nucleation

LIT-1276: Demand Characteristics Confound the Rubber Hand Illusion

3 min read

Properties4
StatusActive — read in full 2026-10-20 (NOTE-1005) from the author's final manuscript on PsyArXiv; worth reading as the direct test of the assumption behind every standard check on the rubber hand illusion, that expectations are matched between the "illusion" and "control" statements and between synchronous and asynchronous brushing. Twenty people who never underwent the illusion read a description of it, watched a 62-second video and rated, for each of Botvinick & Cohen's nine statements, how far they expected to feel it. They expected the illusion statements more than the control statements under synchrony (1.7 against 0.0, d = 1.40) and the illusion statements more under synchrony than asynchrony (1.7 against 0.4, d = 1.05): the pattern published reports show. Read it as showing that the usual controls cannot rule out suggestion, not as showing that the illusion is suggestion; the author says so. Small, pre-registered, sequential Bayesian stopping, one online sample of undergraduates, and no illusion or drift measured.
Tagsself, psychometrics, embodied-cognition, consciousness
Cited as “paper” byNOTE-1005 — Lush — Demand characteristics confound the rubber hand illusion
Cited as “source” byTHEORY-766 — Brushing a seen rubber hand in synchrony with the hidden real hand, against asynchronous brushing, raises reports that touch is felt at the rubber hand and, more weakly and less reliably, that the rubber hand is one's own; the proprioceptive drift toward the rubber hand that accompanies these reports is not a measure of them; the standard comparison conditions do not equalise participants' expectations, so whether the stimulation or the expectation produces the report difference is open

Graph View

Lush, P. (2020). Demand characteristics confound the rubber hand illusion. Collabra: Psychology 6(1), 22. DOI 10.1525/collabra.325. Preprint: PsyArXiv, DOI 10.31234/osf.io/mpncv.

Key takeaways

From the reading, NOTE-1005.

  • The controls rest on an untested assumption. The illusion is separated from suggestion by comparing illusion with control statements, or synchronous with asynchronous brushing. Both comparisons assume that “demand characteristics (and therefore expectancies) for control and illusion statements are closely matched”.
  • The assumption fails. People told about the procedure, but never put through it, expected the illusion statements more than the control statements (d = 1.40), and expected them more under synchrony than asynchrony (d = 1.05). The expectancy profile has the shape of published agreement scores, with the illusion items positive and the control items at or below zero.
  • Naive participants expect it too. The nine who had never heard of the illusion showed the same two differences (d = 2.30 and 1.68).
  • What follows, and what does not. The standard controls “cannot provide evidence that the RHI is not a suggestion effect”. The author is explicit that the results “do not provide evidence that the RHI is a suggestion effect, but rather that existing claims to the contrary are invalid”.
  • Implicit measures inherit the problem. Proprioceptive drift, skin conductance and imaging are tied to ownership only through their correlation with report. The demand characteristics of drift had, to the author’s knowledge, never been tested.
  • A remedy is proposed, not tested. Illusion and control statements should be matched on expectancy and on “suggestion difficulty”, with Bayes factors to support claims of no difference.

Standing in the record

Filed on 2026-10-20 at the owner’s request for experimental work on the rubber hand illusion that bears on the record’s four Proposed THEORYs about it. Read in full from the author’s final manuscript: Active.

It is the test of the gap that NOTE-1006 named in Botvinick & Cohen 1998 (LIT-1277): the intended answer is guessable from the set-up, and the control items guard against blanket agreement, not against suggestion specific to the illusion items. It shows that the guess is in fact made, in the direction and roughly the shape of the published results.

What it does to the four THEORYs, set out in the NOTE:

  • THEORY-765. It measures expectations, as that THEORY’s promote_when asks. But it measures them in people who were not stimulated, and it measures no illusion. So it does not meet the promote_when and does not trigger the refutation. What it shows is that the 1998 asynchronous control is not a suggestion control.
  • THEORY-769. It warns that graded ownership ratings may track graded expectation.
  • THEORY-767. It does not touch interoception.
  • THEORY-768. The author extends the confound to the full-body illusion on which that THEORY leans.

It also bears on QUESTION-048, on what report and “implicit” measures can establish about an experience, and on NOTE-372’s C6.

Filed alongside, in the 2026-10-20 batch on experimental work building on the illusion: Lush et al. 2020, phenomenological control (Nature Communications); Ehrsson et al. 2022 and Lush & Seth 2022, the exchange over it; Roseboom & Lush 2022 (LIT-1280); Rohde, Di Luca & Ernst 2011; Abdulkarim & Ehrsson 2015; Longo et al. 2008; Kalckert & Ehrsson 2012. It carries no instruction for machine-learning practice.

One of the sources of THEORY-766 (filed 2026-10-20), which restates the rubber hand synchrony effect as a difference in reports, separates drift from those reports, and leaves open whether expectation produces the difference.

Backlinks