Illusions With Hands,but Not With Balloons – Comparing Ownership and Referral of Touch for a Corporal and Noncorporal Object After Visuotactile Stimulation

Abstract

The rubber hand illusion is a perceptual illusion of perceiving an object like a model hand as part of the own body. The question whether the illusion can be induced with noncorporal objects that do not look like a human body part is not perfectly resolved yet. In this study, we directly assessed the subjective experience of two different components within the illusion (i.e., ownership and referral of touch) when a model hand and a balloon are stimulated. We observed significantly stronger illusion ratings for the hand as compared with the balloon, and only the hand ratings showed a clear affirmation of the illusion. We further conclude that (a) a significant difference between synchronous and asynchronous conditions may not be sufficient to argue for the successful induction of the illusion and (b) the subcomponents show a different pattern in the different conditions, which may lead to alternative interpretations. These observations call for a more fine-grained interpretation of questionnaire data in rubber hand illusion studies.

Keywords

body perception multisensory/cross-modal processing ownership rubber hand illusion

Introduction

The rubber hand illusion (RHI) is a well-known perceptual illusion in which participants perceive a fake model hand as part of their own body (Botvinick & Cohen, 1998). This illusion provided important insights into the mechanisms underlying the experience of the body. Since its introduction, a number of studies have not only investigated the basic mechanisms (Kilteni, Maselli, Kording, & Slater, 2015) but also extended this paradigm and created a variety of new illusions (Lenggenhager, Tadi, Metzinger, & Blanke, 2007; Petkova & Ehrsson, 2008; Tsakiris et al., 2008).

Typically, the illusion is measured by questionnaires that reflect the illusory experience, often complimented by behaviourally quantifiable measures like proprioceptive drift and skin-conductance response (SCR) after a potential threat (Armel & Ramachandran, 2003; Tsakiris & Haggard, 2005). Participants typically undergo two different conditions: one which induces the illusion and one which does not induce the illusion. Participants rate their experience on a Likert-type scale from −3 to +3 or on a discrete visual analogue scale from 1 to 7/10. A typical experiment observes higher ratings of the illusion statements in Condition A as compared with Condition B and conclude that participants experience the RHI in Condition A but not B (Botvinick & Cohen, 1998). However, there is surprisingly little agreement on the exact procedures how to draw inferences based on the data. When overviewing the literature, we can identify the following approaches:

Two different kinds of illusion statements: The first category of statements reflects the experience of perceiving the hand as part of the own body (i.e., I felt as if I was looking at my own hand; ownership). The other statements reflect the experience of apparent fusion of both the visual and tactile stimuli (i.e., I felt as if the touch I feel originates from the rubber hand; referral of touch). Some studies use only one category (e.g., referral of touch; Lloyd, 2007), and some use both kinds of statements (Abdulkarim & Ehrsson, 2016; Botvinick & Cohen, 1998).

One of the most common designs is the comparison between synchronous and asynchronous stimulations in which there is a temporal delay between the touches applied to the two hands. It has been repeatedly shown that asynchronous stimulation does not lead to an illusion (Botvinick & Cohen, 1998; Tsakiris & Haggard, 2005; Ehrsson et al., 2004).

Previous studies have investigated whether the illusion can be induced with a noncorporeal object. For example, Armel and Ramachandran (2003) suggested that the illusion can be induced when the table instead of a model hand is touched, advocating a bottom-up approach in the generation of the illusion (Armel & Ramachandran, 2003). However, Tsakiris & Haggard (2005) have shown that the illusion cannot be induced when the hand is replaced by a stick. Later, Tsakiris et al. examined this question in more detail and changed the object gradually from a simple wooden block to more hand-like shape (Tsakiris, Carpenter, James, & Fotopoulou, 2010). No illusion was observed for a wooden block or intermediately shaped wooden block. Likewise, Guterstam, Gentile, and Ehrsson (2013) also concluded that the illusion cannot be induced with a wooden block and used it as a control condition for a variant of the RHI, the invisible hand illusion (see Experiment 4). Recently, Ma and Hommel (2015) suggested that the illusion can be induced with a noncorporeal like a balloon, which moved in accordance with the participant’s movements.

In the following, we present data from a RHI experiment in which we compared synchronous and asynchronous visuotactile stimulations with a model hand and a noncorporal object, a balloon. We highlight some methodological issues when it comes to the interpretation of the data. We show that a statistical significant difference between synchronous and asynchronous may not be sufficient to infer successful induction of the illusion and present the results of commonly used statistical approaches. Furthermore, we show that the different categories of statements are rated differently in the different object conditions.

Methods

Participants

Forty undergraduate students (25 females; age range: 18–27 years) participated in the experiment. This study was approved by the ethics committee of the University of Reading Malaysia. Participants were recruited via in internal online platform and compensated with student credits. All participants gave written informed consent.

Materials

A wooden box (length 34 cm × height 14.5 cm × width 24.5 cm) was placed on a table with either a life-sized model hand or a balloon placed on top of the box. The model hand was covered with a latex glove; participants wore an identical glove. The glove and the balloon were of similar white colour. A dark blue cloth was used to cover the participants’ right arm during the experiment, see Figure 1.

Figure 1.

Illustration of the setup to test a model hand (upper) and balloon (lower). Note: Please refer to the online version of the article to view the figures in colour.

Questionnaire

The statements were adopted from questionnaires used in previous RHI studies (Botvinick & Cohen, 1998; Kalckert & Ehrsson, 2014). The questionnaire consisted of eight statements with three different categories: three ownership statements (Own), two referral of touch statements (RoT), and three control statements (C), see Table 1. The control statements were used to eliminate potential compliance effect. Statements were adjusted to the different object conditions by replacing the word balloon and hand, respectively, see Table 1. Participants rated their experiences on a 7-point Likert-type scale, where −3 referred to strongly disagree and 3 as strongly agree; 0 indicated uncertainty.

Table 1.

Questionnaire Items.

Category	Statements
Own 1	I felt as if I was looking at my own hand
Own 2	I felt as if the rubber hand was part of my body
Own 2	I felt as if the balloon was part of my body
Own 3	I felt as if the rubber hand was my hand
Own 3	I felt as if the balloon was my hand
RoT 1	It seemed as if I were feeling the touch in the location where I saw the rubber hand touched
RoT 1	It seemed as if I were feeling the touch in the location where I saw the balloon touched
RoT 2	It seemed as though the touch I felt was caused by the paintbrush touching the rubber hand
RoT 2	It seemed as though the touch I felt was caused by the paintbrush touching the balloon
C1	It seems as if I had more than one hand
C2	It felt as if I had no longer a right hand, as if my right hand had disappeared
C3	I felt as if my real hand was turning rubbery

Note. RoT = referral of touch statement.

Skin-Conductance Response

Two electrodes were connected to a Brainvision actiCHamp amplifier (Brain Products, Gilching, Germany), which recorded the participants’ SCR. Data were recorded with the BrainVision Recorder software (1.21.0303 version). Due to technical failure, it was not possible to analyse these data and is not reported.

Procedure

Participants sat comfortably at a table with their arms resting on the table. The box with the model hand was placed in front of the participant, see Figure 1. Prior to each condition, participants were asked to relax and refrain from making any movement during the stimulation. The stimulation was applied at a frequency of 1 Hz. The experimenter touched the proximal phalanx of the index finger of the model and the real hand with identical soft brushes. The order of conditions was randomised. We tested four different conditions comprising two different objects (model hand vs. balloon) and two different stimulations (synchronous vs. asynchronous stimulation). In the synchronous conditions, the experimenter stroked both the participant’s and the model hand’s right index finger or to a spatially corresponding location of the balloon. In the asynchronous condition, the participant’s index finger was stroked approximately 500 milliseconds prior to the touch of the object. Each condition lasted for 90 seconds, after which the hand/balloon was threatened by a sudden approach with a syringe without actually touching the model hand or balloon. Thereafter, participants were asked to withdraw their hand from the box and answer a computerised version of the questionnaire. During this break, participants moved their hand and relaxed to prevent potential carryover effects.

Analysis

In our analysis, we calculated illusion scores by averaging the individual questionnaire items for their respective category and compared them across conditions. Wilcoxon signed-rank tests were used for individual pairwise comparisons, while Friedman tests were used for comparisons involving three or more conditions. Furthermore, we conducted one-sample tests against 0. Thus, we tested whether scores differ significantly from a level in which participants rated their experience of the illusion as unclear and do not clearly affirm the illusion experience. Next, we classified the proportion of illusion responders with a cutoff score of 1 and 0 for illusion, ownership, and RoT scores (Kalckert & Ehrsson, 2014). Effect sizes were computed with the following formula: r = Z/(√N). All results are two-tailed, and all tests are based on a priori hypotheses unless otherwise stated. Bonferroni corrections were applied in cases of multiple tests. Analysis was conducted using SPSS Version 22 (IBM).

Results

In the following, we present the median ratings for all individual scores (illusion scores/ownership score/RoT score). Scores were high for synchronous RHI condition (1.58/1.5/2) but not for its asynchronous condition (−0.54/−0.5/0). The scores of the synchronous balloon condition were either negative or slightly positive (−0.25/−1/0.5), and all negative for its asynchronous condition (−1.25/−1.67/−1), see Figure 2. All control scores were negatively rated for the four conditions (−0.67/−1/−1.17/−1.5).

Figure 2.

Box plots showing median scores in the four conditions: left: illusion score (average of statements Own 1–3, RoT 1–2), middle: ownership score (average of Own 1–3), and right: referral of touch score (average of RoT 1, 2). RHI = rubber hand illusion; RBI = rubber balloon illusion.

We observed significant differences across the four conditions for all three scores (illusion, ownership, and RoT): illusion, χ²(3, N = 40) = 79.821, p < .001; ownership, χ²(3, N = 40) = 75.641, p < .001; and RoT, χ³(2, N = 40) = 68.523, p < .001.

Rubber Hand Illusion Conditions

We found higher illusion scores in the synchronous condition as compared with the asynchronous condition (Z = −5.374, p < .001, r = .60) and a significantly higher illusion score of synchronous condition as compared with the control score (Z = −5.360, p < .001, r = .59). We performed the same comparison between synchronous and asynchronous conditions with the ownership score (Z = −5.346, p < .001, r = .60) and RoT score (Z = −5.099, p < .001, r = .57).

Rubber Balloon Illusion Conditions

Likewise, we found a significantly higher illusion score in the synchronous condition as compared with asynchronous condition (Z = −3.654, p < .001, r = .40), as well as for the ownership score (Z = −2.787, p = .005, r = .31), and RoT score (Z = −3.383, p = .001, r = .38). The synchronous illusion score was significantly higher to the synchronous control score (Z = −5.337, p < .001, r = .51)

Further Analysis

We ran one-sample Wilcoxon signed-rank tests with a test median of 0 for both the hand and balloon and for the three scores of the synchronous conditions. For the hand, we found that all three scores were significantly different to 0 (illusion: p < .001; ownership: p < .001; and RoT: p < .001). It should be noted that only the RoT score is positive for the balloon condition, ownership and illusion scores are negative. In the balloon condition, the illusion scores were not significant to 0 (illusion: p = .357; ownership: p = .009; and RoT: p = .264), with a value of p adjusted for multiple comparisons of p = .0083.

Comparison Hand Versus Balloon

To examine a difference between the objects, we compared the synchronous RHI condition with the synchronous balloon condition. These were significant for all the three scores: illusion (Z = −5.316, p < .001, r = .59), ownership (Z = −5.186, p < .001, r = .58), and RoT (Z = −4.627, p < .001, r = .52) were higher for the hand as compared with the balloon.

Some studies use an approach in which they subtract scores from specific conditions or categories (Abdulkarim & Ehrsson, 2016). In the first approach, the control scores are subtracted from the illusion scores within the same condition. We did so for both the synchronous hand (M = 2.33) and synchronous balloon conditions (M = 0.83) and found a significant difference between the two (Z = −5.032, p < .001, r = .56). In an alternative procedure, the synchronous and asynchronous illusion scores are subtracted and we found a significant difference between hand (M = 1.38) and balloon (M = .075) conditions (Z = −3.106, p = .002, r = .35).

Illusion Responders

See Table 2.

Table 2.

Percentage of Responders for the Three Scores.

	Illusion score (%)	Ownership score (%)	RoT score (%)
Median > 1
Hand synchronous	70	70	87.5
Balloon synchronous	22.5	20	45
Median > 0
Hand synchronous	92.5	85	97.5
Balloon synchronous	45	32.5	57.5

Note. RoT = referral of touch statement.

Discussion

This study investigated how participants experience synchronous and asynchronous visuotactile stimulations to a model hand and a noncorporeal object, a balloon. In line with previous studies, our results show that participants perceive the RHI after synchronous but not after asynchronous stimulation. When using different ways to statistically compare the illusion for the two objects, we see that the illusion is always stronger for the hand as compared with the balloon (for illusion, ownership, and RoT scores). Albeit significant differences between synchronous and asynchronous conditions have been found for the balloon conditions, we like to propose that it cannot be concluded that the illusion can be experienced with a noncorporeal object. Furthermore, we have observed that the subcomponents of the illusion, that is, ownership and referral of touch, show a different pattern for both objects.

In the following, we like to point out caveats in the interpretation of questionnaire responses. The comparison between synchronous and asynchronous conditions is one of the most commonly used approaches to determine the presence of the illusion (Botvinick & Cohen, 1998; Tsakiris & Haggard, 2005). The logic here is that the illusion is present in the synchronous but not asynchronous condition and that we can observe a statistical significant difference between the two. When applying this logic to our data, we can observe a statistical significant difference for both the hand and the balloon. This could suggest that the illusion or an illusion effect can be induced even for the noncorporeal object (albeit the latter shows lower effect sizes).

What speaks against this interpretation? First, the scores for the balloon condition are either negative (illusion and ownership) or weakly positive (RoT). Thus, most participants actually do not affirm aspects of the illusion experience in the synchronous condition. For example, the illusion score for the synchronous balloon condition is −0.25. This is similar to the score of the asynchronous RHI condition (−0.5), which is typically interpreted as a nonillusion. Second, all three scores of the synchronous balloon condition are not significantly different to 0. When advocating that any score greater than 0 should be interpreted as an affirmation of the questionnaire category (in fact, only the RoT score does for the balloon condition), evidence for a statistical significant difference to 0 can be provided by using a one-sample test. This approach did not show a clear affirmation of any of the illusion scores. Third, when applying a cutoff score of +1, we see that only about 20% of participants can be classified as illusion responders for the synchronous balloon condition. This is in contrast to the rubber hand condition in which more than 70% of participants have a score of >1. This is in line with previous observations, which showed that typically about 70% (60%–90%) experience the illusion (Kalckert & Ehrsson, 2014). When lowering this threshold to 0, we see that the balloon condition induces the illusion at best in 45%.

We have noted differences for the two subcomponents of ownership and referral of touch. The RoT sensation, that is, apparent fusion of visual and tactile stimulus, is always stronger than the ownership sensation, both in terms of relative score and amount of responders. Both these types of statements have been frequently used in studies, often interchangeably. It has been shown that they can represent different aspects or steps within the illusion (Valenzuela Moguillansky, O’Regan, & Petitmengin, 2013). Combining both scores increases ratings in conditions which may permit a RoT but not necessarily an ownership experience. For example, the synchronous balloon condition shows a relatively higher illusion score (−0.25) as compared with its asynchronous condition (−1.25). However, when splitting this score into RoT and ownership, we see that the ownership statements have a negative median score of −1, and the referral of touch +0.5. Thus, both these types of statements show a different pattern in the synchronous balloon condition. It can be speculated that other manipulations may differently affect these scores and potentially lead to different conclusions depending on which subcomponent is used to measure the illusion.

Ma and Hommel (2015) reported an illusion for a virtual balloon, which moved with the participants’ movements. In one of their experiments, they observed a main effect of synchronicity on different questionnaire items tapping into different aspects of the illusion like referral of touch, ownership, and agency, as well as SCR (Ma & Hommel, 2015). As outlined previously, differences between synchronous and asynchronous can occur, even in cases when the ratings suggest that participants do not clearly affirm the illusion-related statements. In the mentioned study and elsewhere by the same authors (Ma, Hommel, & Chen, 2018) the overall ownership ratings are rather low, which indicates a rather ambiguous illusion experience. Thus, we conclude that we cannot replicate an illusion experience with a noncorporal object like a balloon when using visuotactile stimulation.

The induction method is a possible explanation for this difference: Here, we provided visuotactile stimulation, in the study by Ma and Hommel (2015) participants were able to actively control the balloon. This addition of active control, or a sense of agency, may change the experience of ownership or sensory processes involved in the generation of the illusion more generally (Kalckert & Ehrsson, 2012). Previous studies on this matter did not find conclusive evidence for a stronger ownership experience after movement, and the exact mechanism if and how both these senses interact is not perfectly understood yet (Braun, Thorne, Hildebrandt, & Debener, 2014; Kalckert & Ehrsson, 2014).

Given these observations, it seems necessary to more clearly define how we use questionnaire ratings and under which circumstances we conclude that the illusion is present. A difference between synchronous and asynchronous conditions may be not sufficient. In case of the balloon, a statistical difference can be observed, but the scores are in the negative range of the scale (see Figure 2). Such a statistical significant difference can be still interpreted as an illusion effect in some way. The synchronous stimulations lead apparently to an increase of the illusion-related experiences, for both ownership and RoT, as reflected in the relatively higher (i.e., less negative) scores. However, given the relatively low ratings, strictly speaking a disagreement, it seems that most participants do not clearly experience the illusion for the balloon after visuotactile stimulation. In the light of this observation, we suggest that criteria other than a statistical significant difference between synchronous and asynchronous conditions may need to be considered.

In sum, this study highlights some of the caveats when using questionnaire data to assess the RHI. Our results illustrate the problem when drawing inferences based on a statistical significant difference between synchronous and asynchronous conditions. This, however, does not invalidate the general approach of using questionnaires to assess the illusion experience.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Abdulkarim

Ehrsson

H. H.

(2016). No causal link between changes in hand position sense and feeling of limb ownership in the rubber hand illusion. Attention, Perception, and Psychophysics, 78, 707–720. doi:10.3758/s13414-015-1016-0

Armel

K. C.

Ramachandran

V. S.

(2003). Projecting sensations to external objects: Evidence from skin conductance response. Proceedings of the Royal Society B: Biological Sciences, 270, 1499–1506. doi:10.1098/rspb.2003.2364

Botvinick

Cohen

(1998). Rubber hands ‘feel’ touch that eyes see. Nature, 391, 756–756. doi:10.1038/35784

Braun

Thorne

J. D.

Hildebrandt

Debener

(2014). Interplay of agency and ownership: The intentional binding and rubber hand illusion paradigm combined. PLoS One. doi:10.1371/journal.pone.0111967.s001

Ehrsson

H. H.

Spence

Passingham

R. E.

(2004). That’s my hand! Activity in premotor cortex reflects feeling of ownership of a limb. Science, 305, 875–877. doi:10.1126/science.1097011

Guterstam

Gentile

Ehrsson

H. H.

(2013). The Invisible hand illusion: Multisensory integration leads to the embodiment of a discrete volume of empty space. Journal of Cognitive Neuroscience. Retrieved from http://www.mitpressjournals.org/doi/abs/10.1162/jocn_a_00393

Kalckert

Ehrsson

H. H.

(2012). Moving a rubber hand that feels like your own: A dissociation of ownership and agency. Frontiers in Human Neuroscience, 6. doi:10.3389/fnhum.2012.00040

Kalckert

Ehrsson

H. H.

(2014). The moving rubber hand illusion revisited: Comparing movements and visuotactile stimulation to induce illusory ownership. Consciousness and Cognition, 26, 117–132. doi:10.1016/j.concog.2014.02.003

Kilteni

Maselli

Kording

K. P.

Slater

(2015). Over my fake body: Body ownership illusions for studying the multisensory basis of own-body perception. Frontiers in Human Neuroscience, 9, 613. doi:10.3389/fnhum.2015.00141

10.

Lenggenhager

Tadi

Metzinger

Blanke

(2007). Video ergo sum: Manipulating bodily self-consciousness. Science, 317, 1096–1099. doi:10.1126/science.1143439

11.

Lloyd

D. M.

(2007). Spatial limits on referred touch to an alien limb may reflect boundaries of visuo-tactile peripersonal space surrounding the hand. Brain and Cognition, 64, 104–109. doi:10.1016/j.bandc.2006.09.013

12.

Hommel

(2015). Body-ownership for actively operated non-corporeal objects. Consciousness and Cognition, 36, 75–86. doi:10.1016/j.concog.2015.06.003

13.

Hommel

Chen

(2018). The roles of consistency and exclusivity in perceiving body ownership and agency. Psychological Research, 1–10. doi:10.1007/s00426-018-0978-7

14.

Petkova

V. I.

Ehrsson

H. H.

(2008). If I were you: Perceptual Illusion of body swapping. PLoS One, 3, e3832. doi:10.1371/journal.pone.0003832

15.

Tsakiris

Carpenter

James

Fotopoulou

(2010). Hands only illusion: Multisensory integration elicits sense of ownership for body parts but not for non-corporeal objects. Experimental Brain Research, 204, 343–352. doi:10.1007/s00221-009-2039-3

16.

Tsakiris

Haggard

(2005). The rubber hand illusion revisited: Visuotactile integration and self-attribution. Journal of Experimental Psychology: Human Perception and Performance, 31, 80–91. doi:10.1037/0096-1523.31.1.80

17.

Tsakiris

Longo

Kosobud

Tsakiris

Haggard

Maquet

(2008). Looking for myself: Current multisensory input alters self-face recognition. PLoS One. doi:10.1371/journal.pone.0004040

18.

Valenzuela Moguillansky

O’Regan

J. K.

Petitmengin

(2013). Exploring the subjective experience of the “rubber hand” illusion. Frontiers in Human Neuroscience, 7, 1–18. doi:10.3389/fnhum.2013.00659