Experimental Modelling of Body Schema with the Rubber Hand Illusion and the Hand Laterality Judgment Task: Cognitive and Meta-cognitive Levels

Dobrushina, O.R. Research Center of Neurology, Moscow, Russia

Arina, G.A. Lomonosov Moscow State University, Moscow, Russia

Kazantseva, D.A. Lomonosov Moscow State University, Moscow, Russia

Tsypushtanova, M.M. Research Center of Neurology, Moscow, Russia

Makarova, A.G. Research Center of Neurology, Moscow, Russia

Dobrynina, L.A. Research Center of Neurology, Moscow, Russia

Public significance statements

Body schema—the internal representation of the body’s shape and posture—is disrupted in a number of disorders that impair quality of life, including stroke, eating disorders and phantom pain.  Based on the rubber hand illusion and hand laterality judgment task experiments, this study proposes that body schema representations be separated into cognitive and metacognitive levels. Differentiating the body schema concept may open new possibilities for developing diagnostic and therapeutic methods for disorders affecting body schema.

Introduction

The body schema concept, which was proposed more than a century ago (Head, 1920), is still evolving. Originally, “body schema” referred to a constantly updated representation of the body’s shape and posture through the integration of proprioceptive signals.  Later, it became clear that not only proprioception but also visual, sensorimotor, and other sensory inputs contribute to updating the body schema, and that this multisensory integration leads to a supramodal (amodal) representation of the body’s spatial properties (Graziano & Botvinick, 1999). However, not all authors accept that the body schema is unified (Matsumiya, 2021). There is broad agreement that the main function of the body schema is action guidance, including real and mentally simulated actions (Jeannerod, 2001; de Vignemont, 2009), which distinguishes it from body image. Body image, by contrast, refers to body representations unrelated to movement, including perceptual, conceptual and emotional dimensions (Gallagher, 2005). The embodied cognition framework proposes a broader role: the body schema may shape high-level cognitive processes such as a sense of agency (Gallagher, 2005). A further recent advance is the predictive coding account, in which body representations are continuously updated through the interaction between a priori top-down predictions and the current bottom-up sensory input (Apps & Tsakiris, 2014; Di Lernia et al., 2023; Morasso & Mohan, 2020; Owens et al., 2018). 

The ongoing theoretical development, influenced by experimental studies and clinical observations (Sattin et al., 2023; Schwoebel & Coslett, 2005), aims to differentiate between multiple body representations. While there is growing consensus regarding the distinction between body schema and body image (de Vignemont, 2009; Gallagher, 2005), some authors distinguish between body structural description and body semantics (Schwoebel & Coslett, 2005). In addition, Longo (2016) describes six body representations, which belong either to somatoperception (body image, body schema, superficial schema and body model) or to somatorepresentation (semantic domain and structural description). Within the body schema, de Vignemont distinguishes protective and working forms. The protective body schema provides a sense of ownership and includes the affective dimension, whereas the working body schema supports action in space and provides a sense of agency (de Vignemont, 2017). Thus, the taxonomy of bodily representations remains unsettled.

To support theoretical advancement through data-driven approaches, multiple methods of experimental body schema modelling have been developed. Body schema modelling refers to experimental tasks that engage the body schema. The present study, which evaluates components of the body schema, is based on a cross-analysis of performance in two well-validated body schema experiments—the rubber hand illusion (RHI) (Botvinick & Cohen, 1998) and the hand laterality judgment task (HLJT) (Parsons, 1994). During the RHI, the participant’s real hand is hidden from view while they observe an artificial hand placed congruently with the body. Synchronous strokes to the real and artificial hands elicit the illusion of ownership over the artificial hand. Illusion strength is usually measured by the drift of the perceived hand position towards the artificial hand (proprioceptive drift) and by questionnaires assessing the subjective feeling of ownership. These two variables correlate in some but not all experiments, and they are believed to represent different aspects of the illusion rather than being causally related (Abdulkarim & Ehrsson, 2016; Rohde et al., 2011). The RHI shows perceptual and neuronal stability across different settings and over time, so it is believed to reflect trait-like features of body perception (Bekrater-Bodmann et al., 2012).

The RHI is thought to depend on multisensory integration mechanisms, which recently have been conceptualized within the predictive coding framework: the strength of the illusion is believed to depend on the balance between bottom-up processing (visual and tactile stimulation) and top-down influences (the stability of body schema against distortions) (Dobrushina et al., 2021; Samad et al., 2015). Both bottom-up and top-down influences remain subjects of lively debate (Chancel et al., 2021; Ehrsson et al., 2022; Golaszewski et al., 2021; Limanowski, 2021; Lush et al., 2020; Rossi Sebastiano et al., 2022; Santos et al., 2023; Slater & Ehrsson, 2022; Thériault et al., 2022). Interest in the dynamic characteristics of the RHI and in its fading has grown in recent years (Abdulkarim et al., 2021; Finotti et al., 2023; Perepelkina et al., 2018). The present study used a dynamic version of the rubber hand illusion, as it allows more detailed quantification of the processes related to predictive coding. Over the course of stimulation, illusion strength increased gradually, consistent with progressive resolution of visuo-proprioceptive prediction error as the hypothesis that the artificial hand is one’s own came to provide the best explanation of the incoming visuo-tactile input (Perepelkina et al., 2018).

The HLJT is a widely used body schema task that requires participants to distinguish left from right hands presented in different positions (Kim & Yi, 2021; Parsons, 1987, 1994; Sattin et al., 2023). Participants typically perform the task by mentally rotating their own hand into the presented position, which implies reliance on the body schema. This strategy is prevalent in children as early as 5–8 years of age (Spruijt et al., 2015). Moreover, limited experience of hand use impairs performance (Funk & Brugger, 2007). However, alternative accounts dispute the necessity of mental movement (Vannuscorps et al., 2012; Viswanathan et al., 2012). The hand-binding hypothesis explains the HLJT via a cross-modal analysis of the features of a seen hand, whereas mental movement is considered an aftereffect of this feature-based analysis (Viswanathan et al., 2012). Other authors regard mental movement as only one of several possible strategies for completing the HLJT (Moreno-Verdú & Hardwick, 2022). Performance is measured by the number of correct responses and by response time, which is longer for the images of hands rotated into uncommon, uncomfortable positions (Parsons, 1994).  

To our knowledge, the RHI and HLJT have not been used simultaneously in previous studies. The RHI effect is thought to reflect the incorporation of an object into the protective body schema (de Vignemont, 2017), whereas the mental movement involved in HLJT refers to the working body schema (Haslam et al., 2022). Thus, we aimed to map the links between the different components of the body schema using the RHI (proprioceptive drift and ownership) and the HLJT (answer correctness and response time). We considered a stronger RHI to be evidence of a less stable body schema, and successful HLJT completion to be evidence of stronger body representations. We hypothesized that higher scores in the RHI would be associated with lower scores in the HLJT.

Methods

Participants

The study sample included 31 (17 male) right-handed healthy volunteers aged 18–35 years (mean 20.5 ± 3.0). None of the participants had previously taken part in experiments with bodily illusions. Each participant completed two successive experiments—the HLJT followed by the RHI. The order of tasks was not balanced. The HLJT data were collected with the participation of N.V. Engovatova and were partially used in her master’s thesis (Engovatova, 2022). 

Procedure 

Rubber Hand Illusion

We used a previously validated dynamic modification of the RHI (Perepelkina et al., 2018). The participants sat still at a table with their right hand hidden inside a wooden box. The middle finger of the right hand was positioned 20 cm from the edge of the table. A life-sized and gender-matched right-hand prosthesis (“rubber hand”) was placed medial to the real right hand at a distance of 15 cm so that the participant could see it. The right forearm and the forearm of the rubber hand were covered with black fabric (Figure 1).

Figure 1. Experimental setting in the rubber hand illusion

The experiment consisted of two conditions: the emergence of the illusion and the fading of the illusion. During the emergence condition, the experimenter stroked the real and artificial hands slightly irregularly at a mean frequency of 0.5 Hz. During the fading condition, stimulation was absent: the participant could no longer see the rubber hand (the experimenter had removed it), and there was no tactile stimulation. We used a behavioral measure (proprioceptive drift) and a subjective measure (ownership questionnaire) of illusion strength. Proprioceptive drift was measured as the difference between the initial error in right-hand localization (measured before the start of the stimulation) and the perceived position of the hand after 15, 30, 60, 120 and 240 seconds from the onset of each condition. With their eyes closed, participants made a pointing movement to the middle finger of the right hand from each of three starting points, presented in randomized order. The results of these three measurements were averaged.  The ownership questionnaire was completed at the same time points only during the emergence condition. It consisted of three test questions (statements related to the real effects of RHI) and three control questions rated on a Likert scale from -3 to 3; the questionnaire was published previously (Perepelkina et al., 2019).

All participants completed the RHI after the computerized HLJT.

Hand Laterality Judgment Task

We used a computerized HLJT based on Parsons (1994). The stimulus material comprised black-and-white photographs of both hands (left and right) in palm-up and palm-down positions.  These four photographs were digitally manipulated to create images of hands rotated to 0°, 90° medial, 180° and 90° lateral orientations, giving a total of 16 different images (Figure 2). Each image was presented four times in a sequence with a randomized stimulus order (64 trials in total). The presentation of the task and the collection of responses were performed using the Cogent Matlab Toolbox (http://www.vislab.ucl.ac.uk/cogent_2000.php, RRID:SCR_015672).

Figure 2. Examples of Stimuli in the Hand Laterality Judgment Task

Note. Hands differed in handedness (right or left), position (palm up or down) and orientation (0°, 90° medial, 180° and 90° lateral), resulting in 2 × 2 × 4 = 16 different stimuli.

Participants sat at a computer with their left hand resting on the table and two fingers of their right hand on the response keys (“<”—left, “>”—right). Each stimulus appeared on the screen and remained there until the participant responded. Participants were asked to identify the laterality of the hand as quickly and accurately as possible. Before the test, instructions were displayed on the screen, and a practice stimulus was then presented.

Statistical Analysis

Data analysis was performed using mixed linear modelling in R Project 3.6.0 (https://www.R-project.org/, RRID:SCR_001905), packages “lme4”, “lmerTest”, “dplyr”, “ggplot2”, “simr”.  Continuous variables were rescaled.

First, we modelled the data from the two experiments separately in order to verify their validity.  For the RHI, the following model was used for the proprioceptive drift dynamics: 

Proprioceptive drift ~ Time * Condition + (1 | Participant) (1),

where “Time” represented the time of stimulation, “Condition” indicated the emergence or fading phase of the illusion and (1 | Participant) represented random participant-specific effects. For the ownership dynamics, the following model was used:

Ownership questionnaire ~ Time * Subscale + (1 | Participant) (2),

where “Time” represented the time of stimulation and “Subscale” indicated test or control questionnaire statements.

For the HLJT, the following models were used for the answer correctness and response times for each stimulus:

Answer correctness ~ Hand * Palm * Orientation + (1 | Participant) (3),

Log (Response time) ~ Hand * Palm * Orientation + (1 | Participant) (4),

where the stimulus characteristics “Hand” (right or left), “Palm” (up or down) and “Orientation” (0°, 90° medial, 180° or 90° lateral) served as independent variables.  A binomial distribution was used to model answer correctness (0 or 1).

Spearman correlation was used for pairwise evaluation of the dependencies between the total number of correct responses and mean response time, and also between the ownership questionnaire score and proprioceptive drift.

Second, to assess the interaction between the performance in the two experiments, we entered the resulting parameters from the HLJT—the total number of correct responses (Accuracy) and the average correct response time (Response Time)—into the models representing RHI dynamics:

Proprioceptive drift ~ Time * Condition * Accuracy + (1 | Participant) (5)

Proprioceptive drift ~ Time * Condition * Average Response Time + (1 | Participant) (6)

Ownership questionnaire ~ Time * Subscale * Accuracy + (1 | Participant) (7)

Ownership questionnaire ~ Time * Subscale * Average Response Time + (1 | Participant) (8)

The significance threshold for these four confirmatory models was Bonferroni-corrected to α = .05/4 = .0125.

The normality of the residual distributions for all linear mixed models was assessed by skewness and kurtosis in accordance with the reference values proposed by Kim (2013). All models were tested for homoscedasticity of the residuals. Where heteroscedasticity was detected, mixed models with separate residual variances for each of the predictors (fixed and random) were fitted (Nakagawa et al., 2025). Heterogeneity did not affect the results in any case.

Transparency and Openness

We report how we determined our sample size, all data exclusions, all manipulations, and all measures in the study. All data, code and stimuli materials are available in the Mendeley Repository at http://dx.doi.org/10.17632/p7zv55hnd5.1. This study’s design and its analysis were not pre-registered.

The sample size was similar to those of previous studies utilizing the RHI (Perepelkina et al., 2017, 2018) and the HLJT (Parsons, 1987, 1994; Schwoebel, 2001). Power analysis was performed using Monte Carlo simulations and, in the absence of previous studies combining the HLJT and the RHI, was based on the observed effect sizes. For the model describing the dependence of proprioceptive drift on accuracy in the HLJT, the estimated power for the interactions of accuracy with time and with condition was over 99%.  For the model describing the dependence of the ownership questionnaire on average response time in the HLJT, the estimated power for the interaction of average response time with the questionnaire subscale was 91%, and the power for the interaction of average response time, questionnaire subscale, and time was 73%.

Results

Rubber Hand Illusion

As expected, proprioceptive drift depended on time (estimate .08, 95% CI .05 to .11, t = 5.29, p < .001) and condition (estimate .04, 95% CI .02 to .06, higher during fading, t = 3.62, p < .001), with an interaction of time and condition (estimate -.09, 95% CI -.14 to -.05, negative effect of time during fading; t = -4.16, p < .001), along with a dependence of ownership on time (estimate .19, 95% CI .11 to .27, t = 4.80, p < .001) and subscale (estimate .47, 95% CI .41 to .52, higher score for test statements, t = 16.06, p < .001) with an interaction of time and subscale (estimate .13, 95% CI .02 to .24, higher effect of time for test statements, t = 2.27, p = .02). Thus, the dynamics of the emergence and fading phases were captured by both proprioceptive drift and ownership questionnaire responses.

The correlation between proprioceptive drift and the sum of test statements on the ownership questionnaire was not significant at any time point.

Hand Laterality Judgment Task

During the HLJT, errors were observed in 9.9% of trials. The total number of correct responses ranged from 41 to 64 (out of 64) with a median of 59 (25th quartile: 55; 75th quartile: 62). The error rate was higher for the 180° orientation (estimate -1.28, 95% CI -2.25 to -.31, z = -2.60, p =  .01).

The average response time ranged from 1097 to 5965 ms with a median of 2101 ms (25th quartile 1529 ms, 75th quartile 2524 ms).  Response time was longer for the palm-up position (estimate .21, 95% CI .10 to .31, t = 3.82, p = .001), 180° orientation (estimate .57, 95% CI .47 to .68, t = 10.65, p  < .001) and 90° medial orientation (estimate .17, 95% CI .06 to .27, t = 3.10, p = .002), with an interaction of palm-up position and 180° orientation (estimate -.29, 95% CI -.43 to -.14, t = -3.76, p = .01) and also of palm-up position and 90° medial orientation (estimate -.23, 95% CI -.38 to -.08, t = -3.02, p = .03). Post hoc analyses were performed for the Orientation factor and the Orientation × Palm interaction (Tables 1 and 2).

Table 1

Post-hoc Analyses: Pairwise Comparisons by Orientation Factor

Orientation	Mean	SE	df	Lower CL	Upper CL	p-value (0)	p-value (180)	p-value (90lateral)
0	7.34	.06	34.60	7.22	7.47	-	-	-
180	7.78	.06	34.60	7.65	7.90	< .001	-	-
90lateral	7.57	.06	34.60	7.45	7.70	< .001	< .001	-
90medial	7.40	.06	34.60	7.27	7.52	1	< .001	< .001

Note. SE = Standard Error. CL = Confidence Limit. 

Table 2

Post-hoc Analyses: Pairwise Comparisons by Orientation × Palm Interaction

Orientation	Palm	Mean	SE	df	Lower CL	Upper CL	p-value (0)	p-value (180)	p-value (90lateral)
0	Down	7.23	.07	41.30	7.09	7.36	-	-	-
180	Down	7.83	.07	41.30	7.70	7.96	< .001	-	-
90lateral	Down	7.40	.07	41.30	7.27	7.53	< .001	< .001	-
90medial	Down	7.42	.07	41.30	7.29	7.55	< .001	< .001	1
0	Up	7.46	.07	41.30	7.33	7.60	-	-	-
180	Up	7.73	.07	41.30	7.59	7.86	< .001	-	-
90lateral	Up	7.75	.07	41.30	7.61	7.88	< .001	1	-
90medial	Up	7.37	.07	41.30	7.24	7.51	.12	< .001	< .001

Note. SE = Standard Error. CL = Confidence Limit. 

We observed the longest response times in the 180° orientation, intermediate response times in the 90° lateral orientation and the fastest response times in both the 90° medial and 0° orientations (Table 1). In the palm-down position, the longest response time was observed in the 180° orientation, intermediate response times in the 90° medial and 90° lateral orientations, and the fastest response time in the 0° orientation. In the palm-up position, the longest response times were observed in the 180° and 90° lateral orientations, and the fastest in the 0° and 90° medial orientations (Table 2). Thus, more time was required for the mental rotation of the depicted hand into views that correspond to uncommon, uncomfortable positions of a real hand.

No correlation was observed between the total number of correct responses and the average response time (Rho = -.05, p = .75).

Accuracy in the Hand Laterality Judgment Task and Rubber Hand Illusion Dynamics

Accuracy in the mental hand rotation task was linked to the dynamics of proprioceptive drift, but not to the sense of ownership in the RHI. The influence of accuracy in the mental hand rotation task on proprioceptive drift dynamics is shown in the Figure 3 (a median split is used for illustrative purposes). We observed an interaction of accuracy with time (estimate -.25, 95% CI -.37 to -.13, t = -3.96, p  < .001), with condition (estimate -.24, 95% CI -.33 to -.15, t = -5.38, p  < .001) and an interaction of the number of correct responses, condition and time (estimate .38, 95% CI .20 to .56, t = 4.30, p < .001). To distinguish the effects related to the emergence and fading phases, we performed a post hoc mixed-models analysis for these two phases separately.  During the emergence phase, an interaction of accuracy with time was observed (high accuracy related to slower dynamics, estimate -.25, 95% CI -.37 to -.13, t = -4.05, p < .001), whereas during the fading phase higher accuracy was associated with a lower absolute value of drift (estimate -.18, 95% CI -.32 to -.04, t = -2.75, p = .02) and with less prominent dynamics (estimate .13, 95% CI .02 to .25, t = 2.60, p =  .02).  Taken together, higher accuracy in the mental hand rotation task was linked to less pronounced RHI dynamics.

Figure 3. Interaction of the proprioceptive drift dynamics with the number of correct responses in the mental hand rotation task

Note. In the “spaghetti” plot, each line represents the dynamics of proprioceptive drift during the RHI in an individual participant. The time point 240 s corresponds to the end of the emergence phase (discontinuation of stimulation). For illustrative purposes, a median split of the sample was performed based on accuracy in the mental hand rotation task.  Bold lines represent means of the two subsamples.

Average Response Time in the Hand Laterality Judgment Task and Rubber Hand Illusion Dynamics

Response time in the mental hand rotation task was significantly associated with ownership dynamics, but not with proprioceptive drift in the RHI. For ownership, we observed an interaction between response time and question type (estimate .50, 95% CI .22 to .78, t = 3.50, p < .001) and between response time, question type, and time (estimate -.81, 95% CI -1.36 to -.25, t = -2.81, p = .005). To distinguish the effects related to the test and control questionnaire statements, we performed a separate post hoc mixed-models analysis. No effects were found for the control statements, whereas for the test statements a longer response time was associated with a higher absolute value (estimate .56, 95% CI .14 to .98, t = 2.54, p = .01) but slower dynamics (interaction with time: estimate -.57, 95% CI -.96 to -.17, t = -2.82, p = .006) of ownership.  The dynamics of the ownership questionnaire (sum of test statements) as a function of response time are shown in the Figure 4 (median split used for illustrative purposes). Both the modelling and visual inspection indicate that slow performance in the mental hand rotation task was associated with early emergence of the subjective feeling of ownership over the rubber hand.

Figure 4. Interaction of the ownership questionnaire dynamics with response time in the mental hand rotation task

Note. In the “spaghetti” plot, each line represents the ownership questionnaire (sum of test statements) dynamics during the emergence phase of the RHI in an individual participant.  For illustrative purposes, a median split of the sample was performed based on the response time in the mental hand rotation task.  Bold lines represent means of the two subsamples.

Discussion

Our results are consistent with predictive coding accounts of body perception (Apps & Tsakiris, 2014; Morasso et al., 2015), of body ownership (Bertoni et al., 2023) and, in particular, of the RHI (Chancel et al., 2021; Dobrushina et al., 2021; Golaszewski et al., 2021; Samad et al., 2015; Santos et al., 2023). In the predictive coding framework, a stronger RHI indicates the predominance of bottom-up processes. The HLJT can also be conceptualized in terms of predictive coding, since the motor image serves as a prediction that is compared with the depicted hand before a decision is made. Participants with poor performance on the HLJT showed faster dynamics in the RHI experiment, which can be explained by less-developed body schema representations and, thus, weaker top-down influences that protect the body schema from artificial distortions. Results in the HLJT improve with motor practice (Habacha et al., 2014) and are negatively affected by hand immobilization (Debarnot et al., 2018), which suggests that the task reflects the strengths of hand representations. Thus, we propose that performance in the HLJT relies on the availability of bodily representations, i.e., top-down processing. Both proprioceptive drift and the ownership questionnaire in the RHI were negatively associated with performance on the HLJT. These results parallel the absence of RHI effects in expert pianists (Pyasik et al., 2018), who, due to years of practice, may have highly developed internal representations of the hand.

A cross-analysis of the two experiments revealed distinct associations: first, between proprioceptive drift and accuracy in the HLJT, and second, between the sense of ownership over an artificial hand and response time in the HLJT.  At the same time, we did not observe pairwise correlations between proprioceptive drift and ownership scores in the RHI or between accuracy and response time in the HLJT.  Some RHI studies show an association between proprioceptive drift and ownership scores (Botvinick & Cohen, 1998; Longo et al., 2008; Samad et al., 2015; Tosi et al., 2023), whereas others fail to detect it (Holle et al., 2011; Rohde et al., 2011) and show that hand proprioception may change without changes in ownership (Holmes et al., 2006; Makin et al., 2008). The same dissociation between ownership and proprioceptive drift appears in somatic RHI (Gallagher et al., 2020). A study using subtle mechanical manipulation of participants’ hand position during the RHI experiment found no causal link between proprioceptive drift and a sense of ownership (Abdulkarim & Ehrsson, 2016). According to a recent meta-analysis, these measures are related, but the correlation is rather weak (Tosi et al., 2023). The association between accuracy and response time in the HLJT remains largely underexplored. The limited available evidence comes from a few studies with differing designs, which reported a correlation between error rate and response time (Conson et al., 2012; Parsons, 1994).

We suggest that the two pairs of variables (proprioceptive drift and accuracy in the HLJT; ownership questionnaire and response time in the HLJT) characterize distinct domains of body schema organization. By analogy with other cognitive tasks, the two domains identified here may be defined as cognitive and metacognitive.  In our interpretation, proprioceptive drift and accuracy in the HLJT reflect the spatial, cognitive level of the body schema, i.e., the strength of the structural representations. Participants with stronger body representations may be less susceptible to illusion-related errors in localizing the hidden hand and therefore show lower proprioceptive drift.  The sense of ownership and response time in the HLJT may characterize the subjective, metacognitive level of body schema-related processing. A faster response may indicate higher confidence in the answer (the average response time was calculated using only correct responses to minimize the possible influence of misclicks and guesses). Participants with high confidence in their body schema may be more inclined to rely on internal representations rather than on actual sensory input when they develop a sense of body ownership (dominance of top-down processing). Accordingly, they show lower scores on the ownership questionnaire.  Nevertheless, the distinction between the cognitive and metacognitive components of the body schema requires careful verification in further research that controls for metacognitive variables.

Our findings may contribute to the theoretical development of the body schema concept, and they may also have clinical significance. The body schema is affected in a number of disorders that impair quality of life (de Vignemont, 2009). The rubber hand illusion and HLJT are used to characterize disease-related alterations of the body schema in patients with stroke (de Vries et al., 2013; Feenstra et al., 2016; Llorens et al., 2017; Haslam et al., 2022), eating disorders (Campione et al., 2017; Eshkevari et al., 2011), somatoform disorders (Perepelkina et al., 2019) and phantom pain (Giummarra et al., 2010). Theoretical models linking RHI and HLJT parameters to body schema dimensions can be used to interpret diagnostic data and develop targeted therapeutic interventions for patients with body schema distortions.

Conclusion

The results of our study support the predictive coding conceptualization of body schema, which is believed to be continuously updated through an interaction between prior inferences and current sensory input. We observed the expected negative association between RHI strength and HLJT performance. Beyond this, our results can be interpreted as a distinction between the cognitive and metacognitive levels of body schema organization. The cognitive level may correspond to structural representations, whereas the metacognitive level may comprise confidence in those representations. However, this hypothesis requires further confirmation.  The proposed model now needs to be tested in further research using different experimental designs.

Limitations

The most substantial limitation of the study is the order of task presentation, which was not counterbalanced across participants. All participants first completed the HLJT and then took part in the RHI. Thus, the effect of presentation order was not controlled, which means that order, learning or carry-over effects cannot be excluded. For example, according to our hypotheses, we would expect reduced RHI scores after participation in the HLJT. Although the data show that, on average, the RHI was formed successfully, these measures may be underestimated because the preceding task also engaged the body schema. In addition, the small sample size limits the generalizability of the results.

Ethics Statement

The study protocol was approved by the Ethics Committee and the Institutional Review Board of the Research Center of Neurology (protocol number 4-3/21), and all participants gave informed consent.

Conflict of Interest

The authors declare no conflict of interest.

References

Abdulkarim, Z., & Ehrsson, H.H. (2015). No causal link between changes in hand position sense and feeling of limb ownership in the rubber hand illusion. Attention Perception & Psychophysics, 78(2), 707–720. https://doi.org/10.3758/s13414-015-1016-0

Abdulkarim, Z., Hayatou, Z., & Ehrsson, H.H. (2021). Sustained rubber hand illusion after the end of visuotactile stimulation with a similar time course for the reduction of subjective ownership and proprioceptive drift. Experimental Brain Research, 239(12), 3471–3486. https://doi.org/10.1007/s00221-021-06211-8

Apps, M.A., & Tsakiris, M. (2013). The free-energy self: A predictive coding account of self-recognition. Neuroscience & Biobehavioral Reviews, 41, 85–97. https://doi.org/10.1016/j.neubiorev.2013.01.029

Bekrater-Bodmann, R., Foell, J., Diers, M., & Flor, H. (2012). The perceptual and neuronal stability of the rubber hand illusion across contexts and over time. Brain Research, 1452, 130–139. https://doi.org/10.1016/j.brainres.2012.03.001

Bertoni, T., Mastria, G., Akulenko, N., Perrin, H., Zbinden, B., Bassolino, M., & Serino, A. (2023). The self and the Bayesian brain: Testing probabilistic models of body ownership through a self-localisation task. Cortex, 167, 247–272. https://doi.org/10.1016/j.cortex.2023.06.019

Botvinick, M., & Cohen, J. (1998). Rubber hands ‘feel’ touch that eyes see. Nature, 391(6669), 756. https://doi.org/10.1038/35784

Campione, G.C., Mansi, G., Fumagalli, A., Fumagalli, B., Sottocornola, S., Molteni, M., & Micali, N. (2017). Motor-based bodily self is selectively impaired in eating disorders. PLoS ONE, 12(11), e0187342. https://doi.org/10.1371/journal.pone.0187342

Chancel, M., Hasenack, B., & Ehrsson, H.H. (2021). Integration of predictions and afferent signals in body ownership. Cognition, 212, 104722. https://doi.org/10.1016/j.cognition.2021.104722

Conson, M., Mazzarella, E., Donnarumma, C., & Trojano, L. (2012). Judging hand laterality from my or your point of view: Interactions between motor imagery and visual perspective. Neuroscience Letters, 530(1), 35–40. https://doi.org/10.1016/j.neulet.2012.09.051

De Vignemont, F. (2009). Body schema and body image—Pros and cons. Neuropsychologia, 48(3), 669–680. https://doi.org/10.1016/j.neuropsychologia.2009.09.022

De Vignemont, F. (2017). Agency and bodily ownership: the bodyguard hypothesis. The Subject's Matter: Self-Consciousness and the Body, 217. https://doi.org/10.7551/mitpress/10462.003.0013

De Vries, S., Tepper, M., Feenstra, W., Oosterveld, H., Boonstra, A.M., & Otten, B. (2013). Motor imagery ability in stroke patients: The relationship between implicit and explicit motor imagery measures. Frontiers in Human Neuroscience, 7, 790. https://doi.org/10.3389/fnhum.2013.00790

Debarnot, U., Huber, C., Guillot, A., & Schwartz, S. (2018). Sensorimotor representation and functional motor changes following short-term arm immobilisation. Behavioral Neuroscience, 132(6), 595–603. https://doi.org/10.1037/bne0000274

Di Lernia, D., Serino, S., Tuena, C., Cacciatore, C., Polli, N., & Riva, G. (2023). Mental health meets computational neuroscience: A predictive Bayesian account of the relationship between interoception and multisensory bodily illusions in anorexia nervosa. International Journal of Clinical and Health Psychology, 23(4), 100383. https://doi.org/10.1016/j.ijchp.2023.100383

Dobrushina, O.R., Arina, G.A., Dobrynina, L.A., Novikova, E.S., Gubanova, M.V., Belopasova, A.V., Vorobeva, V.P., Suslina, A.D., Pechenkova, E.V., Perepelkina, O.S., Kremneva, E.I., & Krotenkova, M.V. (2021). Sensory integration in interoception: Interplay between top-down and bottom-up processing. Cortex, 144, 185–197. https://doi.org/10.1016/j.cortex.2021.08.009

Ehrsson, H.H., Fotopoulou, A., Radziun, D., Longo, M.R., & Tsakiris, M. (2022). No specific relationship between hypnotic suggestibility and the rubber hand illusion. Nature Communications, 13(1), 564.  https://doi.org/10.1038/s41467-020-18591-6

Ehrsson, H.H., Fotopoulou, A., Radziun, D., Longo, M.R., & Tsakiris, M. (2022). No specific relationship between hypnotic suggestibility and the rubber hand illusion. Nature Communications, 13(1), 564. https://doi.org/10.1038/s41467-022-28177-z

Engovatova, N.V. (2022). Individual'no-psikhologicheskie faktory resheniia zadachi po opredeleniiu lateralizatsii ruki. [Master's thesis, Lomonosov Moscow State University].

Eshkevari, E., Rieger, E., Longo, M.R., Haggard, P., & Treasure, J. (2011). Increased plasticity of the bodily self in eating disorders. Psychological Medicine, 42(4), 819–828. https://doi.org/10.1017/s0033291711002091

Feenstra, W., Tepper, M., Boonstra, A.M., Otten, B., & De Vries, S. (2016). Recovery of motor imagery ability in the first year after stroke. International Journal of Rehabilitation Research, 39(2), 171–175. https://doi.org/10.1097/MRR.0000000000000162

Finotti, G., Garofalo, S., Costantini, M., & Proffitt, D.R. (2023). Temporal dynamics of the Rubber Hand Illusion. Scientific Reports, 13(1), 7526. https://doi.org/10.1038/s41598-023-33747-2

Funk, M., & Brugger, P. (2007). Mental rotation of congenitally absent hands. Journal of the International Neuropsychological Society, 14(1), 81–89. https://doi.org/10.1017/S1355617708080041

Gallagher, M., Colzi, C., & Sedda, A. (2020). Dissociation of proprioceptive drift and feelings of ownership in the somatic rubber hand illusion. Acta Psychologica, 212, 103192. https://doi.org/10.1016/j.actpsy.2020.103192

Gallagher, S. (2005). How the body shapes the mind. Clarendon press.

Giummarra, M.J., Georgiou-Karistianis, N., Nicholls, M.E.R., Gibson, S.J., & Bradshaw, J.L. (2010). The phantom in the mirror: a modified rubber-hand illusion in amputees and normals. Perception, 39(1), 103–118. https://doi.org/10.1068/p6519

Golaszewski, S., Frey, V., Thomschewski, A., Sebastianelli, L., Versace, V., Saltuari, L., Trinka, E., & Nardone, R. (2021). Neural mechanisms underlying the Rubber Hand Illusion: A systematic review of related neurophysiological studies. Brain and Behavior, 11(8), e02124. https://doi.org/10.1002/brb3.2124

Graziano, M.S.A., & Botvinick, M.M. (2002). How the brain represents the body: Insights from neurophysiology and psychology. In Common mechanisms in perception and action: Attention and performance XIX (pp. 136–157). https://doi.org/10.1093/oso/9780198510697.003.0006

Habacha, H., Molinaro, C., Tabben, M., & Lejeune-Poutrain, L. (2014). Implementation of specific motor expertise during a mental rotation task of hands. Experimental Brain Research, 232(11), 3465–3473. https://doi.org/10.1007/s00221-014-4029-3

Haslam, B.S., Butler, D.S., Cocks, T.S., Kim, A.S., & Carey, L.M. (2022). Body schema as assessed by upper limb Left/Right judgment tasks is altered in stroke: Implications for motor imagery training. Journal of Neurologic Physical Therapy, 47(1), 26–34. https://doi.org/10.1097/NPT.0000000000000412

Head, H.  (1920).  Studies in neurology, Vol. 2.  Oxford University Press.

Holle, H., McLatchie, N., Maurer, S., & Ward, J. (2011). Proprioceptive drift without illusions of ownership for rotated hands in the “rubber hand illusion” paradigm. Cognitive Neuroscience, 2(3–4), 171–178. https://doi.org/10.1080/17588928.2011.603828

Holmes, N.P., Snijders, H.J., & Spence, C. (2006). Reaching with alien limbs: Visual exposure to prosthetic hands in a mirror biases proprioception without accompanying illusions of ownership. Perception & Psychophysics, 68(4), 685–701. https://doi.org/10.3758/BF03208768

Jeannerod, M. (2001). Neural simulation of action: A unifying mechanism for motor cognition. NeuroImage, 14(1), S103–S109. https://doi.org/10.1006/nimg.2001.0832

Kim, A., & Yi, C. (2021). Understanding the Left Right Judgement Test: A literature review. Physical Therapy Korea, 28(4), 235–244. https://doi.org/10.12674/ptk.2021.28.4.235

Kim, H. (2013). Statistical notes for clinical researchers: assessing normal distribution (2) using skewness and kurtosis. Restorative Dentistry & Endodontics, 38(1), 52. https://doi.org/10.5395/rde.2013.38.1.52

Limanowski, J. (2021). Precision control for a flexible body representation. Neuroscience & Biobehavioral Reviews, 134, 104401. https://doi.org/10.1016/j.neubiorev.2021.10.023

Llorens, R., Borrego, A., Palomo, P., Cebolla, A., Noé, E., Badia, S.B.I., & Baños, R. (2017). Body schema plasticity after stroke: Subjective and neurophysiological correlates of the rubber hand illusion. Neuropsychologia, 96, 61–69. https://doi.org/10.1016/j.neuropsychologia.2017.01.007

Longo, M.R., Schüür, F., Kammers, M.P., Tsakiris, M., & Haggard, P. (2008). What is embodiment? A psychometric approach. Cognition, 107(3), 978–998. https://doi.org/10.1016/j.cognition.2007.12.004

Longo, M. (2016). Types of body representation. In Y. Coello & M.H. Fischer (Eds.), Perceptual and emotional embodiment, 1 (pp. 117–134). Routledge.

Lush, P., Botan, V., Scott, R.B., Seth, A.K., Ward, J., & Dienes, Z. (2020). Trait phenomenological control predicts experience of mirror synaesthesia and the rubber hand illusion. Nature Communications, 11(1), 4853. https://doi.org/10.1038/s41467-020-18591-6

Makin, T.R., Holmes, N.P., & Ehrsson, H.H. (2008). On the other hand: Dummy hands and peripersonal space. Behavioural Brain Research, 191(1), 1–10. https://doi.org/10.1016/j.bbr.2008.02.041

Matsumiya, K. (2022). Multiple representations of the body schema for the same body part. Proceedings of the National Academy of Sciences, 119(4). https://doi.org/10.1073/pnas.2112318119

Morasso, P., & Mohan, V. (2020). The body schema: Neural simulation for covert and overt actions of embodied cognitive agents. Current Opinion in Physiology, 19, 219–225. https://doi.org/10.1016/j.cophys.2020.11.009

Morasso, P., Casadio, M., Mohan, V., Rea, F., & Zenzeri, J. (2015). Revisiting the Body-Schema concept in the context of Whole-Body Postural-Focal dynamics. Frontiers in Human Neuroscience, 9, 83. https://doi.org/10.3389/fnhum.2015.00083

Moreno-Verdú, M., & Hardwick, R. (2022). Measuring the mind’s eye: Does the Hand Laterality Judgement Task really assess motor imagery ability? PsyArXiv. https://doi.org/10.31234/osf.io/g2cj3

Nakagawa, S., Ortega, S., Gazzea, E., Lagisz, M., Lenz, A., Lundgren, E., & Mizuno, A. (2025). Location–scale models in ecology and evolution: Heteroscedasticity in continuous, count and proportion data. Methods in Ecology and Evolution, 17(2), 554–566. https://doi.org/10.1111/2041-210x.70203

Owens, A.P., Friston, K.J., Low, D.A., Mathias, C.J., & Critchley, H.D. (2018). Investigating the relationship between cardiac interoception and autonomic cardiac control using a predictive coding framework. Autonomic Neuroscience, 210, 65–71. https://doi.org/10.1016/j.autneu.2018.01.001

Parsons, L.M. (1987). Imagined spatial transformations of one’s hands and feet. Cognitive Psychology, 19(2), 178–241. https://doi.org/10.1016/0010-0285(87)90011-9

Parsons, L.M. (1994). Temporal and kinematic properties of motor behavior reflected in mentally simulated action. Journal of Experimental Psychology Human Perception & Performance, 20(4), 709–730. https://doi.org/10.1037/0096-1523.20.4.709

Perepelkina, O., Boboleva, M., Arina, G., & Nikolaeva, V. (2017). Higher emotional intelligence is associated with a stronger rubber hand illusion. Multisensory Research, 30(7–8), 615–637. https://doi.org/10.1163/22134808-00002577

Perepelkina, O., Romanov, D., Arina, G., Volel, B., & Nikolaeva, V. (2019). Multisensory mechanisms of body perception in somatoform disorders. Journal of Psychosomatic Research, 127, 109837. https://doi.org/10.1016/j.jpsychores.2019.109837

Perepelkina, O., Vorobeva, V., Melnikova, O., Arina, G., & Nikolaeva, V. (2018). Artificial hand illusions dynamics: Onset and fading of static rubber and virtual moving hand illusions. Consciousness and Cognition, 65, 216–227. https://doi.org/10.1016/j.concog.2018.09.005

Pyasik, M., Salatino, A., & Pia, L. (2018). Do movements contribute to sense of body ownership? Rubber hand illusion in expert pianists. Psychological Research, 83(1), 185–195. https://doi.org/10.1007/s00426-018-1137-x

Rohde, M., Di Luca, M., & Ernst, M.O. (2011). The rubber hand illusion: Feeling of ownership and proprioceptive drift do not go hand in hand. PLoS ONE, 6(6), e21659. https://doi.org/10.1371/journal.pone.0021659

Rossi Sebastiano, A., Bruno, V., Ronga, I., Fossataro, C., Galigani, M., Neppi-Modona, M., & Garbarini, F. (2022). Diametrical modulation of tactile and visual perceptual thresholds during the rubber hand illusion: A predictive coding account. Psychological Research, 86(6), 1830–1846. https://doi.org/10.1007/s00426-021-01608-0

Samad, M., Chung, A.J., & Shams, L. (2015). Perception of body ownership is driven by Bayesian sensory inference. PLoS ONE, 10(2), e0117178. https://doi.org/10.1371/journal.pone.0117178

Santos, D., Agostinho, M., Treister, R., & Canaipa, R. (2023). Correlations between within-subject variability of pain intensity reports and rubber hand illusion proprioceptive drift. Neuroscience Letters, 810, 137319. https://doi.org/10.1016/j.neulet.2023.137319

Sattin, D., Parma, C., Lunetta, C., Zulueta, A., Lanzone, J., Giani, L., Vassallo, M., Picozzi, M., & Parati, E.A. (2023). An overview of the body schema and body image: Theoretical models, methodological settings and pitfalls for rehabilitation of persons with neurological disorders. Brain Sciences, 13(10), 1410. https://doi.org/10.3390/brainsci13101410

Schwoebel, J. (2001). Pain and the body schema: Evidence for peripheral effects on mental representations of movement. Brain, 124(10), 2098–2104. https://doi.org/10.1093/brain/124.10.2098

Schwoebel, J., & Coslett, H.B. (2005). Evidence for multiple, distinct representations of the human body. Journal of Cognitive Neuroscience, 17(4), 543–553. https://doi.org/10.1162/0898929053467587

Slater, M., & Ehrsson, H.H. (2022). Multisensory integration dominates hypnotisability and expectations in the rubber hand illusion. Frontiers in Human Neuroscience, 16, 834492. https://doi.org/10.3389/fnhum.2022.834492

Spruijt, S., Jongsma, M.L.A., Van Der Kamp, J., & Steenbergen, B. (2015). Predictive models to determine imagery strategies employed by children to judge hand laterality. PLoS ONE, 10(5), e0126568. https://doi.org/10.1371/journal.pone.0126568

Thériault, R., Landry, M., & Raz, A. (2022). The Rubber Hand Illusion: Top-down attention modulates embodiment. Quarterly Journal of Experimental Psychology, 75(11), 2129–2148. https://doi.org/10.1177/17470218221078858

Tosi, G., Mentesana, B., & Romano, D. (2023). The correlation between proprioceptive drift and subjective embodiment during the rubber hand illusion: A meta-analytic approach. Quarterly Journal of Experimental Psychology, 76(10), 2197–2207. https://doi.org/10.1177/17470218231156849

Vannuscorps, G., Pillon, A., & Andres, M. (2012). Effect of biomechanical constraints in the hand laterality judgment task: Where does it come from? Frontiers in Human Neuroscience, 6, 299. https://doi.org/10.3389/fnhum.2012.00299

Viswanathan, S., Fritz, C., & Grafton, S.T. (2012). Telling the right hand from the left hand. Psychological Science, 23(6), 598–607. https://doi.org/10.1177/0956797611429802