A UX designer with a strong background in industrial design and over four years of experience. Daria is proficient in Figma, Adobe Creative Suite, 3D modeling, and project management, aiming to blend creative & technical skills to enhance user experiences.
This project explores the psychological impact of avatar design in virtual reality (VR), specifically focusing on how realistic and idealised avatars influence women’s self-esteem and their acceptance of plastic surgery. The study is situated within the broader context of increasing digital representation in social and professional environments, with implications for both user experience (UX) design and mental health.
Women experience more social pressure to conform to beauty standards.
Research indicates that there is a connection between low self-esteem, plastic surgery acceptance and intention, and the use of AR beauty filters, specifically in women from adolescence to adulthood (Alsaggaf, 2021; Javornik et al., 2022; Hermans et al., 2022; Maes and de Lenne, 2022; Chen et al., 2019; Eshiet, 2020).
All of the research on the impact of XR on plastic surgery acceptance was done using AR beauty filters.
Bessière et al., (2007) stated that the use of overly idealised avatars can be related to lowered self-esteem.
Avatar type definition was based on self-discrepancy theory, which is the most used theory to define why people create avatars that differ from themselves. This theory can be described as negative feelings when the user’s actual self has discrepancies with the ought or ideal self.
Avatar-type creation methodology was based on the Lascowitz and Huffstadt (2023) study that focused on similar topics and used a cartoonish avatar style to describe idealised avatars and 3D scanned to describe realistic avatars.
The development of the prototypes was guided by the design thinking framework, which allowed for an iterative, user-centred approach.
The prototype development focused on three main aspects: environment design, avatar creation, and task interaction.
The iterative process was not linear, with steps being retraced because of user feedback and technological limitations.
These stepbacks were addressed through design adjustments and by focusing on the research aim and questions.
I started with the empathy stage for the environment, using short interviews and card sorting to gain users’ expectations from both environment types.
3 platforms that were user-tested for realistic avatars and two highlighted were chosen for this study. It used the Ready Player Me avatar creation platform that had integration into the Spatial.io
I ideated 2 ideas each and body-stormed them with users.
Participants preferred small talk interaction with AI chatbot-generated dialogue.
I developed some sketches based on the findings from the empathy stage and moved through the levels of fidelity.
Unfortunately, I could not use the environments shown, due to technical limitations of the platforms. So, the pre-existing environment that followed the findings from the empathy stage was used in spatial.io. -platform that was chosen to host the prototype.
To allow embodiment realistic avatars had to be rigged, however, the process was toolong and did not produce high-quality results. Another limitation was the implementation of the avatars to the platform, which was not possible.
So, it was chosen to only expose participants to the realistic avatar.
To ensure the study was controlled and reliable, specific tasks and interactions were developed for users. Participants engaged in small talk with another avatar, simulating networking or dating scenarios, with dialogue driven by a combination of scripted and ChatGPT-generated interactions.
On the right, you can see specific things either chosen or incorporated to follow the user-centred design.
Based on empathy stage findings, the curvature of the walls and animated avatars were chosen. Mirrors were added based on the research aim and questions
A mixed-methods, within-subject study design was employed, where 12 female participants interacted with both avatar types in both environments. Self-esteem was measured using The Rosenberg Self-esteem Scale (Rosenberg, 1965), and plastic surgery acceptance was assessed using The Acceptance of Cosmetic Surgery Scale (Henderson-King and Henderson-King, 2005). Qualitative data were collected through exit interviews and analysed thematically.
Thematic analysis of the qualitative data shows an interesting dichotomy.
Both avatar types had their advantages and disadvantages, based on user feedback.
The idealised avatars have a positive impact on self-esteem; however, qualitative data suggests that they might increase plastic surgery considerations for some women because of increased self-scrutiny.
Idealised avatars prompted some participants to consider enhancing their appearance, though not directly leading to plastic surgery intentions. This aligns with similar effects seen in AR face filters found by Maes and de Lenne, (2022).
Realistic avatars were seen as more professional and trustworthy but may lead to an increase in self-consciousness and negative body image.
Realistic, 3D-scanned avatars did not enhance self-esteem and often led to discomfort, linked to the uncanny valley effect defined by Mori, (1970). Participants felt less connected, similar to the findings by Mancini and Sibilla (2017). Another interesting topic was VR dating, users agreed that they needed some form of verification to feel safe in this context.
EcoSpark is a mobile app for environmentally-aware teens, that encourages them to explore and implement sustainable lifestyles in a meaningful way, overcomes their worries about environmental crisis and helps them find like-minded peers so that their individual impact is supported by collective efforts and they are reassured that their actions are meaningful and sufficient.
ComPals is a platform where every new journey begins with guidance and friendship. Inspired by the word “compass,” ComPals is designed to be your guiding star in the world of remote onboarding. We believe that starting a new role should feel like joining a community, not just a company. That’s why we’ve infused our platform with the essence of “Pals,” reflecting our commitment to fostering genuine connections and support among team members.
Ribblr AI is an AI-powered feature integrated into the Ribblr knitting app. This feature identifies and provides solutions for knitting mistakes by analyzing photos of the user’s project, enhancing the knitting experience for novices and intermediates. The project focused on ensuring high accuracy and user satisfaction, with strategic integration into Ribblr’s subscription model to increase engagement and revenue.
