The short version of our research is that some scientists got together and had this conversation:

“Hey, hey, you know that iconic study where researchers made people think a rubber hand belonged to them?!”
“Yeah! and how it’s been used in research about racism, pain, empathy, and like, basically everything?”
“WAIT. WAIT. What if we did that. But with someone’s WHOLE BODY?”
“Oh, that’s been done before too.”
“No, no, no, what if we did it with a different SIZED body?”

…and then some scientists turned that into a serious sounding grant application and now here we are, making people believe they inhabit slimmer and larger bodies.

But why? I mean, besides the sheer fun of it, of course. Because eating disorders. One of the common clinical assumptions about eating disorders, especially anorexia nervosa, has been that a component of not-eating comes from inaccurately perceiving body size. Specifically, people with anorexia think they’re larger than they are₁. An interesting demonstration of this phenomenon involves making people with anorexia walk between progressive narrower spaces, and then making non-anorexic controls do the same. From the study [AN = anorexia nervosa participants, HC = healthy controls]:

AN patients started rotating for openings 40% wider than their own shoulders, while HC started rotating for apertures only 25% wider than their shoulders. The results imply abnormalities in AN even at the level of the unconscious, action oriented body schema. Body representation disturbances in AN are thus more pervasive than previously assumed: They do not only affect (conscious) cognition and perception, but (unconscious) actions as well.  (Here’s an interesting followup investigating this behavior in perspective-taking)

Anyways, the question was, it seems like perception of your body size being inaccurate would result in an eating disorder symptoms…except that thus far, we’d not determined a causal connection. BUT, what if we could make people perceive their body incorrectly?  If we could make them think their body was larger or smaller than they thought, would we see a measurable increase or decrease in eating disorder pathology? Okay, but how do we do this?

Creepy headless mannequins.

Obviously.

The illusion takes inspiration from the rubber hand studies. Picture this: you hide the participants arm from their field of view–put it in a box or somesuch. Then, in full view, you put a rubber hand. Tickle the hand with a feather, while simultaneously tickling the participant’s hand. Repeat, with a variety of sensations: tapping, tickling, rubbing. Then threaten the rubber hand: jab it with a pin₂, make as though you’re about to stab it with a knife. The participant will react as if you’d just jabbed them with a pin, often experiencing pain in the location of the pin in the rubber hand, or sharply increasing heart rate and comical flinching when menaced with a knife.

Our proprioception–the sense of knowing where your body is in space–is frighteningly malleable. If it appears that that detatched hand on the table in front of you is yours, and if you feel sensation when you see the hand touched, well, then of course that thing in front of you is your hand.

Further, after a bunch of duplications and manipulations of the rubber [bodypart] illusion, we also know that post-experiment, people have weird beliefs about their actual body, thinking features are longer or larger. So, if we trick you into believing the rubber hand with the long fingers is yours, then stop the experiment, when you look down at your actual hands, you might perceive the fingers to be longer. Now, let’s do it with a whole body.

So of course, the first thing you have to do is pull the head off the mannequin and replace it with a camera, angled downwards so the mannequin is facing downwards, and looking at the shape of its body. That camera feed is then linked to a set of goggles, which the participant wears. One RA, who will probably avoid clearly describing her job on her resume, stands between the mannequin and the participant, and touches both simultaneously.

Participant feels: contact of RA touching him.
Participant sees: view from mannequin perspective, with a small or large body, but also sees a hand touching the mannequin.

Do this for enough trials, and the participant’s brain will synchronize the incoming information and, rather than keep trying to remember that the mannequin is a mannequin, decide that all the incoming sensory information lines up, and the view through the goggles is the view when looking down.

Then the RA threatens the participant with a knife₃.

Well, not quite. The RA pretends to slash the mannequin with the knife, which is captured by the camera and relayed to the participant. Skin conductance, a known measure of fear, is taken, and we can get a fairly good guess about whether the participant thought we had made them sign a consent form and then stabbed them. (This is the scientific equivalent of “Made ya look, didn’t I? Didn’t I?)

This process is repeated with the slimmer mannequin and the larger mannequin for each participant. Between each illusion, they’re run through a battery of tests measuring body satisfaction and measures of eating disorder psychopathology. Two results stand out.

1. Firstly, in the larger body condition (LB), the participants didn’t seem to have much of an emotional reaction. Body satisfaction didd’t change significantly, nor did participants misperceive their actual body to be larger. The mannequin was 115% of the size of the participants (who all clustered around the same BMI). Increasing your size 15% is significant, but unlike in the smaller body (SB) condition, participants seemed non-reactive.

A few theories here. It might be that the participants just entered with inaccurate perceptions about their bodies, believing them to be larger than they are. So, when presented with a larger body, in front of them, they just didn’t notice much change. Also interestingly, it seems as though size-change illusions aren’t asymmetric. Participants in other research more readily believed that they possessed doll-sized legs than giant legs.

The third part of this isn’t directly related to the results, but the researchers noted that the LB condition still had a flat stomach and muscle definition, so it may be that the participants still viewed themselves as having a socially desirable body. I’d want a replication with a flabbier dummy.

2. ‘Owning’ a smaller body did have an impact on body satisfaction. Predictably, it increased. However, I want to point out that all of these participants had normal-BMIs, and the slim mannequins were approximate 85% the size of the participants. Or, to put it another way, the bodies were at the size considered the cutoff for anorexia.

Secondly, after the SB condition had ended, and participants were away from the mannequin, they believed their actual body to be smaller, and gave smaller estimated hip-size numbers than prior to the condition. This, as noted in 1, didn’t occur with the large body.

—

I admit, I completely expected the LB condition to have the larges effect, emotionally and in terms of ED psychopathology. Even funhouse mirrors can startle and upset me, and I assumed that believing you owned a larger body would cause the same. Importantly, these were non-clinical participants–without an eating disorder diagnosis. It may be possible that they simply had more immunity to such an experience, or that their interactions with proprioception are fundamentally different.

Either way, I think we can agree that creepy mannequin studies are creepy.

Full paper located here. [I didn’t distinguish strongly between Experiment 1 & 2, as found in the paper, and didn’t cover all of the scales and correlations]

₁Caveat: historically, anorexia has included an upper bound of weight in the definition, meaning that not-eating-while-overweight wasn’t called anorexia. This radically impacts who is studied and while it leaves me frustrated at research, I can’t conjure up better methodology retroactively. So, this statement reads better as “people in a specific definition of anorexia who would probably be described as slim already think they’re larger than they are”

₂This is a stickup!
…I’m so sorry. 

₃“And what were your duties while working at the Body and Self Laboratory?” “Well, I beheaded mannequins and threatened people with knives.” “You’re hired!”