper still contains some deficiencies, the authors are advised to revise the paper, considering the following points to improve its readability and contribution:
- The introduction does not clearly address how the proposed VRST differs from existing VR-based Stroop tasks. More direct references to existing VR-Stroop studies in the literature should be provided, and the methodological or conceptual novelty of this study should be stated in the introduction. - The rationale for selecting healthy older adults as the target population should be discussed in more detail in the introduction. Simply stating that "inhibitory control may be reduced in older adults" is insufficient.
- It is stated that some of the dataset overlaps with a previous study. However, it is not entirely clear which variables, analyses, or subsets of data are shared. For transparency, the variables used in the previous study and those analyzed in this study should be clearly stated, and, if possible, a comparison should be provided in a table. - Although the total sample size is 224, the test-retest reliability analysis was only performed on 30 participants, which is methodologically limiting. It should be stated how this subsample was selected and whether there was any randomness in the selection.
- VRST completion time has different values in the table and in the text. It is given as both 72.65 s and 82.12 s. Why are these values different? If there is an error, it should be corrected.
- ICC results are reported only as high values; 95% confidence intervals or interpretation of the measures should be added. - Showing the correlation results with a scatter plot can present both the relationship and the data distribution more clearly; it should be added if possible.

Author Response

1. The introduction does not clearly address how the proposed VRST differs from existing VR-based Stroop tasks. More direct references to existing VR-Stroop studies in the literature should be provided, and the methodological or conceptual novelty of this study should be stated in the introduction.

RESPONSE#1: We sincerely appreciate the reviewer’s constructive feedback regarding the need to clarify our study's novelty. We have significantly revised the Introduction to provide a more detailed comparison between our Virtual Reality-based Stroop Test (VRST) and existing VR-based Stroop paradigms.

Specifically, we have added references to key studies such as the Virtual Stroop Task.

Parsons, T.D.; Barnett M.D. Virtual apartment stroop task: Comparison with computerized and traditional stroop tasks. J. Neurosci. Methods 2018, 309, 35–40.

Parsons, T.D.; Courtney, C.G.; Arizmendi, B.; Dawson, M. Virtual reality Stroop task for neurocognitive assessment. Stud. Health Technol. Inform. 2011, 163, 433–439.

While these pioneering studies have successfully enhanced ecological validity by situating the task in realistic environments, we noted that they primarily focus on terminal outcomes—such as verbal responses or discrete button presses—to measure performance.

The methodological and conceptual novelty of our study lies in addressing the motor output often overlooked in previous VR-Stroop literature. Grounded in the framework of embodied cognition, the VRST treats inhibitory control as a continuous process integrated with physical movement. Unlike existing tools that record a simple binary response (e.g., whether a button was clicked), the VRST captures high-resolution kinematic trajectories. This allows for the measurement of "interference-in-motion," such as subtle motoric hesitations and path deviations, providing a more granular assessment of how cognitive conflict manifests in real-time motor behavior.

2. The rationale for selecting healthy older adults as the target population should be discussed in more detail in the introduction. Simply stating that "inhibitory control may be reduced in older adults" is insufficient.

RESPONSE#2: We sincerely appreciate the reviewer’s insight. We agree that a more robust justification for targeting cognitively healthy older adults was necessary. We have revised the Introduction to emphasize that this population is critical for identifying the preclinical window of cognitive decline.

As detailed in the revised manuscript, while cognitively healthy older adults maintain general functional independence, they often undergo a subtle erosion of inhibitory control that serves as a precursor to pathological aging. However, traditional paper-and-pencil or computerized assessments frequently encounter ceiling effects in these high-functioning seniors, making it difficult to detect early, sub-clinical lapses in cognitive-motor regulation.

We have added a discussion explaining that by extracting continuous kinematic metrics within this specific demographic, the VRST can detect nuanced behavioral fluctuations (e.g., micro-hesitations) that remain invisible in binary "correct/incorrect" scores. Thus, establishing normative benchmarks in healthy cohorts is a fundamental prerequisite for distinguishing successful aging from pathological decline. We have explicitly clarified this rationale in the fourth paragraph of the Introduction.

3. It is stated that some of the dataset overlaps with a previous study. However, it is not entirely clear which variables, analyses, or subsets of data are shared. For transparency, the variables used in the previous study and those analyzed in this study should be clearly stated, and, if possible, a comparison should be provided in a table.

RESPONSE#3: We appreciate the reviewer’s request for transparency regarding the data overlap. To clarify, the current study was originally designed to establish the psychometric foundation (reliability and validity) of the VRST. Following this, the project was expanded to validate its clinical utility in detecting MCI, which was recently published.

As requested, we have added Table 1 to the Methods section to explicitly delineate the differences between the two studies. While both utilize the same kinematic markers (e.g., 3D trajectory, hesitation latency), the current study provides new longitudinal evidence through two-week test-retest reliability (ICC), which ensures that these behavioral markers are stable traits rather than transient performance fluctuations. Furthermore, we have now clarified that the present study performs a concurrent validity analysis against the gold-standard paper-based Stroop test, an analysis not conducted in the previous clinical study. We believe this table and the accompanying revisions provide the necessary transparency to demonstrate the unique contribution of this work.

4. Although the total sample size is 224, the test-retest reliability analysis was only performed on 30 participants, which is methodologically limiting. It should be stated how this subsample was selected and whether there was any randomness in the selection.

RESPONSE#4: We acknowledge the reviewer’s comment regarding the sample size for the test-retest reliability analysis. While a larger sample is generally preferred, a subsample of n=30 is widely recognized as a methodologically acceptable size for establishing reliability in the development of clinical and neuropsychological tools.

To address the reviewer’s question regarding the selection process, we have added a detailed description of the randomization procedure in the Methods section. Specifically, to minimize selection bias, the 30 participants were selected from the pool of volunteers who expressed willingness to participate in a follow-up session using a computerized simple random selection process. Each consenting participant was assigned a unique identification number, and a computer-based random number generator was employed to select the final 30 participants. This procedure ensured that the selection was entirely objective and that each volunteer had an equal probability of being included in the reliability analysis, thereby reducing potential researcher bias.

5. VRST completion time has different values in the table and in the text. It is given as both 72.65 s and 82.12 s. Why are these values different? If there is an error, it should be corrected.

RESPONSE#5: We sincerely apologize for this discrepancy and thank the reviewer for their meticulous review. Upon re-examining our dataset, we confirmed that there was a clerical error during the manuscript preparation. The value of 72.65 s represents the correct mean completion time for the total healthy cohort (N=224). The inconsistent value of 82.12 s was inadvertently included during the drafting process. We have now standardized all values throughout the text and tables to ensure consistency.

6. ICC results are reported only as high values; 95% confidence intervals or interpretation of the measures should be added. - Showing the correlation results with a scatter plot can present both the relationship and the data distribution more clearly; it should be added if possible.

RESPONSE#6: We agree with the reviewer that reporting 95% confidence intervals and providing a visual representation of data distribution are essential for a robust evaluation of test-retest reliability.

First, we have updated the Results section to include the 95% confidence intervals for all ICC values and added a standardized interpretation of these measures based on the clinical guidelines, confirming the excellent reliability of the VRST metrics.

Second, as suggested, we have added the figure (Scatter plots) to the Results section (Figure 3). These plots illustrate the relationship between the initial and follow-up (two-week) assessments. By visualizing the data points along with a regression line and a y=x identity line, we can more clearly present the strong linear relationship and the consistent distribution of scores within our cohort. We believe these additions significantly enhance the transparency and clarity of our reliability findings.

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Thank you for the revised submission. I have no substantive concerns, and I find the revisions careful and of good quality. My remaining comments are primarily about reporting and writing standards to ensure transparency and reproducibility:

Accessibility and reproducibility of the VR Stroop test. The manuscript describes a “novel virtual reality Stroop test developed based on principles of embodied cognition.” Please specify how other researchers can access this method. For example: Where can the software/application be obtained (e.g., repository link, institutional website, or upon request)? What hardware is required (HMD model, controllers, tracking system, PC specifications), and are the relevant hardware details provided? If specialized equipment is needed, please indicate where it can be sourced (manufacturer/vendor information) so that the experiment can be independently replicated.
Verification of participant involvement and data authenticity. The manuscript states that “224 cognitively healthy older adults (mean age = 71.51 years) randomly completed the VRST and the traditional Stroop test.” To strengthen confidence in the study’s integrity, please clarify what documentation or procedures were used to verify that participants indeed completed the protocols (e.g., attendance logs, signed consent forms, testing timestamps, system usage logs, audit trails, preregistration, or independent monitoring). If video recordings are not available due to privacy/ethics constraints, please explicitly state this and describe alternative verification measures that protect participant confidentiality while supporting data authenticity.
Comparison with similar tools/products and added value. It would be helpful to benchmark the proposed VRST against existing comparable VR-based Stroop or cognitive assessment tools. Please consider adding a brief comparative discussion (or a table) outlining similarities and differences, including advantages and limitations. In particular, clarify what this VRST improves relative to alternatives (e.g., ecological validity, sensitivity, user experience, feasibility, cost, time efficiency), and provide a clear rationale for why readers/researchers or potential users should choose this tool.

I believe addressing these points will significantly improve the manuscript’s clarity, credibility, and practical impact.

Author Response

1. Accessibility and reproducibility of the VR Stroop test.The manuscript describes a “novel virtual reality Stroop test developed based on principles of embodied cognition.” Please specify how other researchers can access this method. For example: Where can the software/application be obtained (e.g., repository link, institutional website, or upon request)? What hardware is required (HMD model, controllers, tracking system, PC specifications), and are the relevant hardware details provided? If specialized equipment is needed, please indicate where it can be sourced (manufacturer/vendor information) so that the experiment can be independently replicated.

RESPONSE#1: We sincerely appreciate the reviewer’s emphasis on the reproducibility of our findings. We have addressed the concerns regarding accessibility and technical specifications as follows:

First, regarding software accessibility, the VRST application was custom-developed by the research team using the Unity 3D engine. To facilitate independent replication while maintaining the integrity of the developed tool, we have included a dedicated statement in the Acknowledgments section at the end of the manuscript, noting that the software is available for academic and non-commercial research purposes upon reasonable request to the corresponding author.

Second, we have significantly expanded the description of the hardware and PC specifications in the Methods section. We have specified that the study was conducted on a PC running Windows 10, equipped with an Intel Core i7-10700 CPU and 16 GB of RAM, using the HTC VIVE system and SteamVR Tracking 1.0. These detailed parameters ensure that other researchers can accurately replicate the high-precision kinematic data acquisition environment.

2. Verification of participant involvement and data authenticity.The manuscript states that “224 cognitively healthy older adults (mean age = 71.51 years) randomly completed the VRST and the traditional Stroop test.” To strengthen confidence in the study’s integrity, please clarify what documentation or procedures were used to verify that participants indeed completed the protocols (e.g., attendance logs, signed consent forms, testing timestamps, system usage logs, audit trails, preregistration, or independent monitoring). If video recordings are not available due to privacy/ethics constraints, please explicitly state this and describe alternative verification measures that protect participant confidentiality while supporting data authenticity.

RESPONSE#2: We fully agree with the reviewer that ensuring research integrity and data authenticity is paramount. To provide objective evidence of participant involvement and the authenticity of the collected data, we have included a comprehensive Attendance Log for all 224 participants as Non-Published Material. This log documents the specific testing dates, participant IDs, session numbers, and precise entry/exit timestamps for each individual. For the 30 participants who participated in the test–retest reliability analysis, their follow-up sessions (Session 2) are also explicitly recorded.

3. Comparison with similar tools/products and added value.It would be helpful to benchmark the proposed VRST against existing comparable VR-based Stroop or cognitive assessment tools. Please consider adding a brief comparative discussion (or a table) outlining similarities and differences, including advantages and limitations. In particular, clarify what this VRST improves relative to alternatives (e.g., ecological validity, sensitivity, user experience, feasibility, cost, time efficiency), and provide a clear rationale for why readers/researchers or potential users should choose this tool.

RESPONSE#3: We appreciate the reviewer’s insightful comment. As Reviewer #2 also pointed out a similar need for clarification regarding our study's novelty, we have extensively revised both the Introduction section to address this.

As noted in the revised Introduction, existing VR-based Stroop paradigms have significantly improved ecological validity by using immersive environments. However, these tools still primarily rely on terminal, discrete responses (e.g., verbal answers or button presses).

Parsons, T.D.; Barnett M.D. Virtual apartment stroop task: Comparison with computerized and traditional stroop tasks. J. Neurosci. Methods 2018, 309, 35–40.

Parsons, T.D.; Courtney, C.G.; Arizmendi, B.; Dawson, M. Virtual reality Stroop task for neurocognitive assessment. Stud. Health Technol. Inform. 2011, 163, 433–439.

The added value and conceptual novelty of our VRST lie in the following:

Embodied Cognition Framework: Unlike previous tools, our VRST treats inhibitory control as a continuous process integrated with physical movement. By utilizing a dressing-room scenario that requires 6-DoF grabbing and sorting, it engages the motor system more deeply than simple pointer-clicking.
Kinematic Sensitivity: The VRST captures high-resolution trajectories, allowing us to measure "interference-in-motion." This provides a granular assessment of cognitive conflict (e.g., motoric hesitation, path deviation) that traditional binary metrics (correct/incorrect) cannot detect.

We believe these revisions clearly demonstrate why our tool offers significant methodological improvements over existing alternatives.

Reviewer 4 Report (Previous Reviewer 4)

Comments and Suggestions for Authors

This manuscript introduces a virtual reality–based Stroop test designed to assess inhibitory control in cognitively healthy older adults using an embodied cognition framework. The task requires participants to perform a virtual clothing-sorting task while ignoring conflicting color cues, and performance is quantified using both cognitive and kinematic metrics

My comments are as follows:

1) The VR task involves sorting clothing items into color-coded boxes using a controller. But it is not clear the exact trial structure, difficulty of manipulation, learning effects, or whether task randomization was controlled. Additionally, the manuscript reports 30 trials per session but later mentions 20 stimuli, creating ambiguity about the total number of trials.

2) Several terms are inconsistent throughout the manuscript, like "competition time" vs. "completion time." These should be standardized.

3 ) The diagram on page 5 illustrates the VR setup and interaction flow but lacks detailed labeling of sensors and tracking components.

Author Response

My comments are as follows:

1. The VR task involves sorting clothing items into color-coded boxes using a controller. But it is not clear the exact trial structure, difficulty of manipulation, learning effects, or whether task randomization was controlled. Additionally, the manuscript reports 30 trials per session but later mentions 20 stimuli, creating ambiguity about the total number of trials.

RESPONSE#1: We sincerely apologize for the lack of clarity regarding the task structure and the typographical error in the stimulus count. We have extensively revised the Methods (Section 2.2. Procedure) to address these points.

Clarification of Trial Numbers: We have corrected the inconsistency. Each session consisted of exactly 30 trials, utilizing 30 distinct incongruent stimuli. The previous mention of "20 stimuli" was a typographical error and has been corrected to "30" throughout the manuscript.
Trial Structure and Randomization: To ensure experimental rigor and prevent predictability, the sequence of the 30 stimuli was fully randomized for each participant and session via the software’s algorithm. This counterbalancing was designed to mitigate potential order effects.
Difficulty and Learning Effects: To control for the difficulty of 6-DoF manipulation and minimize learning effects, we implemented a standardized tutorial and practice session. As now specified in the revised manuscript, this session continued until each participant demonstrated sufficient proficiency in manipulating the VR controller. This ensured that the recorded performance reflected cognitive inhibitory control rather than technical unfamiliarity or the learning curve of the VR interface.
Standardized Task Parameters: We have also clarified that factors such as the distance to storage boxes (1.5 meters), haptic vibration, and auditory feedback were fixed to ensure a standardized testing environment across all 224 participants.

2. Several terms are inconsistent throughout the manuscript, like "competition time" vs. "completion time." These should be standardized.

RESPONSE#2: We sincerely appreciate the reviewer for identifying these inconsistencies. We have conducted a comprehensive review of the entire manuscript to ensure terminological consistency. Specifically, all instances of "competition time" (which was a typographical error) have been corrected to "completion time" to accurately reflect the duration required to complete the task. Furthermore, we have standardized other key terms related to our metrics, such as "movement distance" and "hesitation latency," throughout the text to ensure clarity and coherence.

3. The diagram on page 5 illustrates the VR setup and interaction flow but lacks detailed labeling of sensors and tracking components.

RESPONSE#3: We agree with the reviewer that a more detailed technical illustration is necessary for a complete understanding of the VRST hardware infrastructure. Accordingly, we have significantly revised Figure 1 to provide precise labeling of all sensors and tracking components.

The updated diagram now explicitly illustrates the following technical details:

Lighthouse Base Stations (SteamVR Tracking 1.0): We have added labels specifying their wide Field of View (120° Horizontal and Vertical) and their optimized installation parameters, including a downward tilt of 30–45 degrees to maximize the tracking volume.
Tracking Mechanism (Laser-Sensor Interaction): To clarify how the system captures movement, we have illustrated that the Base Stations function as laser emitters rather than simple cameras. The diagram now shows the Photo-diode sensors embedded on the surfaces of both the HMD and the 6-DoF handheld controllers, which detect the laser sweeps to calculate high-precision spatial coordinates.
Active Tracking Interaction (Laser-Sensor Recognition): To clarify how the system captures movement, we have illustrated the dynamic interaction between the Base Stations (Motion Sensors) and the 6-DoF handheld controllers using signal indicators (e.g., lightning bolt-style paths). These visuals, along with added text, demonstrate how the Motion Sensors actively recognize controller movements through the Photo-diode sensors embedded on the controllers, which detect laser sweeps to calculate high-precision spatial coordinates.
Active Tracking Volume: The updated figure visualizes the synchronized tracking area (Active Tracking Volume) where sub-millimeter precision is maintained.

We believe these enhancements provide the technical rigor required for reproducibility and clearly demonstrate the precision of our measurement system.

Round 2

Reviewer 2 Report (Previous Reviewer 2)

Comments and Suggestions for Authors

The authors have revised the manuscript by taking the revision comments into account, and the paper is acceptable in its current form.

Author Response

1. The authors have revised the manuscript by taking the revision comments into account, and the paper is acceptable in its current form.

RESPONSE#1: We are sincerely grateful to the reviewer for the positive evaluation and for recommending our manuscript for publication. We appreciate the constructive feedback provided in the previous rounds, which greatly helped improve the quality and clarity of this work

Reviewer 3 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Regarding software accessibility, the VRST application was custom-developed by the research team using the Unity 3D engine. To facilitate independent replication while maintaining the integrity of the developed tool, we have included a dedicated statement in the Acknowledgments section at the end of the manuscript, noting that the software is available for academic and non-commercial research purposes upon reasonable request to the corresponding author. I fully understand the confidentiality requirements for this section. Could you please provide a trial version for reviewers or editors? Otherwise, we have no way of knowing whether the software is genuine or fabricated. At least some photos and screenshots of the user interface are needed.

Comments on the Quality of English Language

Author Response

1. Regarding software accessibility, the VRST application was custom-developed by the research team using the Unity 3D engine. To facilitate independent replication while maintaining the integrity of the developed tool, we have included a dedicated statement in the Acknowledgments section at the end of the manuscript, noting that the software is available for academic and non-commercial research purposes upon reasonable request to the corresponding author. I fully understand the confidentiality requirements for this section. Could you please provide a trial version for reviewers or editors? Otherwise, we have no way of knowing whether the software is genuine or fabricated. At least some photos and screenshots of the user interface are needed.

RESPONSE#1: We appreciate the reviewer’s request to verify the authenticity of our custom-developed software. Regarding the provision of a trial version, we would like to clarify that the VRST application is deeply integrated with specific VR hardware (HTC VIVE system) and the SteamVR environment, making it technically challenging to provide a standalone executable that functions correctly on a standard PC without the required VR interface and sensors.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The author should clarify the novelty of the VRST compared with existing VR-Stroop paradigms, especially in terms of embodied kinematic features and their theoretical contribution.
The author should consider reviewing and citing related studies on digital framework development, such as “Developing a Digital Literacy Framework for Rural Farmers in China,” to strengthen the theoretical grounding and demonstrate awareness of recent work on framework construction, expert validation (Delphi), and multi-level indicator design, which could provide useful methodological parallels and support the rigor of the present study.
Although concurrent validity is reported, the manuscript should include effect size interpretation and confidence intervals to strengthen the statistical rigor.
The author should better justify why error count showed poor reliability and consider task difficulty adjustment or signal-detection–based metrics.
Although the sample size is adequate, the author should discuss potential sampling bias related to digital literacy and VR familiarity in older adults.
The author should report hardware specifications (HMD model, tracking accuracy, sampling rate) to improve reproducibility of the kinematic measures.
Although embodied cognition is well discussed, the manuscript should more explicitly link each kinematic variable to specific inhibitory control mechanisms using computational or neurocognitive models.
The author should add a short limitations paragraph on cross-sectional design and the absence of predictive validity for conversion to MCI or dementia.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors are encouraged to revise the manuscript taking the following points into consideration:
- In the abstract, the content of the VR-based Stroop task (e.g., task duration, difficulty level, ratio of congruent/incongruent trials) is presented superficially and should be described more explicitly.
- Considering the study’s focus, the keyword list could be expanded to include "older adults."
- It remains unclear until the end of the introduction how the VRST differs from existing VR-based Stroop tasks or other ecological tests. This distinction should be clarified in the introduction.
- Although age-related changes in inhibitory control are mentioned, the rationale for specifically targeting healthy older adults and the theoretical or clinical significance of this population should be addressed in the introduction.
- While the classic Stroop paradigm is well explained, how this paradigm was adapted to the VR environment and how Stroop interference was operationalized in the virtual task should be explicitly described.
- Although partial overlap with a previous study dataset is mentioned, it is not clearly specified which participants, variables, or analyses were shared. This information should be clearly detailed.
- Were all participants retested after two weeks, or only a subset? Please clarify.
- The method of randomization for the VRST and traditional Stroop tests should be described.
- Technical details necessary for reproducibility—such as the HMD model, resolution, refresh rate, field of view, controller type, and software platform—should be provided.
- The lack of correlation between VRST variables and traditional Stroop error count should be discussed in terms of how it affects the VRST’s sensitivity to detecting cognitive errors.
- Only Pearson correlations were used to compare the VRST and traditional Stroop tests. The rationale for not using more advanced statistical analyses should be explained.
- It is noted that the VR hardware and software did not provide fully natural interaction and immersion. This limitation and its potential impact on the embodied cognition framework should be analyzed in the discussion section.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses a potentially relevant topic and, in principle, does not appear to suffer from fundamental conceptual or analytical flaws. However, in its current form, the study lacks sufficient clarity, consistency, and methodological transparency to allow a proper scientific evaluation. Substantial revisions are therefore required before the manuscript can be considered for acceptance.

Inconsistency in participant numbers (major concern)

A critical issue concerns the inconsistent reporting of participant numbers. The Abstract states that 224 older adults were recruited, whereas the Methods section later reports 58 male and 44 female participants. This discrepancy raises serious concerns regarding sample description and data reporting. The authors must clarify the actual sample size, recruitment flow, and participant allocation, and ensure consistency throughout the manuscript.
Insufficient description of the VRST intervention (major concern)

At present, the VRST intervention is not described in sufficient detail to allow readers to clearly understand what was actually implemented. The manuscript does not provide enough information on the structure, content, duration, or delivery of VRST. As a result, it is difficult to assess whether this study represents a concrete experimental intervention or remains largely conceptual.
Reproducibility and accessibility of the VRST system (major concern)

The lack of information regarding access to the VRST software is a major limitation. The authors should clearly state whether the software is commercially available, custom-developed, or restricted, and provide instructions on how other researchers could obtain, access, or validate the system. Without such information, the study cannot be meaningfully replicated or independently verified, which substantially limits its scientific value.
Lack of supporting evidence for real-world implementation

To strengthen confidence in the authenticity and execution of the study, the authors are strongly encouraged to provide supplementary materials (e.g., demonstration videos, screenshots, or photographs of the experimental setup). Such materials would greatly enhance transparency and help reviewers and readers understand how VRST was applied in practice.
Overall assessment and recommendation

Overall, while the study does not appear to contain obvious analytical errors, the current presentation lacks sufficient methodological detail to support its claims. The manuscript reads more like a proof-of-concept or descriptive report than a fully documented empirical study. Unless the authors substantially improve transparency, provide concrete methodological details, and address the issues outlined above, the manuscript cannot be recommended for acceptance.

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript investigates the concurrent validity and test–retest reliability of a newly developed virtual reality–based Stroop test (VRST) grounded in the framework of embodied cognition. The authors evaluated the VRST in a relatively large sample of 224 cognitively healthy older adults, comparing VR-derived cognitive and kinematic measures with a traditional paper-based Stroop test.

My comments are below.

1) It is not fully clear whether the VRST primarily measures inhibitory control or a broader executive motor construct. Please discuss more explicitly how inhibitory control is isolated within the VRST.

2) Both concurrent validity and test–retest reliability analyses show that error count performs poorly compared to other VRST metrics. Despite this, error count is still presented as a core outcome measure. Please clarify whether error count should be retained as a primary VRST outcome, or whether it should be considered exploratory.

3) The study exclusively includes cognitively healthy older adults. The conclusions should more clearly emphasize that findings cannot be generalized to clinical populations.

4) Prior experience with digital devices or VR was not assessed. Please note more clearly how digital familiarity may have influenced performance and reliability, especially in older adults.