.@Laramaister had the genius idea of reverse-correlating one's self, and @mattlongo Sophie De Beukelaer & @manos_tsakiris followed suit. And what happens?
Preprint alert 'The Self in the Mind’s Eye' with @Laramaister @mattlongo Sophie De Beukelaer @manos_tsakiris from BIAS project @Warburg_News @RHULPsychology @BirkbeckUoL @BangorUni
psyarxiv.com/f2b36/
psyarxiv.com/f2b36/
How do we see ourselves in our mind’s eye ? The question of how we represent our self has been at the centre of cultural practices across centuries (look at self-portraits), & at the centre of our understanding of mental health issues such as body-image disorders. 
Looking at Van Gogh’s self-portraits across time. Self-portraits are not just a representation of the actual physical appearance of the artist, but also an exploration of the artist’s identity, emotions, and beliefs.
And of course the question of self-(re)presentation has major consequences for modern cultural practices and mental health
In mental health , questions about how we mentally represent our body-image are at the centre of disorders relating to body-image dissatisfaction and body-dysmorphia as well as in cultural practices (e.g. self-objectification & selfies)
Is there a way to visually depict, for all to see, how people "see" themselves with their minds’ eyes? And if so, what can we tell about the mental portraits of ourselves that we hold in our minds?
For the first time, we use a computational reverse-correlation technique to explore individuals’ mental ‘self-portraits’ in an unbiased, data-driven way.
Reverse-correlating one’s self : an average base face (A) is superimposed by a randomly-generated sinusoidal noise to produce (B) & with the same noise pattern inverted (C). Pairs of faces are presented to participants & they select the one that looks most like themselves.
This is a visualization of a participant’s production of her mental self-portrait. Each frame represents her chosen frame as more self-looking and the video shows the progression over 400 trials and the last frame is what we call the mental self-portrait
This is another participant
Starting from the same base face, these two participants ended up with different mental self-portraits
We ask :
Do mental self-portraits look like the participant?
Can external raters reliably infer personality traits from self-portraits?
Are the features of self-portraits influenced by the psychological self?
Which individual differences explain variance in the self-portraits?
Do mental self-portraits look like the participant?
Can external raters reliably infer personality traits from self-portraits?
Are the features of self-portraits influenced by the psychological self?
Which individual differences explain variance in the self-portraits?
Overall, our technique was successful insofar people’s mental self-portraits were accurate as assessed objectively using a face-recognition algorithm as well as external raters. So far so good!
Moreover, external observers could perceive in the mental self-portraits the self-reported personality traits. The higher participants rated themselves on a certain personality trait, the more facial features associated with that trait were present in their self-portrait.
Finally, the higher the participants’ self-esteem with regards to social interactions, the more accurate and ‘true to life’ their self-portraits were.
Next we turned our attention to the mental self-portraits of bodies
Unlike the facial self-portraits, the body-shape portraits had negligible direct relationships with individuals’ actual body shape, but as with faces, people with lower body self-esteem visually represented their bodies as wider.
By implementing a reverse-correlation technique to measure self-representations, we were able to visualise participants’ mental images of both their faces & their body-shapes in a data-driven, unconstrained way, allowing us to visually depict how the self is ‘seen’ in our minds
We find individual differences in how accurate people’s mental pictures of their own appearance really are. Furthermore, these self-representations are biased, essentially ‘coloured in’ by individuals’ attitudes towards themselves, and their beliefs about their personalities
Marcel Proust suggested that we fill the physical appearance of the individual we see with all the notions we have about him, and of the total picture that we form for ourselves, these notions certainly occupy the greater part.
It seems we do the same for ourselves
It seems we do the same for ourselves
Our evidence shows the close interaction between different aspects of self-representation, and raises intriguing possibilities for understanding body-image disorders and our cultural practices of portraying the self. Happy readings!
And any artists who may be interested in using our approach to "draw" portraits please get in touch
Lastly many many kudos to @Laramaister , one of my favourite #WomenInScience , who came up with the original idea and saw through that project to such a fine end !
