F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.161090.2 Research Article Articles Exploring the relationship between preferred bubble tube speeds in sensory rooms and physiological–psychological factors: A study on interoceptive sensitivity, subjective time perception, visual discomfort levels, and anxiety levels

[version 2; peer review: 1 approved, 2 approved with reservations]

 Su Anjie Data Curation Formal Analysis Funding Acquisition Investigation Writing – Original Draft Preparation 1 Shen Junyi Data Curation Formal Analysis Investigation 1 Koyama Shinichi Conceptualization Funding Acquisition Investigation Methodology Supervision Validation Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0001-6280-5750 a 2 Doctoral Program in Design, University of Tsukuba, Tsukuba, Ibaraki Prefecture, Japan Institute of Art and Design, University of Tsukuba, Tsukuba, Ibaraki Prefecture, Japan skoyama@geijutsu.tsukuba.ac.jp

No competing interests were disclosed.

 27 3 2026 2025 14 214 24 2 2026 Copyright: © 2026 Su A et al. 2026 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background

This study examined how preferred bubble-tube motion speeds in sensory rooms relate to individual physiological and psychological characteristics, including interoceptive sensitivity, subjective time perception, visual discomfort, and anxiety levels.

 Methods

Fifty adult participants took part in a controlled laboratory experiment using a method-of-adjustment procedure to select their most comfortable motion speed for a simulated bubble tube, presented as an upward-moving Random Dot Motion (RDM) stimulus. Subjective time perception was evaluated using a 60-second time-estimation task, and interoceptive sensitivity was measured via a heartbeat-tracking task. Visual discomfort and anxiety were assessed using the Japanese versions of the Visual Discomfort Scale (VDS-J), Trypophobia Questionnaire (TQ-J), and State-Trait Anxiety Inventory (STAI).

 Results

The results from the method of adjustment indicated that the preferred speed varied widely, from 1.09 to 13.86 degrees per second. Spearman’s correlation analysis revealed that higher interoceptive awareness was associated with a preference for slower speeds, whereas higher anxiety levels were associated with a preference for faster speeds. In addition, multiple regression analysis showed that subjective time-perception accuracy and visual discomfort levels were significant predictors of participants’ preferred RDM speeds.

 Conclusion

The results indicate that interoceptive sensitivity, subjective time perception, visual discomfort, and anxiety levels play significant roles in determining preferred RDM stimulation speeds. These findings highlight the importance of considering individual differences in physiological and psychological states when designing therapeutic sensory environments, such as sensory rooms and bubble tubes, to support comfort, well-being, and therapeutic outcomes.

 Sensory Hypersensitivity Sensory Room Bubble Tubes Interoceptive Sensitivity Subjective Time Perception Visual Discomfort Anxiety Japan Society for the Promotion of Science AFD29210 18H03663 21H03758 Japan Science and Technology Agency JPMJSP2124 This study was supported by the Topic-Setting Program to Advance Cutting-Edge Humanities, Social Sciences Research Area Cultivation (AFD29210), a Grant-in-Aid for Scientific Research (A) (18H03663), and a Grant-in-Aid for Scientific Research (B) (21H03758) awarded to Shinichi Koyama. Additionally, a JST SPRING grant (Grant Number: JPMJSP2124) was awarded to Su Anjie. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Amendments from Version 1

In the abstract, we made minor revisions (highlighted parts) for the gramatical correction and clalification purpose, based on the consultation with a professional English proofreader.

Reflecting the recommendation from reviewer 1, we added citations to Green et al. (2012), MacLennan et al. (2020, 2021), and Verhulst et al. (2022) in the Introduction. These references now contextualize the theoretical link between sensory over-responsivity (SOR) and anxiety.

Reflecting the recommendation [1] from reviewer 2, we avoided using the word ASD, and rephrase the sentence to more neurodiversity-community-friendly one.

Reflecting the recommendation [5] from Reviewer 2, we added key citations (Champagne & Stromberg, 2004; Sutton et al., 2013; Scanlan & Novak, 2015) to contextualize the mechanisms of multisensory modulation within the existing literature.

Reflecting the recommendation [6] from Reviewer 2, we strengthened parallels to existing research in sensory processing and neurodivergent experiences by incorporating key evidence from observational studies and systematic reviews (e.g., Breslin et al., 2020; Haig & Hallett, 2023).

Reflecting the recommendation [7] from Reviewer 2, we expanded the conceptual framework to explicitly distinguish interoceptive dimensions (referencing Garfinkel et al., 2015)

Reflecting the recommendation [7] from Reviewer 2, we strengthened the theoretical framework to explicitly explain how subjective time perception relates to interoceptive processing and anxiety, resolving the previous disconnection.

Reflecting the recommendation [7] from Reviewer 2, we expanded the discussion on visual discomfort, elaborating on pattern-glare mechanisms and the impact of spatial/temporal statistics with additional references.

Reflecting the recommendation [7] from Reviewer 2, we clarified the clinical distinction between state and trait anxiety to better contextualize the STAI measurements used in this study.

Introduction

Sensory overload is a common experience among individuals with sensory hypersensitivity, in which everyday stimuli such as light ( Lucker, 2013), noise ( Khalfa et al., 2004; Landon et al., 2016), and touch ( Blakemore et al., 2006) can become overwhelming. Individuals may find it challenging to navigate routine environments ( Lucker, 2013), which can affect their ability to function effectively in daily life ( Kim et al., 2000). Researchers and practitioners have developed various methods and tools to provide sensory support ( Koegel et al., 2004). These approaches aim not only to reduce stress associated with sensory overload ( Mazurek et al., 2013) but also to enhance quality of life and agency within everyday environments ( Hannant et al., 2016). Studies have shown that sensory hypersensitivity can significantly influence behavior and emotional experiences and is associated with heightened distress responses ( Engel-Yeger et al., 2016). Consistent with this, work on sensory over-responsivity (SOR) in autism demonstrates robust links with anxiety across development, including theoretical and longitudinal evidence that SOR and anxiety can exacerbate one another ( Green & Ben-Sasson, 2010; Green et al., 2012). More recent studies further link sensory reactivity differences to anxiety subtypes in autistic children and to perceived causal relations in autistic adults ( MacLennan et al., 2020, 2021; Verhulst et al., 2022).

Designed to provide a controlled environment that supports self-regulation and relaxation, sensory rooms are typically equipped with assistive technologies and specialized elements. A prominent example of the Snoezelen room, a multisensory environment developed in the Netherlands in the 1970s ( Hulsegge & Verheul, 2005). Snoezelen rooms combine sensory experiences to create a soothing atmosphere in which users can explore stimuli at their own pace, thereby promoting relaxation and reducing stress ( Hulsegge & Verheul, 2005). These environments are particularly beneficial in psychiatric inpatient care, where they provide sensory-accessible spaces that help individuals navigate environmental stressors. Research has shown that sensory rooms, including Snoezelen rooms, significantly enhance emotional well-being and reduce distress, especially among individuals who experience heightened distress responses or anxiety ( Haig & Hallett, 2023; West et al., 2017). They also serve as valuable tools for self-management and offer alternatives to more restrictive practices ( Barbic et al., 2019). In forensic mental health settings, these rooms have been shown to reduce inpatient stress, support recovery, and enhance overall lived experience within facilities ( Wiglesworth & Farnworth, 2016).

Among commonly used sensory-room equipment, bubble tubes often serve as a central element, providing visual and auditory stimulation. These tall, water-filled cylinders produce continuously rising bubbles via an internal air pump and are often paired with color-changing LED lights, supporting visual tracking, color engagement, and gentle auditory cues. The continuous movement and changing colors of the bubbles can soothe and attract attention ( Fujisawa, 2021). Bubble tubes are frequently described as calming for autistic people and others with sensory processing differences; observational studies indicate that they are among the most popular items in multisensory environments ( Unwin et al., 2024).

However, the precise mechanisms by which bubble tubes promote relaxation remain poorly understood. Research on sensory modulation in psychiatric inpatient settings and multisensory environments suggests that low-demand, user-controlled multisensory input can support arousal regulation and reduce distress ( Champagne & Stromberg, 2004; Sutton et al., 2013; Scanlan & Novak, 2015). From a sensory-integration perspective, foundational models emphasize individual differences in sensory thresholds and self-regulatory capacity that likely shape responses to such environments ( Dunn, 2001; Lane et al., 2019). Syntheses of evidence across multisensory environment implementations report generally positive within-session effects on calm and engagement but variable outcomes across populations and programs, underscoring the need to identify person-level moderators ( Breslin et al., 2020; Van Weert et al., 2005; Haig & Hallett, 2023).

We focused on four constructs that plausibly shaped comfort during visual stimulation. Interoception can be described in at least three related facets: (i) accuracy—objective performance on tasks indexing detection of internal bodily signals; (ii) sensibility—self-reported beliefs and attention toward internal sensations; and (iii) awareness—metacognitive insight into one’s interoceptive performance ( Garfinkel et al., 2015; Critchley & Garfinkel, 2017). In the present study, we focused on interoceptive accuracy, assessed using an objective heartbeat-tracking task, because objective indices are most directly comparable with behavioral preferences measured in our speed-selection paradigm ( Garfinkel et al., 2015). Perception of internal bodily signals (e.g., heartbeat) supports monitoring and adjustment of emotional and physiological states, shaping responses to external sensory input ( Garfinkel & Critchley, 2013; Pollatos et al., 2012; Tanaka et al., 2021).

Subjective time perception, shaped by attention, emotion, and sensory input, can influence how dynamic visual stimuli are experienced and may covary with interoceptive processes; timing errors may bias preferred motion speeds toward what feels subjectively “right” ( Yabe & Yamada, 2023). Time processing in the seconds range recruits the insular cortex and covaries with interoceptive accuracy, such that individuals with more precise detection of cardiac signals tend to show more accurate duration judgments ( Meissner & Wittmann, 2011; Vicario et al., 2020). Affective states also modulate temporal judgments: meta-analytic and experimental work indicates that fear and anxiety can distort perceived duration via changes in arousal and attentional allocation, with anxiety in particular biasing underestimation in threat-of-shock paradigms ( Droit-Volet & Meck, 2007; Sarigiannidis et al., 2020). Given that bubble tubes are intended to enhance well-being and are used by people with sensory processing differences, it is pertinent to examine how interoceptive accuracy and time perception relate to preferred bubble-tube speeds.

Visual discomfort can also shape responses to bubble-tube stimulation. Individuals reporting higher visual discomfort often perform visual tasks more slowly, and uncomfortable visual patterns are associated with stronger yet less frequent neural responses, consistent with less efficient coding ( Conlon & Humphreys, 2001; Wilkins, 2016). Accordingly, the speed of visual motion may either alleviate or exacerbate discomfort, making it a design-relevant parameter in sensory rooms. Visual discomfort is reliably elicited by images whose spatial statistics deviate from natural 1/f structure—typically with excess mid-spatial-frequency energy—and by patterns that provoke pattern glare (e.g., high-contrast stripes), with elevated susceptibility in migraine and related conditions ( Evans & Stevenson, 2008; Penacchio & Wilkins, 2015). Analogous principles apply in the temporal domain: flicker and temporal profiles that depart from natural temporal statistics increase discomfort, and adaptation can systematically shift both perceived temporal sharpness and discomfort ( Yoshimoto et al., 2019). These observations motivate treating motion speed as a design-relevant parameter that can either alleviate or exacerbate discomfort in sensory rooms.

Anxiety is likewise important. Elevated state anxiety is linked to heightened sensitivity to environmental cues and greater distress under sensory load ( Åhs et al., 2013), and greater sensory-processing sensitivity is associated with higher anxiety, depression, and stress, particularly when mindfulness and acceptance are low ( Bakker & Moulding, 2012). Research on SOR in autism indicates sustained associations with anxiety, including bidirectional influences observed longitudinally in toddlers ( Green et al., 2012) and systematic links to anxiety subtypes in childhood as well as perceived causal pathways in adults ( MacLennan et al., 2020, 2021; Verhulst et al., 2022; Green & Ben-Sasson, 2010). In the present study, we assessed both state anxiety (a transient, context-dependent anxious affect) and trait anxiety (a relatively stable dispositional tendency) using the simplified State-Trait Anxiety Inventory ( Spielberger, 1983; Koizumi et al., 1998). Conceptually, state anxiety primarily indexes momentary arousal that can shift sensory gating, whereas trait anxiety has been linked to broader differences in attentional control that may heighten stimulus-driven processing ( Eysenck et al., 2007). Taken together, these considerations underscore the relevance of interoceptive accuracy, subjective time perception, visual discomfort, and anxiety in understanding individual variation in preferred bubble-tube speed.

Although we did not recruit a clinically characterized cohort, adjacent work indicates that person-level sensory profiles and user control of sensory inputs shape engagement and calm in multi-sensory environments ( Dunn, 2001; Lane et al., 2019; Scanlan & Novak, 2015; Champagne & Stromberg, 2004; Unwin et al., 2022, 2024). In autistic populations, reviews document distinctive motion and multisensory processing that can shift preferred parameters of dynamic visual input ( Robertson & Baron-Cohen, 2017). These observations motivate treating bubble-tube motion speed as a design-tunable parameter and considering person-level characteristics that plausibly modulate comfort, without presupposing clinical diagnosis.

We used an upward-moving random dot motion (RDM) display to isolate the motion-speed component of bubble-tube viewing while minimizing confounds from color, texture, and form. RDM/RDK paradigms are canonical tools for probing global motion, with well-established neural underpinnings in primate area MT/V5 and tight psychophysics–neurophysiology links ( Newsome & Paré, 1988; Britten et al., 1992; Salzman et al., 1992). In neurodiversity research, RDM tasks are widely used to assess global motion processing in autistic participants, with mixed but convergent evidence for small group-level differences and strong parameter control across studies (e.g., Milne et al., 2002; Koldewyn et al., 2010; Van der Hallen et al., 2019; see also Robertson & Baron-Cohen, 2017). This makes RDM an appropriate, translationally useful proxy for bubble-tube motion when the design-tunable variable of interest is speed.

In this study, we tested whether interoceptive accuracy, subjective time perception, anxiety, and visual discomfort levels jointly predict preferred stimulation speed. This approach provides design-relevant estimates that can be translated into adjustable settings for sensory-room equipment.

 Methods Participants

The sample size was determined through a priori power analysis using G*Power (version 3.1). Based on the consideration of three key parameters—a large effect size (r = 0.50) following Cohen’s (1988) guidelines to reflect the anticipated substantive and meaningful association between variables, a conventional significance level ( α = 0.05) to maintain an acceptable balance between Type I error rates and the ability to detect true effects, and a statistical power (1- β = 0.80) to ensure an 80% likelihood of detecting a true effect if one existed—G*Power recommended a minimum sample size of 42 participants. To account for potential data loss or participant dropout, we recruited 50 participants to enhance the robustness and reliability of the statistical analyses.

Participants were not screened for clinical diagnoses (e.g., autism, anxiety, or mood-related conditions), and psychotropic medication was not recorded. Our aim was to model inter-individual differences in a non-clinical adult sample rather than to estimate effects by diagnostic subgroup; we did not solicit diagnostic disclosures to minimize privacy burden and because our predictors (interoceptive accuracy, time-estimation error, state anxiety, visual discomfort) are non-diagnostic mechanistic indices relevant to design-tunable parameters.

We recruited 50 participants (27 females) aged 22–35 years (M = 25.94, SD = 2.74) from the University of Tsukuba between December 25, 2023, and March 22, 2024. Prior to participation, all participants provided written informed consent as approved by the Institutional Review Board (IRB) of the Institute of Art and Design, the University of Tsukuba (IRB No. [GEI021-15]). On the day of the experiment, participants were asked to abstain from alcohol, caffeine, and cigarettes to support consistency of physiological state across participants. Participants reported adequate sleep and typical or corrected-to-typical vision.

 Stimuli

We used a Random Dot Motion (RDM) stimulus program developed using the Flutter SDK. The source code (v1.0) is available under the MIT License on Zenodo ( Su, 2025; https://doi.org/10.5281/zenodo.14795461), with a Windows executable build concurrently archived ( Su, 2025d; https://doi.org/10.5281/zenodo.14795194). The program ran on a Lenovo laptop (screen dimensions: 36.3 cm × 23.8 cm, resolution: 1920 × 1080, model number: 115423562). Our RDM program generated dots that moved vertically upward within a circular aperture centered on the screen ( Figure 1). All dots moved at the same speed, simulating the upward movement of the bubbles in the tube. This circular area had a diameter of 1000 pixels, corresponding to a visual angle of 21.88°. A white fixation cross was placed at the center of the circular area. This helped ensure that participants maintained fixation on a specific location and sustained visual attention, facilitating detection of subsequent stimuli. During the experiment, the program randomly generated 200 dots per second within a circular area. Each dot had a radius of 12 pixels, corresponding to a visual angle of 0.26°. All dots had the same brightness and chromaticity, set at 128 cd/m 2, X = 0.23, Y = 0.28 in the CIE 1931 color space, ensuring consistent visual presentation.

Schematics of the Random Dot Motion (RDM) stimulus.

The RDM stimulus allowed parametric control over upward motion speed (°/s) while holding dot density, lifetime, and luminance statistics constant, thereby aligning the laboratory manipulation with the adjustable bubble-rise rate used in sensory-room equipment. This choice is grounded in extensive basic-science validation of RDM for motion perception and in its widespread use to characterize motion processing in autistic samples, facilitating comparison with prior neurodiversity research ( Newsome & Paré, 1988; Britten et al., 1992; Salzman et al., 1992; Milne et al., 2002; Koldewyn et al., 2010; Van der Hallen et al., 2019).

During the experiment, participants adjusted the upward speed of the dots using a keyboard. The program recorded the current dot speed when the “Enter” key was pressed, allowing for later review. The upward motion speed ranged from 0 pixel/s (0°/s) to 1000 pixels/s (21.88°/s). Bellini et al. (2023) reported that sensory environments typically use soft, dim lighting to create a tranquil atmosphere that supports comfort, relaxation, and reduced stress. Therefore, the experiment was conducted in a darkened room designed to simulate a sensory-room environment ( Figure 2).

Experimental setup shows a participant adjusting the speed of the dots in the Random Dot Motion (RDM) stimulus. Apparatus

A fingertip pulse oximeter (CMS50D, Contec Medical Systems Co., Ltd.) was used to monitor the participants’ heart rates throughout the experiment. We also used a stopwatch to measure participants’ subjective time perception.

 Procedures

To ensure consistent viewing of the RDM stimuli, a chin rest was used to maintain a viewing distance of 57 cm between the participant’s eyes and the screen. Participants were given control over the upward movement speed of the dots, adjusting it using arrow keys until they identified the most comfortable speed, which they confirmed by pressing the Enter key.

A consolidated summary of all measures is presented in Tables 1A (experimental tasks) and 1B (questionnaires), including key settings, scoring, and validation references.

Experimental tasks.

Construct Task Key settings Scoring Validation refs
Preferred stimulation speed Method of Adjustment with upward RDM 6 runs; ascending & descending starts; aperture 21.88°; 200 dots/s Mean of 6 speeds (deg/s) Wier et al., 1976
Subjective time perception 60-s time-estimation Eyes closed; self start/stop stopwatch |recorded − 60|in seconds Castellotti et al., 2022
Interoceptive accuracy Heartbeat-tracking 60 s; no palpation; pulse oximeter |counted−recorded| Schandry, 1981

Questionnaires.

Construct Instrument Items/Scale Scoring Validation refs
Visual discomfort (General) VDS-J 23 items; 0–3 Sum/mean (higher = more) Imaizumi et al., 2018
Visual discomfort (Trypophobia) TQ-J 17 items; 1–5 Sum/mean Imaizumi et al., 2016
Anxiety (State & Trait) STAI-Y (JP short form) 4-point; Y-1 & Y-2 Separate state/trait sums Koizumi et al., 1998

Open science note. All materials required to reproduce the stimuli, measures, and analyses are openly available; see the Data, code, and materials availability section for persistent links (DOIs).

We used a method-of-adjustment procedure with an upward random-dot motion (RDM) display to isolate the motion-speed component of bubble-tube viewing while minimizing color/texture/form confounds. Each participant completed eight runs (two practice, six recorded) with randomized ascending (0→21.88°/s) and descending (21.88→0°/s) starting speeds. Participants adjusted speed until it felt most comfortable and pressed “Enter” to record the value. The primary dependent variable was the mean of the six recorded speeds (°/s). This classical method-of-adjustment approach is standard in psychophysics and supports precise parameter control with RDM.

Following Castellotti et al. (2022), participants completed a 60-s time-estimation task with eyes closed, starting and stopping a stopwatch when they judged that 60 s had elapsed. We computed absolute error (|recorded − 60|in seconds).

We administered the heartbeat-tracking task following Schandry (1981). Participants silently counted heartbeats during a 60-s interval without palpation, while a fingertip pulse oximeter recorded actual beats. Interoceptive accuracy was computed as|counted − recorded|(lower values indicate better accuracy).

General tendency to visual discomfort was assessed using the Japanese Visual Discomfort Scale (VDS-J; 23 items, 0–3); the Japanese version shows a unidimensional Rasch structure, Rasch person reliability of approximately 0.82, good test–retest stability, and construct validity (higher scores reported in migraine) ( Imaizumi et al., 2018). Discomfort to clustered patterns was assessed with the Japanese Trypophobia Questionnaire (TQ-J; 17 items, 1–5), which demonstrates a one-factor structure and adequate internal consistency and test–retest reliability in Japanese adults ( Imaizumi et al., 2016).

We measured both state and trait anxiety using the simplified State-Trait Anxiety Inventory (STAI; Koizumi et al., 1998) to assess anxiety.

These scales provided detailed information about participants’ sensory and psychological profiles. All participants completed three questionnaires in a separate bright and quiet room, and each questionnaire score was recorded separately for further analysis. We summarized all measures in Tables 1A, 1B and the essential task settings and scoring rules are reported below.

 Statistical analysis

To explore the relationship between the average preferred speed (APS) (i.e., the mean of the six speeds that participants identified as most comfortable) for RDM stimulation and factors such as interoceptive sensitivity (measured by the absolute difference between perceived and actual heart rates), subjective time perception (measured by actual time and the absolute difference between estimated time and actual time), visual discomfort levels (measured by the VDS-J and TQ-J scores), and anxiety levels (measured by the STAI scores), we conducted correlation analyses. Prior to inference, we screened the data for entry errors and outliers and inspected Q–Q plots and Shapiro–Wilk tests; several variables (e.g. Absolute Difference between Estimated Time and Actual Time, Absolute Difference between the Perceived and Actual Heart Rate) departed from normality. Accordingly, Spearman’s rank correlations were used for bivariate associations. We computed Spearman’s rs and 95% confidence intervals (CI) via bootstrap (5,000 resamples). We examined the correlation between the APS for RDM stimulation and several variables. All statistical analyses were conducted using two-tailed tests, with the significance level set at α = 0.05.

While the Spearman correlation analysis revealed significant relationships between certain variables and the APS, this analysis examined only bivariate associations. To examine the combined effects of multiple variables on APS, we conducted a multiple regression analysis. This analysis identified factors that significantly predicted APS, providing a deeper understanding of the underlying mechanisms influencing participants’ comfort levels.

All analyses were conducted using IBM SPSS Statistics v27 (64-bit) via the graphical interface; no custom code was written. To facilitate reuse, we provide de-identified data files (.xlsx and .sav), complete SPSS output files (.spv) containing the exact procedures and parameter settings and a step-by-step analysis recipe (PDF) that enumerates each transformation and test from raw variables to the reported tables and figures.

 Results Descriptive statistics

Table 2 summarizes means, standard deviations, ranges, and distribution shape (skewness, kurtosis) for all variables (N = 50). APS showed wide inter-individual variability (M = 249.47, SD = 116.82; range = 50.00–633.33). TimeDiff and the visual-discomfort measures (VDS-J, TQ-J) were positively skewed, whereas other variables were approximately symmetric; accordingly, bivariate associations used Spearman’s rs, and regression inference relied on residual diagnostics.

Descriptive statistics for all variables.

Variable N Mean SD Range Skewness Kurtosis
APS 50 249.47 116.82 50.00–633.33 0.95 1.40
AHR 50 79.14 9.20 60.00–102.00 0.28 −0.29
PHR 50 63.96 16.82 33.00–100.00 0.30 −0.62
HRDiff 50 17.46 13.66 1.00–44.00 0.69 −0.66
ST 50 67.07 11.84 39.56–102.91 0.44 0.78
TimeDiff 50 10.53 8.83 0.31–42.91 1.37 2.56
VDS-J 50 13.38 9.34 1.00–55.00 2.00 6.83
TQ-J 50 28.82 12.50 17.00–63.00 1.41 1.05
STAI total 50 21.18 5.41 10.00–31.00 −0.24 −0.93
Spearman correlation coefficients and p-values for variables relative to the average preferred speed of RDM stimulation

Spearman’s rank correlation analysis revealed significant correlations between APS and several variables at α = 0.05. A strong positive correlation ( rs = 0.49, 95% CI [.267,.660], p < .001) was found between the Absolute Difference between the Perceived and Actual Heart Rate (HRDiff) and the APS. This indicates that individuals with a greater discrepancy between their actual and perceived heart rates, reflecting lower interoceptive sensitivity, tended to perceive faster RDM stimuli as more comfortable. This finding underscores the relevance of interoceptive sensitivity in preferred RDM speed.

In addition, a negative correlation ( rs = -0.31, 95% CI [-.555, -.025], p = .030) was observed between the APS and the Perceived Heart Rate (PHR). This indicates that individuals who perceived a higher number of heartbeats tended to find slower speeds more comfortable, suggesting a preference for less-intense sensory input when interoceptive awareness is heightened.

Furthermore, the APS was positively correlated with the Standard Deviation of the Six Recorded Speeds (SD, rs = 0.49, 95% CI [.204, .709], p < .001). This suggests that individuals who preferred faster speeds also exhibited greater variability across the six trials, indicating a wider range of preferred RDM speeds. This finding was further supported by the significant positive correlation ( rs = 0.36, 95% CI [.108,.560], p = .011) observed between SD and HRDiff. This suggests that participants with lower heartbeat-tracking accuracy tended to show greater variability in preferred speeds across the six trials. This finding further highlights the association between interoceptive sensitivity and preferred RDM speed.

Finally, a positive correlation ( rs = 0.36, 95% CI [.060, .601], p = .011) was found between APS and STAI scores, reflecting participants’ anxiety levels during the experiment. This indicates that participants with higher STAI scores tended to perceive faster visual motion stimuli as more comfortable, suggesting that faster RDM stimulation speeds may be more effective in modulating anxiety levels.

However, Actual Heart Rate (AHR), Subjective Time (ST), Absolute Difference between Estimated Time and Actual Time (TimeDiff), and visual discomfort (VDS-J and TQ-J scores) did not show significant correlations with APS at α = 0.05. This suggests that while interoceptive sensitivity and anxiety levels play primary roles, other factors may influence the preferred speed to a lesser extent or in more nuanced ways. Table 3 presents the Spearman correlation coefficients and p-values for the relationships between the APS and various variables.

Spearman Correlation Coefficients and P-values for variables relative to the Average Preferred Speed (APS).
Variable Spearman Correlation Coefficient P-Value
Standard Deviation of Six Recorded Speeds 0.49 <.001 **
Actual Heart Rate 0.10 0.485
Perceived Heart Rate -0.31 0.030 *
Absolute Difference between Perceived and Actual Heart Rate 0.49 <.001 **
Subjective Time 0.10 0.478
Absolute Difference between Estimated Time and Actual Time 0.20 0.162
VDS-J Scores 0.01 0.963
TQ-J Scores 0.07 0.614
STAI Scores 0.31 0.028 *

The correlation is significant at the 0.05 level (2-tailed).

The correlation is significant at the 0.01 level (2-tailed).

 Multiple linear regression analysis

We specified a single multiple-regression model for APS with a limited, theory-driven set of candidate predictors (PHR, HRDiff, ST, TimeDiff, VDS-J, TQ-J, STAI). Because inference targeted the model as a whole and the number of pre-specified predictors was modest, we did not apply a formal multiplicity correction (two-sided α = .05). Instead, we emphasize effect sizes and 95% CI, consistent with guidance cautioning against routine Bonferroni-type adjustments in planned analyses ( Rothman, 1990; Perneger, 1998). To obtain a parsimonious subset we used forward stepwise selection (F-to-enter ≤ .05; F-to-remove ≥ .10) and interpret selected predictors cautiously given known limitations of automated selection ( Babyak, 2004; Derksen & Keselman, 1992; Whittingham et al., 2006). Analyses were conducted in SPSS (v27, 64-bit).

Model assumptions were checked and met, including linearity (residuals vs. fitted), independence (Durbin–Watson = 2.17), normality (skew = 0.37, kurtosis = 0.20), and homoscedasticity (residual plots).

The final model was significant, F (3, 46) = 7.14, p < .001, explaining 31.8% of the variance in APS (R 2 = .318, adjusted R 2 = .273). The detailed coefficients are presented in Table 4.

Multicollinearity was assessed using the variance inflation factor (VIF) for each predictor in the model. All VIF values were well below the threshold of 10, with a highest VIF value being 1.199. This indicates that multicollinearity is not a concern in the model, as each predictor variable exhibits low intercorrelation with the others. Additionally, no multiplicity correction was applied; p values are two-sided and should be interpreted descriptively alongside effect sizes and CIs ( Rothman, 1990; Perneger, 1998).

Our linear regression analysis revealed that three predictors significantly contributed to APS:

STAI: STAI scores positively predicted APS. This suggests that participants with higher STAI scores tend to prefer faster RDM stimulation speeds.

TimeDiff: TimeDiff was a significant positive predictor of APS. This suggests that participants with larger time-estimation errors (i.e., lower accuracy) prefer faster visual motion.

VDS-J: VDS-J scores negatively predicted APS. Higher visual discomfort (VDS-J) predicted lower APS (i.e., slower preferred speeds).

Multiple linear regression results for predicting Average Preferred Speed.
Predictor B SE B β t p VIF 95% CI for B
(Constant) 51.49 58.13 0.886 0.380
STAI 9.54 2.76 0.44 3.46 0.001 1.128 3.92 to 15.17
TimeDiff 4.77 1.69 0.36 2.83 0.007 1.094 1.38 to 8.16
VDS-J -4.06 1.68 -0.33 -2.42 0.019 1.199 -7.42 to -0.70

Note: N = 50. B, unstandardized regression coefficient; SE B, standard error of the coefficient; β = standardized coefficient; VIF, Variance Inflation Factor; STAI, State-Trait Anxiety Inventory; TimeDiff, Absolute Difference between Estimated Time and Actual Time; VDS-J, Visual Discomfort Scale Japanese version scores.

The standardized coefficients ( β) indicated that STAI scores ( β = 0.44) had the strongest relative influence on APS, followed by TimeDiff ( β = 0.36) and VDS-J scores ( β = -0.33).

The variance inflation factor (VIF) values for all predictors were close to 1, indicating no serious multicollinearity issues in the final model. Other variables, including HRDiff, PHR, AHR, ST, and TQ-J scores, did not contribute significantly to the prediction of APS in this model.

These findings provide insight into factors influencing preferred RDM speed. However, it is important to note that while the model explains a substantial portion of the variance in the APS (31.8%), unexplained variability remains, indicating that other factors not included in this model may also influence the preferred bubble tube speed.

 Discussion

This study aimed to investigate the relationship between various physiological and psychological factors and APS during RDM stimulation. Using Spearman’s correlation and multiple regression analyses, we identified several significant predictors of APS, providing insight into the interplay between interoceptive sensitivity, subjective time perception, visual discomfort levels, anxiety levels, and the preferred speeds for bubble tubes.

Our pattern of associations aligns with adjacent literatures on sensory processing, neurodivergent experience, and multi-sensory therapeutic environments. Sensory-integration frameworks propose individual differences in sensory thresholds and modulation that shape comfort under dynamic input ( Dunn, 2001; Lane et al., 2019). In autism research, atypical motion processing and heightened visual sensitivity can bias preferences toward slower, steadier dynamics when discomfort is high ( Robertson & Baron-Cohen, 2017). Clinical and community studies of multi-sensory environments further indicate that low-demand, user-controlled inputs reduce distress and improve engagement ( Champagne & Stromberg, 2004; Scanlan & Novak, 2015; Unwin et al., 2022, 2024). Framing bubble-tube speed as a design-tunable parameter therefore accords with theory and practice, while our data identify person-level predictors that can guide initial settings and individualized adjustments.

The significant positive correlation between the absolute difference between perceived and actual heart rate (HRDiff) and APS ( rs = 0.49, 95% CI [.238, .681], p < .001) underscores the pivotal role of interoceptive sensitivity in shaping the preferred RDM stimulation speeds. Interoceptive sensitivity and sensory processing are related via a clear mechanistic pathway. Enhanced interoceptive sensitivity enables more precise detection of internal physiological states, which, in turn, facilitates higher temporal resolution in sensory processing. This heightened internal awareness allows individuals to detect subtle physiological responses to sensory stimuli more accurately. Recent research by Grist et al. (2023) supports this connection by demonstrating that interoceptive awareness is correlated with sensory processing capabilities in neurotypical children, suggesting this relationship is fundamental to human development.

People with higher interoceptive sensitivity preferred slower visual motion, which may reflect two complementary mechanisms. First, reduced speed decreases information load and prevents sensory system overload. Second, slower presentation speeds provide extended processing windows, allowing for more thorough signal integration. This pathway enhances sensory comfort in two ways: by reducing sensory overload, and by improving processing precision. Slower information presentation lowers neural stress and fatigue, while better signal-to-noise ratios and prediction accuracy promote psychological ease. Together, these factors enable effortless engagement with the sensory environment, facilitating effective tracking and integration of stimuli for an optimal sensory experience.

The positive correlation between anxiety levels and APS ( rs = 0.31, 95% CI [.076, .582] p = .028), as confirmed by the multiple regression analysis ( β = 0.44, p = .001), suggests that faster RDM stimulation speeds may be preferred by individuals with higher anxiety levels. This finding is consistent with that of Park and Youn (2022) who found that high-intensity visual stimuli can replenish cognitive resources and reduce anxiety. Faster speeds may induce a state of physiological arousal that counteracts the heightened arousal associated with anxiety, thereby promoting relaxation and calmness.

The multiple regression analysis not only confirmed the independent contribution of anxiety levels to the APS, but also revealed the significant influence of subjective time perception. Although not significantly correlated with APS in the correlation analysis, the Absolute Difference between Estimated Time and Actual Time (TimeDiff) emerged as a predictor in the regression model ( β = 0.36, p = .007), suggesting that subjective time perception may interact with other factors to influence the preferred RDM stimulation speed in a complex manner. Individuals with lower subjective time-perception accuracy, in the same way as those with lower interoceptive sensitivity, may perceive faster RDM stimulation speeds as more comfortable. This finding suggests a potential link between time perception and interoceptive processing, as proposed by Di Lernia et al. (2018). Subjective time appears to covary with interoceptive processes: people with greater interoceptive accuracy may prefer slower visual motion, consistent with our observation that larger time-estimation errors were associated with faster preferred speeds. Furthermore, while visual discomfort did not show significant correlations with APS in the correlation analysis, the multiple regression analysis revealed additional effects. Specifically, higher VDS-J scores were significantly associated with lower APS ( β = -0.33, p = .019). These results indicate that although visual discomfort may not independently predict the APS, it contributes to the overall model of the preferred speed. This suggests that individuals with higher visual discomfort levels perceive slower speeds as more comfortable. This finding aligns with that of Penacchio et al. (2023) who suggested that slower visual stimulus speeds can alleviate discomfort in individuals with visual sensitivity. This emphasizes the importance of tailoring the sensory stimuli to individual comfort levels to maximize the therapeutic benefits of RDM stimulation.

These insights contribute to our understanding how physiological and psychological factors relate to preferred RDM speed and suggest potential pathways for designing sensory interventions. Our findings showed that the participants’ preferred RDM stimulation speeds varied according to their physiological and psychological factors, underscoring the need for personalized design in bubble tubes and sensory rooms.

Limitations

This study has several limitations that qualify the interpretation of the findings. First, the sample size (N = 50) constrains statistical power and external validity. Although the final regression model explained a meaningful proportion of variance, estimates may be unstable in small samples, and the precision of effects is limited. Replication with larger, prospectively powered cohorts is needed to refine effect sizes and improve generalizability.

Second, the multiple linear regression analysis used a stepwise variable-selection procedure after considering several candidate predictors. Data-driven selection can capitalize on chance and inflate Type I error, even when diagnostics such as VIF are satisfactory. Future work should confirm the model using a priori–specified predictors, cross-validation or a hold-out set, and consider regularized regression (e.g., ridge/LASSO) to reduce overfitting.

Third, the sample comprised a non-clinical convenience group. As such, the results may not generalize to autistic or other neurodivergent populations, nor to broader clinical settings. Multi-site studies that include autistic participants and other relevant user groups are required to establish transportability.

Fourth, several measures relied on self-report questionnaires (VDS-J, TQ-J, STAI), which are susceptible to response styles and situational influences. Moreover, time perception was indexed by a single 60-s estimation task, and interoception by the heartbeat-tracking task; both are single-method operationalizations with known variability. Future work should adopt multi-method batteries (e.g., heartbeat discrimination or respiratory interoception tasks; MAIA for sensibility) and report internal consistency/test–retest where applicable.

Fifth, regarding ecological validity, the upward random dot motion (RDM) stimulus isolates motion speed but does not reproduce the full chromatic, acoustic, and tactile dynamics of physical bubble tubes. Consequently, the preferred speeds identified here should be treated as laboratory proxies. Follow-up studies should verify whether preferences translate to device-controlled bubble tubes or video stimuli and to multi-sensory environments with user-controlled settings.

Finally, the study was exploratory and was not preregistered. While we mitigate this by sharing de-identified data, SPSS outputs, and a step-by-step analysis recipe, future research would benefit from preregistration, independent replication, and expanded physiological recording (e.g., HRV/EDA during stimulation) to test mechanistic accounts.

 Conclusions

This study provides insight into the interplay between physiological and psychological factors that influence preferred RDM stimulation speeds. Our findings reveal that interoceptive sensitivity, subjective time perception, visual discomfort, and anxiety are significant predictors.

These results demonstrate that individuals with lower interoceptive sensitivity tend to prefer faster RDM stimulation speeds, possibly as a compensatory mechanism for reduced internal bodily awareness. Observed associations between interoceptive sensitivity and preferred stimulation speed highlight the value of personalizable sensory environments—e.g., adjustable bubble-tube speeds—that account for individual physiological differences.

Anxiety levels emerged as a significant factor, with anxiety significantly and positively correlated with preferred RDM speed. This finding suggests that faster and more intense visual stimuli may have applications in anxiety management. Saiu and Grosso (2022) demonstrated that combined audio-visual stimulation significantly reduced anxiety levels, systolic blood pressure, and heart rate in patients undergoing surgery, supporting the potential of multi-sensory interventions in anxiety management.

Subjective time-perception ability was also related to the preferred RDM speed, with participants showing lower accuracy tending to prefer faster speeds. Previous research has shown that interoceptive focus significantly influences subjective time perception, with heightened interoceptive awareness amplifying the time-dilating effects of fear and the time-accelerating effects of amusement ( Pollatos et al., 2014). This, in turn, may have affected participants’ perceptions of comfort in response to RDM stimulation speeds.

Although visual discomfort was not independently correlated with preferred speed in the Spearman correlation analysis, it emerged as a significant predictor in the multiple regression analysis. Specifically, participants with higher visual sensitivity tended to prefer slower speeds. This finding highlights the importance of designers considering individual levels of visual discomfort when designing visual stimuli in bubble tubes or sensory rooms and offering tailored sensory experiences that cater to each person’s diverse sensory needs.

As Ziegler (2015) noted in her research on multi-sensory design in healthcare settings, environments that allow users to customize sensory inputs contribute to stress reduction and improved well-being. Garzotto et al. (2019) developed an extensible multilayer software and hardware platform to connect and manage different devices in a sensory room, thereby enabling therapists to fully customize activities in multi-sensory environments for different children. These findings suggest that personalizable sensory environments—for example, installations with adjustable bubble-tube speeds—may better accommodate diverse sensory needs and support user comfort and self-regulation. Building on our previous research ( Su et al., 2024), which demonstrated that sensory-hypersensitive individuals have greater variations in their preferences for wallpaper colors and patterns, this study further underscores the necessity for customizable sensory rooms tailored to each person’s unique sensory needs.

These findings should be interpreted in light of the study’s limitations. A key limitation is the lack of clinical characterization: participants were not screened for neurodevelopmental or anxiety disorders, and medication status was not recorded. Consequently, generalizability to autistic or clinically anxious populations is uncertain.

In conclusion, this study enhances our understanding of the factors influencing preferred bubble tube speeds and lays the groundwork for the development of more personalized and effective sensory interventions to improve outcomes for individuals with diverse sensory needs.

Ethical considerations

This study recruited 50 adult participants (27 females) aged 22–35 years (M = 25.94, SD = 2.74) from the University of Tsukuba between December 25, 2023, and March 22, 2024. The research was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of the Institute of Art and Design, the University of Tsukuba (IRB No. [GEI021-15]) on March 22, 2022. Prior to participation, all participants provided written informed consent. To ensure privacy, all participant data were anonymized, and no identifying information was collected or retained.

Written informed consent for publication of the participants details was obtained from the participants.

 Data availability

All data, materials, and software are available under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license.

Underlying data

Zenodo: Preferred Bubble Tube Speed and Physiological–Psychological Factors. DOI: 10.5281/zenodo.14633771 ( Su, A., 2025b).

De-identified participant-level dataset (N = 50) containing trial-level and summary variables used in the analyses.

 Analysis outputs and protocol

Zenodo: Preferred Bubble Tube Speed & Physiological and Psychological Factors — SPSS Outputs and Analysis Recipe (v1.0). DOI: 10.5281/zenodo.17045820 ( Su, A., 2025c).

 Reporting guidelines

Zenodo: Checklist for “Exploring the relationship between preferred bubble tube speeds in sensory rooms and physiological–psychological factors: A study on interoceptive sensitivity, subjective time perception, visual discomfort levels, and anxiety levels”. DOI: 10.5281/zenodo.14633918 ( Su, A., 2025a). License: CC-BY 4.0.

 Extended data

Zenodo: Scales used to assess participants’ levels of visual discomfort and anxiety. DOI: https://doi.org/10.5281/zenodo.14770842 ( Su, A., 2025f).

 Software

GitHub: https://github.com/Lem0n-SAJ/RDM_Program_Su

Zenodo (source code): https://doi.org/10.5281/zenodo.14795461 ( Su, A, 2025d)

Zenodo (Windows executable): https://doi.org/10.5281/zenodo.14795194 ( Su, A, 2025e)

License: MIT

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

 References Åhs F Miller S Gordon AR : Aversive learning increases sensory detection sensitivity. Biol. Psychol. 2013;92:135–141. 23174695 10.1016/j.biopsycho.2012.11.004 Babyak MA : What you see may not be what you get: a brief, nontechnical introduction to overfitting in regression-type models. Psychosom Med. 2004;66(3):411–421. 10.1097/01.psy.0000127692.23278.a9 Bakker K Moulding R : Sensory-processing sensitivity, dispositional mindfulness, and negative psychological symptoms. Personal. Individ. Differ. 2012;53(4):341–346. 10.1016/j.paid.2012.04.006 Barbic SP Chan N Rangi A : Health provider and service-user experiences of sensory modulation rooms in an acute inpatient psychiatry setting. PLoS One. 2019;14(11):e0225238. 31751373 10.1371/journal.pone.0225238 PMC6874072 Bellini EV Macchi A Setola N : Sensory design in the birth environment: Learning from existing case studies. Buildings. 2023;13(3):604. 10.3390/buildings13030604 Blakemore S-J Tavassoli T Calò S : Tactile sensitivity in Asperger syndrome. Brain Cogn. 2006;61(1):5–13. 16500009 10.1016/j.bandc.2005.12.013 Breslin L Guerra N Ganz L : Clinical utility of multisensory environments for people with intellectual and developmental disabilities: A scoping review. Am. J. Occup. Ther. 2020;74(1):7401205060p1–7401205060p12. 32078517 10.5014/ajot.2020.037267 PMC7018453 Britten KH Shadlen MN Newsome WT : The analysis of visual motion: a comparison of neuronal and psychophysical performance. J. Neurosci. 1992;12(12):4745–4765. 1464765 10.1523/JNEUROSCI.12-12-04745.1992 PMC6575768 Castellotti S D’Agostino O Biondi A : Influence of motor and cognitive tasks on time estimation. Brain Sci. 2022;12(3):404. 35326362 10.3390/brainsci12030404 PMC8946194 Champagne T Stromberg N : Sensory approaches in inpatient psychiatric settings: Innovative alternatives to seclusion & restraint. J. Psychosoc. Nurs. Ment. Health Serv. 2004;42(9):34–44. 15493494 10.3928/02793695-20040901-06 Cohen J :1988; Statistical power analysis for the behavioral sciences. 2nded: Routledge. Conlon E Humphreys L : Visual search in migraine and visual discomfort groups. Vis. Res. 2001;41(23):3063–3068. 11704243 10.1016/S0042-6989(01)00204-8 Critchley HD Garfinkel SN : Interoception and emotion. Curr. Opin. Psychol. 2017;17:7–14. 10.1016/j.copsyc.2017.04.020 Derksen S Keselman HJ : Backward, forward and stepwise automated subset selection algorithms: frequency of obtaining authentic and noise variables. Br. J. Math Stat. Psychol. 1992;45(2):265–282. 10.1111/j.2044-8317.1992.tb00992.x Di Lernia D Serino S Pezzulo G : Feel the time: Time perception as a function of interoceptive processing. Front. Hum. Neurosci. 2018;12(74):29559902. 29559902 10.3389/fnhum.2018.00074 PMC5845687 Droit-Volet S Meck WH : How emotions colour our perception of time. Trends Cogn. Sci. 2007;11(12):504–513. 10.1016/j.tics.2007.09.008 Dunn W : The sensations of everyday life: Empirical, theoretical, and pragmatic considerations. Am J Occup Ther. 2001;55(6):608–620. 12959225 10.5014/ajot.55.6.608 Engel-Yeger B Muzio C Rinosi G : Extreme sensory processing patterns and their relation with clinical conditions among individuals with major affective disorders. Psychiatry Res 2016;236:112–118. 26738981 10.1016/j.psychres.2015.12.022 Evans BJW Stevenson SJ : The Pattern Glare Test: a review and determination of normative values. Ophthalmic Physiol. Opt. 2008;28(4):295–309. 18565084 10.1111/j.1475-1313.2008.00578.x Eysenck MW Derakshan N Santos R : Anxiety and cognitive performance: Attentional control theory. Emotion. 2007;7(2):336–353. 10.1037/1528-3542.7.2.336 Fujisawa K : The effects of a multi-sensory environment using bubble tubes on physiological indicators of students with intellectual disabilities and students without disabilities, and the significance of developmental support. Jpn. J. Hum. Growth Dev. Res. 2021;2021(90):37–45. 10.5332/hatsuhatsu.2021.90_37 Garfinkel SN Critchley HD : Interoception, emotion and brain: new insights linking internal physiology to social behaviour. Curr. Opin. Neurobiol. 2013;8(3):231–234. 23482658 10.1093/scan/nss140 PMC3594730 Garfinkel SN Seth AK Barrett AB : Knowing your own heart: distinguishing interoceptive accuracy from interoceptive awareness. Biol. Psychol. 2015;104:65–74. 10.1016/j.biopsycho.2014.11.004 Garzotto F Gelsomini M Gianotti M : Engaging children with neurodevelopmental disorder through multisensory interactive experiences in a smart space. Soro A Brereton M Roe P , editors. Social Internet of Things . Springer;2019; pp.167–184, 10.1007/978-3-319-94659-7_9 Green SA Ben-Sasson A : Anxiety disorders and sensory over-responsivity in children with autism spectrum disorders: Is there a causal relationship? J Autism Dev. Disord. 2010;40(12):1495–1504. 20383658 10.1007/s10803-010-1007-x PMC2980623 Green SA Ben-Sasson A Soto TW : Anxiety and sensory over-responsivity in toddlers with autism spectrum disorders: Bidirectional effects across time. J. Autism Dev. Disord. 2012;42(6):1112–1119. 21935727 10.1007/s10803-011-1361-3 PMC4199633 Grist N Brown T Yu ML : An exploration of the association between sensory processing and interoceptive awareness in neurotypical school-aged children. J. Occup. Ther. Sch. Early Interv. 2023:1–20. 10.1080/19411243.2023.2215759 Haig S Hallett N : Use of sensory rooms in adult psychiatric inpatient settings: A systematic review and narrative synthesis. Int. J. Ment. Health Nurs. 2023;32(1):54–75. 36082841 10.1111/inm.13065 Hannant P Tavassoli T Cassidy S : The role of sensorimotor difficulties in autism spectrum conditions. Front. Neurol. 2016;7:124. 27559329 10.3389/fneur.2016.00124 PMC4978940 Hulsegge J Verheul A :2005; Snoezelen: Another world. A practical book explaining the basic elements of sensory development over 30 years of Snoezelen practice. reviseded: Rompa. Imaizumi S Furuno M Hibino H : Development of the Japanese version of trypophobia questionnaire. Jpn. J. Personal. 2016;25(2):171–173. 10.2132/personality.25.171 Imaizumi S Koyama S Tanno Y : Development of the Japanese version of the Visual Discomfort Scale. PLoS One. 2018;13(1):e0191094. 29324819 10.1371/journal.pone.0191094 PMC5764345 Khalfa S Bruneau N Rogé B : Increased perception of loudness in autism. Hear. Res. 2004;198(1-2):87–92. 15617227 10.1016/j.heares.2004.07.006 Kim JA Szatmari P Bryson SE : The prevalence of anxiety and mood problems among children with autism and Asperger syndrome. Autism. 2000;4(2):117–132. 10.1177/1362361300004002002 Koegel RL Openden D Koegel LK : A systematic desensitization paradigm to treat hypersensitivity to auditory stimuli in children with autism in family contexts. Res. Pract. Pers. Sev. Disabil. 2004;29(2):122–134. 10.2511/rpsd.29.2.122 Koizumi N Fujita D Ninomiya R : Screening test by statistically reducing the number of the State-Trait Anxiety Inventory (STAI) items. Sangyo Eiseigaku Zasshi. 1998;40(4):107–112. 9777671 10.1539/sangyoeisei KJ00001990548 Koldewyn K Whitney D Rivera SM : The psychophysics of visual motion and global form processing in autism. Brain. 2010;133(2):599–610. 19887505 10.1093/brain/awp272 PMC2858014 Landon J Shepherd D Lodhia V : A qualitative study of noise sensitivity in adults with autism spectrum disorder. Res. Autism Spectr. Disord. 2016;32:43–52. 10.1016/j.rasd.2016.08.005 Lane SJ Mailloux Z Schoen S : Neural foundations of Ayres Sensory Integration®. Brain Sci. 2019;9(7):153. 10.3390/brainsci9070153 Lucker JR : Auditory hypersensitivity in children with autism spectrum disorders. Focus Autism Other Dev. Disabil. 2013;28(3):184–191. 10.1177/1088357613475810 MacLennan K Roach L Tavassoli T : The relationship between sensory reactivity differences and anxiety subtypes in autistic children. Autism Res. 2020;13(5):785–795. 31909874 10.1002/aur.2259 MacLennan K Rossow T Tavassoli T : The relationship between sensory reactivity, intolerance of uncertainty and anxiety subtypes in preschool-age autistic children. Autism. 2021;25(8):2305–2316. 34030518 10.1177/13623613211016110 Mazurek MO Vasa RA Kalb LG : Anxiety, sensory over-responsivity, and gastrointestinal problems in children with autism spectrum disorders. J. Abnorm. Child Psychol. 2013;41:165–176. 22850932 10.1007/s10802-012-9668-x Meissner K Wittmann M : Body signals, cardiac awareness, and the perception of time. Biol. Psychol. 2011;86(3):289–297. 21262314 10.1016/j.biopsycho.2011.01.001 Milne E Swettenham J Hansen P : High motion coherence thresholds in children with autism. J. Child Psychol. Psychiatry. 2002;43(2):255–263. 11902604 10.1111/1469-7610.00018 Newsome WT Paré EB : A selective impairment of motion perception following lesions of the middle temporal visual area (MT). J. Neurosci. 1988;8(6):2201–2211. 3385495 10.1523/JNEUROSCI.08-06-02201.1988 PMC6569328 Park Y Youn N : Social distancing and prosocial behaviors: The replenishing effect of intense visual stimulation. J. Bus. Res. 2022;149:983–996. 36407286 10.1016/j.jbusres.2022.05.086 PMC9652121 Penacchio O Otazu X Wilkins AJ : A mechanistic account of visual discomfort. Front. Neurosci. 2023;17:1200661. 37547142 10.3389/fnins.2023.1200661 PMC10397803 Penacchio O Wilkins AJ : Visual discomfort and the spatial distribution of Fourier energy. Vision Res. 2015;108:1–7. 25576380 10.1016/j.visres.2014.12.013 Perneger TV : What’s wrong with Bonferroni adjustments. BMJ. 1998;316(7139):1236–1238. 9553006 10.1136/bmj.316.7139.1236 PMC1112991 Pollatos O Füstös J Critchley HD : On the generalised embodiment of pain: How interoceptive sensitivity modulates cutaneous pain perception. Pain. 2012;153(8):1680–1686. 22658270 10.1016/j.pain.2012.04.030 Pollatos O Laubrock J Wittmann M : Interoceptive focus shapes the experience of time. PLoS One. 2014;9(1):e86934. 24489807 10.1371/journal.pone.0086934 PMC3906083 Robertson CE Baron-Cohen S : Sensory perception in autism. Nat. Rev. Neurosci. 2017;18(11):671–684. 10.1038/nrn.2017.112 Rothman KJ : No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):43–46. 10.1097/00001648-199001000-00010 Saiu S Grosso E : Controlled audio-visual stimulation for anxiety reduction. Comput. Methods Prog. Biomed. 2022;223:106898. 35780520 10.1016/j.cmpb.2022.106898 Salzman CD Murasugi CM Britten KH : Microstimulation in visual area MT: effects on direction discrimination performance. J. Neurosci. 1992;12(6):2331–2355. 1607944 10.1523/JNEUROSCI.12-06-02331.1992 PMC6575906 Sarigiannidis I Grillon C Ernst M : Anxiety makes time pass quicker while fear has no effect. Cognition. 2020;197:104116. 31883966 10.1016/j.cognition.2019.104116 PMC7033556 Scanlan JN Novak T : Sensory approaches in mental health: A scoping review. Aust. Occup. Ther. J. 2015;62(5):277–285. 26358506 10.1111/1440-1630.12224 Schandry R : Heartbeat perception and emotional experience. Psychophysiology. 1981;18(4):483–488. 7267933 10.1111/j.1469-8986.1981.tb02486.x Spielberger CD :1983; Manual for the State-Trait Anxiety Inventory (Form Y) Palo Alto (CA): Consulting Psychologists Press. Su A Shen J Zhang X : Greater variance in wallpaper color evaluation in users with visual sensory hypersensitivity revealed in a virtual reality experiment. J. Sci. Des. 2024;8(1):21–30. Su A : A completed STROBE checklist. Zenodo . 2025a. 10.5281/zenodo.14633918 Su A : Preferred bubble tube speed and physiological–psychological factors. [Data set]. Zenodo. 2025b. 10.5281/zenodo.14633771 SU A : Preferred Bubble Tube Speed & Physiological–Psychological Factors — SPSS Outputs and Analysis Recipe (v1.0) [Data set]. Zenodo. 2025c. 10.5281/zenodo.17045820 Su A : Random Dot Motion (RDM) experiment program Source Code (V1.0). Zenodo. 2025d. 10.5281/zenodo.14795461 Su A : Random Dot Motion Experiment Program – Windows Executable. Zenodo. 2025e. 10.5281/zenodo.14795194 Su A : Scales used to assess participants’ levels of visual discomfort and anxiety (Japanese version used in research and English version for reference). Zenodo. 2025f. 10.5281/zenodo.14770842 Sutton D Wilson M Van Kessel K : Optimizing arousal to manage aggression: A pilot study of sensory modulation. Int. J. Ment. Health Nurs. 2013;22(6):500–511. 23374543 10.1111/inm.12010 Tanaka Y Terasawa Y Umeda S : Effects of interoceptive accuracy in autonomic responses to external stimuli based on cardiac rhythm. PLoS One. 2021;16(8):e0256914. 34464424 10.1371/journal.pone.0256914 PMC8407568 Unwin KL Powell G Price A : The use of multi-sensory environments with autistic children: Exploring the effect of having control of sensory changes. Autism. 2022;26(6):1379–1394. 34693744 10.1177/13623613211050176 PMC9340127 Unwin KL Powell G Price A : Patterns of equipment use for autistic children in multi-sensory environments: Time spent with sensory equipment varies by sensory profile and intellectual ability. Autism. 2024;28(3):644–655. 37421130 10.1177/13623613231180266 PMC10913295 Van der Hallen R Manning C Evers K : Global motion perception in autism spectrum disorder: a meta-analysis. J. Autism Dev. Disord. 2019;49(12):4901–4918. 31489542 10.1007/s10803-019-04194-8 PMC6841654 Van Weert JCM Van Dulmen AM Spreeuwenberg PMM : Behavioral and mood effects of Snoezelen integrated into 24-hour dementia care. J. Am. Geriatr. Soc. 2005;53(1):24–33. 15667372 10.1111/j.1532-5415.2005.53006.x Verhulst I MacLennan K Haffey A : The perceived causal relations between sensory reactivity differences and anxiety symptoms in autistic adults. Autism Adulthood. 2022;4(3):183–192. 36606154 10.1089/aut.2022.0018 PMC9648696 Vicario CM Nitsche MA Salehinejad MA : Time processing, interoception, and insula activation: a mini-review on clinical disorders. Front. Psychol. 2020;11:1893. 32973605 10.3389/fpsyg.2020.01893 PMC7461974 West M Melvin G McNamara F : An evaluation of the use and efficacy of a sensory room within an adolescent psychiatric inpatient unit. Aust. Occup. Ther. J. 2017;64(3):253–263. 28138979 10.1111/1440-1630.12358 Whittingham MJ Stephens PA Bradbury RB : Why do we still use stepwise modelling in ecology and behaviour? J. Anim. Ecol. 2006;75(5):1182–1189. 16922854 10.1111/j.1365-2656.2006.01141.x Wier CC Jesteadt W Green DM : A comparison of method-of-adjustment and forced-choice procedures in frequency discrimination. Percept. Psychophys. 1976;19(1):75–79. 10.3758/BF03198701 Wiglesworth S Farnworth L : An exploration of the use of a sensory room in a forensic mental health setting: Staff and patient perspectives. Occup. Ther. Int. 2016;23(3):255–264. 27237722 10.1002/oti.1428 Wilkins A : A physiological basis for visual discomfort: Application in lighting design. Light. Res. Technol. 2016;48(1):44–54. 10.1177/1477153515612526 Yabe Y Yamada S : A pilot study of how the past, present, and future are represented in three-dimensional space. Front. Psychol. 2023;14:1071917. 37034943 10.3389/fpsyg.2023.1071917 PMC10075084 Yoshimoto S Jiang F Takeuchi T : Adaptation and visual discomfort from flicker. Vision Res. 2019;160:99–107. 31091424 10.1016/j.visres.2019.04.010 PMC7153578 Ziegler U : Multi-sensory design as a health resource: Customizable, individualized, and stress-regulating spaces. Des. Issues. 2015;31(1):53–62. 10.1162/DESI_a_00309 10.5256/f1000research.195119.r474368 Reviewer response for version 2 Haigh Sarah M 1 Referee https://orcid.org/0000-0003-2400-4412 University of Nevada Reno, Reno, Nevada, USA

Competing interests: No competing interests were disclosed.

 12 6 2026 Copyright: © 2026 Haigh SM 2026 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

I can see that revisions have already been made in response to comments and so I will endeavor not to add contradiction.

The topic is interesting and highly relevant based on the current conversation regarding understanding sensory sensitivity in clinical and nonclinical populations and the use of sensory rooms for those who need them. However, I have a couple of questions and suggestions for clarifying the study.

The introduction is focused on clinical (particularly autistic) populations and yet the sample recruited were non-clinical and no subclinical measures related to autism or other conditions were collected, except for anxiety. This in itself would be fine, but there were no checks to see if any of the participants had a diagnosis, which throws into question whether the sample were non-clinical and if there were any medication confounds. Is there any way the authors can address this concern? I see it is included towards the end of the discussion but does not resolve the use of these data when understanding the variability in preferred bubble speed.

The introduction introduces multiple components and currently reads a little choppy. Subheadings or a light introduction into the topics to be covered might help readability (minor concern). May also help to highlight the variability in some of the scales (anxiety, visual discomfort etc.) in the general population that tends to be more extreme in clinical populations, making the non-clinical/convenience sample for this study seem more appropriate. In other words, taking a course look at variability in preferred speed in relation to a variety of scales without subdividing would be representation of the typical populations walking through an airport (for example) and may want to use a sensory room.

For methods, please clarify which measures were recorded once and why. Are there concerns about task uncertainty affecting results? The counting tasks do appear to be a bit unusual. Why these over more standard temporal precision tasks?

Highlighting what the direction of the DVs mean consistently would help. For example, counting 60 seconds was calculated as absolute difference with actual, why? Negative values would also be interesting. Closer to zero means more accurate, correct? Same for the heartbeat counting task, absolute measures only?

Why trypophobia? Comes out of nowhere and there is a debate as to the relationship with visual discomfort. Can the authors add a description of the questionnaires to give the reader a better understanding of what is being asked? Some measures, for example of visual discomfort, only ask about one component (reading or light sensitivity) whereas others are more comprehensive. This ensures that the reader does not have to read the citations to understand what is going on here.

Seem to be multiple comparisons being conducted - any correction for Type 1 error? Regression analysis only reports significant effects but more variables were included in the model? Unclear as to how the limited factors were decided upon. 

Scatterplots showing the data would be helpful to illustrate relationships. Were measures related to one another?

Minor point in discussion: Penacchio et al. (2023) reference is stated to say something about motion and visual discomfort. As far as I am aware, it does not. Can the authors clarify what they are referring to?

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

Visual discomfort

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

 10.5256/f1000research.177078.r375957 Reviewer response for version 1 Savickaite Sarune 1 Referee University of Exeter, Exeter, UK

Competing interests: No competing interests were disclosed.

 22 8 2025 Copyright: © 2025 Savickaite S 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

This study explored how individuals’ physiological and psychological characteristics affect their preferred speeds of bubble tube stimulation in sensory rooms, using a simulated upward-moving visual task. Results showed that people with higher interoceptive sensitivity preferred slower speeds, while those with higher anxiety or less accurate time perception preferred faster speeds. Visual discomfort also played a role, with those experiencing more discomfort tending to prefer slower visual motion. A combination of interoceptive accuracy, anxiety levels, time estimation ability, and visual sensitivity significantly predicted preferred stimulation speed. The findings highlight the importance of personalized sensory environments tailored to individual needs for optimal therapeutic impact.

1. Terminology.  Use “autism” instead of “ASD” throughout for a more inclusive and widely accepted terminology, especially in line with neurodiversity research.

2. Avoid Ableist Language. Refrain from using medicalized or deficit-based terms such as “impairment.” Instead, adopt terminology grounded in the social model of disability, which focuses on societal barriers rather than individual deficits.

3. Identity-First Language. Use “autistic individuals” rather than “individuals with autism” to align with current preferences within the autistic community and neurodiversity literature.

4. Improve Rationale Structure. Streamline the background by clearly introducing the key constructs (e.g., interoception, time perception, anxiety, visual discomfort), then presenting a focused rationale for the current study. Avoid repeating the aim multiple times.

5. Reference Gaps. In the paragraph starting “However, the precise mechanisms by which the bubble tubes induce relaxation remain poorly understood…”, add references to prior sensory integration or multi sensory environments research to contextualize the claim.

6. Existing Research. Although this is a novel area, relevant adjacent literature does exist. Draw clearer parallels to studies in sensory processing, neurodivergent experience, and multi-sensory therapeutic environments.

7. Conceptual Clarity

Interoception. Define which aspects of interoception are relevant (e.g., accuracy, awareness, sensibility), and clarify why they matter for this context. Reference foundational literature (e.g., Garfinkel et al., 2015; Critchley & Garfinkel, 2017).

Time Perception. Explicitly explain how time perception relates to interoception and anxiety in this study. Currently, it appears disconnected.

Anxiety and Visual Discomfort. Define anxiety more clinically (e.g., state vs. trait), and expand discussion of visual discomfort with more references. Both constructs are complex and under-contextualized in the paper.

8. Participant Characteristics. Clarify whether participants were asked about clinical diagnoses (e.g., anxiety, neurodevelopmental conditions), given their relevance to the research aims. If not assessed, provide a justification.

9. Stimuli Choice. Justify the use of Random Dot Motion (RDM) stimuli, particularly in the context of interoception and neurodiversity research, where its utility is not well-established.

10. Measures.  Provide a clearer, consolidated presentation of the scales and tasks used to measure anxiety, interoception, time perception, and visual discomfort—ideally in a dedicated table with psychometric properties and validation references.

11. Open Science Considerations. While SPSS is a common statistical tool, open science practices would benefit from clear data and protocol sharing. Explicitly state where readers can access the materials.

12. Statistical Rationale. Explain why Spearman’s correlation was used (e.g., non-normal distribution), and clarify whether regression models were corrected for multiple comparisons. Justify the model selection process more clearly.

13. Sample Size Concerns. The relatively small sample (N=50) may limit the power and generalizability of the regression analysis. Discuss limitations related to sample size and possible risks of overfitting or underpowered findings.

14. Streamline Results. Reporting of statistics (e.g., means, SDs, p-values, CI) could be streamlined for clarity. Use visual aids (tables or figures) to consolidate key outcomes for easier interpretation.

15. Limitations Section: Explicitly address the study’s limitations, including sample size, generalizability, reliance on self-report measures, and the exploratory nature of the design. This is currently underdeveloped and needed for transparency.

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

No

Is the study design appropriate and is the work technically sound?

Partly

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

I work with neurodiversity, perceptual differences and interoception. This is my area of expertise and this article requires major revisions.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

References : Discrepancies between dimensions of interoception in autism: Implications for emotion and anxiety. Biological Psychology .2016;114: 10.1016/j.biopsycho.2015.12.003 117-126 10.1016/j.biopsycho.2015.12.003 : Interoception and emotion. Current Opinion in Psychology .2017;17: 10.1016/j.copsyc.2017.04.020 7-14 10.1016/j.copsyc.2017.04.020 Koyama Shinichi Institute of Art and Design, University of Tsukuba, Tsukuba, Ibaraki Prefecture, Japan

Competing interests: No competing interests were disclosed.

 14 1 2026

Response to reviewer 2.

Thank you very much for your valuable comments.

1. Terminology.  Use “autism” instead of “ASD” throughout for a more inclusive and widely accepted terminology, especially in line with neurodiversity research.

=> We replaced ASD with autism throughout the paper. We also replaced autism disorder-based terminology and sensory processing with neurodiversity-affirming language.

2. Avoid Ableist Language. Refrain from using medicalized or deficit-based terms such as “impairment.” Instead, adopt terminology grounded in the social model of disability, which focuses on societal barriers rather than individual deficits.

=> We removed medicalized/deficit-based phrasing and reframed descriptions using the social model of disability.

   address these challenges => provide sensory support (Page 2, Line 48)

   functional abilities => agency within everyday environments (Page 2, Line 50)

   history of aggression => experience heightened distress responses (Page 3, Line 67)

3. Identity-First Language. Use “autistic individuals” rather than “individuals with autism” to align with current preferences within the autistic community and neurodiversity literature.

=> We now use autistic children/adults/people/populations/participants rather than individuals with autism, except when citing instruments or studies that use diagnostic labels (Pages 3, 5, 7, 13).

4. Improve Rationale Structure. Streamline the background by clearly introducing the key constructs (e.g., interoception, time perception, anxiety, visual discomfort), then presenting a focused rationale for the current study. Avoid repeating the aim multiple times.

=> We revised the Introduction to follow a four-step structure: background, mechanisms/needs, key constructs (interoception, time perception, anxiety, and visual discomfort), and presentation of the present study and hypotheses. We streamlined the rationale and removed duplicated statements of the aim to ensure a clearer progression (Pages 2-5).

5. Reference Gaps. In the paragraph starting “However, the precise mechanisms by which the bubble tubes induce relaxation remain poorly understood…”, add references to prior sensory integration or multi-sensory environments research to contextualize the claim.

=> We added key citations (Champagne & Stromberg, 2004; Sutton et al., 2013; Scanlan & Novak, 2015) (Page 3, Line 83) to contextualize multisensory modulation mechanisms.

6. Existing Research. Although this is a novel area, relevant adjacent literature does exist. Draw clearer parallels to studies in sensory processing, neurodivergent experience, and multi-sensory therapeutic environments.

=> We strengthened the parallels to existing research in sensory processing, neurodivergent experiences, and multi-sensory environments. Specifically, we incorporated evidence from observational studies and systematic reviews (e.g., Breslin et al., 2020, Page 3, Lines 86-89; Haig & Hallett, 2023, Page 3, Lines 67-71) to contextualize the use of sensory equipment and highlight the variable outcomes across different populations. These additions underscore the necessity of identifying person-level moderators, such as motion processing differences in autistic individuals, to better understand engagement in therapeutic spaces.

7. Conceptual Clarity

Interoception. Define which aspects of interoception are relevant (e.g., accuracy, awareness, sensibility), and clarify why they matter for this context. Reference foundational literature (e.g., Garfinkel et al., 2015; Critchley & Garfinkel, 2017). Time Perception. Explicitly explain how time perception relates to interoception and anxiety in this study. Currently, it appears disconnected.

Anxiety and Visual Discomfort. Define anxiety more clinically (e.g., state vs. trait), and expand discussion of visual discomfort with more references. Both constructs are complex and under-contextualized in the paper.

=> We expanded the conceptual definitions of interoception in the Introduction (Page 4, Line 90). We justified our specific focus on interoceptive accuracy as an objective index (Page 4, Line 94).

=> Beyond interoception, we also strengthened the theoretical framework by:

For interoception, we now explicitly define accuracy, sensibility, and awareness, citing foundational literature (Garfinkel et al., 2015; Critchley & Garfinkel, 2017). Subjective Time Perception: linking it to interoception and anxiety (Page 4). Visual Discomfort: Elaborating on the mechanisms of pattern-glare and the impact of spatial/temporal statistics (Page 4). Anxiety: Clarifying the distinction between state and trait anxiety as measured by the STAI short form (Pages 4-5).

8. Participant Characteristics. Clarify whether participants were asked about clinical diagnoses (e.g., anxiety, neurodevelopmental conditions), given their relevance to the research aims. If not assessed, provide a justification.

=> We clarified that clinical diagnoses (such as anxiety or neurodevelopmental conditions) were not recorded, as the study was designed as a foundational step to investigate person-level predictors of sensory comfort within a general population sample. To address this, we justified this non-clinical focus in the Methods section (Page 6, Line 171) and added a corresponding acknowledgment in the Limitations section (Page 13, Line 419). In the latter, we further propose that future follow-up work should specifically involve autistic and other neurodivergent cohorts to validate these findings and enhance the generalizability of the results within clinical settings.

9. Stimuli Choice. Justify the use of Random Dot Motion (RDM) stimuli, particularly in the context of interoception and neurodiversity research, where its utility is not well-established.

=> We added a dedicated rationale for the use of RDM stimuli, emphasizing their utility in isolating the motion-speed component of bubble-tube viewing while minimizing confounding variables such as color, texture, and form. In the revised Introduction (Page 5, Line 148), we explicitly linked the RDM paradigm to canonical neurophysiological work in the MT/V5 area and its established application in neurodiversity research, particularly in assessing global motion processing in autistic populations. Furthermore, we have expanded the Methods section to clarify how the RDM provides precise parametric control over the upward motion speed while maintaining a constant dot density and luminance. These additions justify the translational value of RDM as a proxy for bubble-tube motion within the context of interoceptive and sensory-processing research.

10. Measures.  Provide a clearer, consolidated presentation of the scales and tasks used to measure anxiety, interoception, time perception, and visual discomfort—ideally in a dedicated table with psychometric properties and validation references.

=> In response to the request for a more consolidated presentation of our measures, we reorganized all experimental tasks and scales into two comprehensive tables (Page 23). Table 1A summarizes the experimental tasks (e.g., interoception and time perception), whereas Table 1B details the questionnaires used for anxiety and visual discomfort. These tables provide a clear overview of the key settings, scoring procedures, psychometric properties, and validation references with consistent citations. These changes significantly improve the transparency of our operational definitions and streamline the referencing of specific constructs in the Results section.

11. Open Science Considerations. While SPSS is a common statistical tool, open science practices would benefit from clear data and protocol sharing. Explicitly state where readers can access the materials.

=> We fully agree with the importance of open science practices and included a dedicated Data Availability Statement (Page 16, Line 510) with persistent DOIs to ensure the transparency and reproducibility of our findings. We provide a comprehensive suite of materials, including de-identified participant-level data in .sav format, original SPSS output files (.spv), and a step-by-step analysis protocol in PDF format. As noted in the revised Methods section (Page 9, Line 269), all statistical procedures were conducted via the graphical interface of IBM SPSS Statistics v27. By sharing the original output files along with a detailed "analysis recipe," we enable other researchers to fully verify and replicate our analytical steps.

12. Statistical Rationale. Explain why Spearman’s correlation was used (e.g., non-normal distribution), and clarify whether regression models were corrected for multiple comparisons. Justify the model selection process more clearly.

=> Response: We updated the Methods and Results sections to provide a more detailed statistical rationale. First, we clarified that Spearman’s rank correlation was employed because several variables exhibited non-normal distributions and outliers, as determined by skewness and kurtosis (Page 8, Lines 256-263). Regarding model selection, we now transparently describe the stepwise regression criteria (entry p < .05, removal p > .10) while adding caution regarding potential overfitting (Page 10, Line 314). To address the issue of multiple comparisons, we framed our correlation analyses as exploratory and descriptive, and prioritized the interpretation of effect sizes and two-tailed 95% confidence intervals. We have added a corresponding note in the Limitations section to ensure transparency regarding the risk of Type I errors (Page 13, Lines 424-428).

13. Sample Size Concerns. The relatively small sample (N=50) may limit the power and generalizability of the regression analysis. Discuss limitations related to sample size and possible risks of overfitting or underpowered findings.

=> We acknowledge the limitations associated with our sample size (N = 50) and added a dedicated subsection to the Limitations (Page 13, Lines 419-423) to address the concerns regarding statistical power and generalizability. In this section, we discuss how a relatively small sample size may limit the precision of our estimates and increase the risk of overfitting, particularly in the context of stepwise regression selection.

14. Streamline Results. Reporting of statistics (e.g., means, SDs, p-values, CI) could be streamlined for clarity. Use visual aids (tables or figures) to consolidate key outcomes for easier interpretation.

=> We streamlined the Results section to enhance clarity and interpretability by reducing the statistical density of the text and consolidating key data into tables. Descriptive statistics for all variables are shown in Table 2 (Page 23), allowing for a more focused narrative flow in the text. Throughout the Results section, we now report a standardized and consistent set of statistics, including means, standard deviations, Spearman correlation coefficients (rs) with 95% confidence intervals and p-values, as well as comprehensive regression parameters (B, SE, β, 95% CI, and VIF). This reorganization ensures that all critical outcomes are presented transparently and are easily accessible to the reader while maintaining a concise and readable manuscript.

15. Limitations Section: Explicitly address the study’s limitations, including sample size, generalizability, reliance on self-report measures, and the exploratory nature of the design. This is currently underdeveloped and needed for transparency.

=> We expanded the Discussion section by adding a dedicated Limitations subsection (Page 13) to ensure full transparency and provide a balanced interpretation of our findings. This section comprehensively addresses the statistical constraints of our sample size (N = 50), including precision limits and potential overfitting risks associated with our model selection. We also discuss the generalizability of our results, noting that the use of a non-clinical convenience sample may not fully represent the experiences of autistic or other neurodivergent groups. Furthermore, we acknowledge the reliance on self-report measures and ecological validity considerations when using RDM stimuli as a proxy for physical bubble tubes. Finally, we clarify the exploratory nature of the design and the absence of registration, highlighting how our commitment to sharing open materials mitigates these concerns and supports future replication efforts.

 10.5256/f1000research.177078.r382825 Reviewer response for version 1 Tavassoli Teresa 1 Referee University of Reading, Reading, England, UK

Competing interests: No competing interests were disclosed.

 13 6 2025 Copyright: © 2025 Tavassoli T 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

This paper is about an interesting topic of the speed of bubbles and how its relates to interoception and visual discomfort. Here are a few comments :

In the introduction also mention the work on sensory hyperreactivity and anxiety such as Green et al., 2010 and 2012, MacLennan at el, 2020, 2021, Verhulst et al., 2022. 

Please comment on why healthy adults were used for this study. The  introduction covers anxiety, autism and mental health conditions and thus it might have been more informative to use one of these groups rather then healthy controls.

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Yes

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

autism, sensory reactivity

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Koyama Shinichi Institute of Art and Design, University of Tsukuba, Tsukuba, Ibaraki Prefecture, Japan

Competing interests: No competing interests were disclosed.

 14 1 2026

Response to reviewer 1

Thank you very much for your comment.

In the introduction also mention the work on sensory hyperreactivity and anxiety such as Green et al., 2012, MacLennan et al, 2020, 2021, Verhulst et al., 2022. Please comment on why healthy adults were used for this study. The introduction covers anxiety, autism and mental health conditions and thus it might have been more informative to use one of these groups rather than healthy controls.

=> We have now incorporated the references you mentioned in the Introduction section to better contextualize our work within the existing literature on sensory hyperreactivity and anxiety: Green et al., 2012 (Page 3, Line 55); MacLennan et al., 2020, 2021 (Page 3, Lines 55-56); Verhulst et al., 2022 (Page 3, Line 55-56).

=>The purpose of the study was to examine the correlation between sensory sensitivity (ranging from very low to very high sensitivity) and preferred dot speed. We therefore recruited participants with a wide variance in sensitivity, from below average to highly sensitive, regardless of the presence or absence of a clinical diagnosis.