Student Projects

My doctoral research examines how virtual reality and augmented reality shape body perception, especially body-weight perception, and applies these insights to intervention design for anorexia nervosa therapy. I run controlled studies that vary avatar self-similarity, agency, embodiment strength, latency, and the use of augmented versus fully virtual settings to model their effects on perceived body weight, self-identification, and body ownership.

My doctoral research explores the use of virtual reality and embodied AI for the treatment of mental health disorders. Especially, the effects of ethical principles and human-centered design are the focus of this research effort. Within controlled studies, we examine self-disclosure and trust towards the embodied conversational agent, the monitoring of physiological measures during the interaction and the overall effectiveness of our applications.

Humans often feel ownership for physical objects, even if they do not legally belong to them. But since the digital world has taken a hold of our lives, we also have taken a hold (and ownership) of digital and virtual objects. My thesis aims to explore how the "offline" concept of Psychological Ownership presents itself in VR and XR, if (spoiler: yes) and how we develop ownership of virtual objects and what effects these feelings of ownership have on our self, our actions and how we interact with the virtual world.

Harassment in social VR can be a deeply distressing experience, intensified by the immersive sense of presence and embodiment. This doctoral research examines how embodiment and avatar design influence both the likelihood and the impact of harassment — and how harassment, in turn, reshapes users’ identification with and experience of their avatars. By combining observations in social VR communities with controlled experimental studies, the project investigates how factors such as realism, personalization, anonymity, and identity cues affect vulnerability, coping, and social interaction. The ultimate goal is to develop evidence-based design strategies that reduce harm and support safer, more inclusive, and empowering social VR environments.

This dissertation investigates the impact of different devices on key factors such as self- and other-perception in social virtual interactions. Specifically, it examines how varying device configurations influence collaboration in virtual environments, with a focus on ensuring high accessibility and supporting various devices. The goal is to answer the question of whether social virtual interactions remain effective when users with different device configurations engage with one another.

How should social VR applications be designed to support and promote intercultural and transcultural competencies and sensitivity? This work will contribute to the evaluation of how different immersive features of (social) VR affect inter- and transcultural competence and sensitivity. Different approaches will be investigated. Does the representation of the virtual environment or the embodiments of peers influence facets of intercultural competence, such as acceptance of stereotypes or tolerance for ambiguity? Can implicit stereotypical behaviors be deduced with the help of virtual objects?

The aim of my work is to create a basis for the question whether virtual reality (VR) and virtual embodiment can be a suitable medium to treat disturbances of sensory body awareness. This involves a series of investigations that examine (a) how and to what extend VR can draw attention towards the physical body compared to non-VR methods and (b) the relation between body awareness and the illusion of virtual body ownership.

By research through design, I’m going to explore use and effect of Image Schema in the field of Data Physicalisation. In an iterative design research process I’m going to create design artefacts and evaluate them by qualitative methods, in order to develop an Image Schema based toolkit to support the design process of Data Physicalisation. 

This doctoral thesis mainly addresses the questions of how body dimensions of virtual representations are perceived in virtual environments and by which factors they are influenced. It investigates how exposure to virtual self-representations in virtual environments can change the perception of one's own body and how sustainable this effect is.

In virtual social interactions, embodied users' virtual representations, so-called avatars, can have different characteristics of appearance and presented behavior. This doctoral thesis aims to investigate the impact of the attributes of virtual others' avatars, with whom one has social contact in XR environments, on a user's perceived quality of communication, and possibly also his/her self-perception.

Today's virtual environments offer only limited sensory modalities, especially for collaborative manual tasks this can become a complex challenge for participants involved. This thesis deals with the development and analysis of sensor substitution techniques to increase usability and UX in collaborative tasks within a distributed virtual environment.

In this work, intuitive use is defined as the extent to which a product can be used in a mentally efficient and effective manner, which is accompanied by a strong metacognitive feeling of fluency. Since available methods for the formative and summative evaluation of intuitive use lack sufficiently high time efficiency, this work proposes two new rhythm based evaluation methods for intuitive use using the secondary task paradigm: 1) IntuiBeat-F as a formative method and 2) IntuiBeat-S as a summative method.

The aim of this project is to explore whether embodying an avatar with increased attractiveness influences self-esteem. To increase the avatar's attractiveness, the participants can modify the body weight of their avatar to match their ideal body weight. We implement a mirror exposure in virtual reality, where participants will embody personalized photorealistic avatars configured to their self-defined ideal body weight.

Due to the large amount of people suffering from mental disorders worldwide, the deployment of Virtual Reality (VR) and Artificial Intelligence (AI) in a therapeutic context are being researched. This master thesis aims to assess the impact of increased transparency regarding the AI's function and provided autonomy regarding the data collection on the perceived trust towards the AI therapist as well as the level of self-disclosure during the interaction.

Professional negotiations can have a significant impact on one's personal future, and a lack of negotiation skills can negatively influence salary and contract conditions. Virtual Reality (VR) training with embodied intelligent agents as negotiation partners can improve negotiation skills. Based on a professional negotiation scenario, the developed VR application can be used to train negotiation skills, as well as research aspects of human-agent interaction.

Empathy, especially affective Empathy, is known to be an effective inhibitor of cyberbullying actions. Therefore this thesis elaborates how to foster Empathy in the context mentioned using Virtual Reality as a medium and Personal Space Invasions as a stylistic mean to transport the feeling of Cybervictims.

From day to day navigation tasks become more complicated and stressful. One promising technology supporting driver safety and providing good user experience is augmented reality. This thesis develops and evaluates innovative augmented reality concepts for navigation tasks.

Does a crossmedial assembly assistance system - combining smart glasses and tablet computer - facilitate procedural instructions for a manual bike part assembly? A prototype system was built and evaluated to examine synergy effects of the devices.

Immersive technologies offer effective facilitators for effective and sustainable behavioral change - an example: "I feel that I can do it in VR, so I can do it in reality" (high feeling of effectiveness and control). VR can, for instance, be used as a new and innovative intervention to reduce the attitude-behavior gap. By altering the perspective and/or the interactivity, the individual feeling of effectiveness and thus the individual environmental behavior should be strengthened. 

In the future, autonomous vehicles will enable motorists to no longer have to concentrate on traffic. As passengers, former car drivers will be able to use their time while driving for other purposes, such as the consumption of media content. This thesis deals with the possible application of technologies and methods from the research area of virtual reality (VR) in an autonomous driving vehicle.

In this thesis an interactive carbon footprint calculator is developed, which gives users immersive feedback about their carbon dioxide emission. Besides calculating the individual carbon footprint, changes for future pro-environmental actions are designed.

Attention A.I. helps to manage drivers’ attention between the joyful experience of watching a video on a head-up display and the safety of monitoring the driving environment in automated vehicles. The system predicts when drivers are likely to switch their attention due to events outside of the vehicle by using machine learning and compares them with the actual behaviour to decide if an intervention is necessary.

The project investigates how full-body VR systems influence presence and sense of embodiment, central components for a convincing immersion in virtual worlds. To this end, a minimal hardware-based system (Meta Movement) and a marker-based outside-in system (HTC Vive) are systematically compared in order to evaluate the effect of the system on user experience and performance. The aim is to clarify the extent to which the reduction in VR hardware affects presence and sense of embodiment and which solution offers the best balance between quality of experience, costs and setup for practical contexts such as education or therapy.

This thesis presents the first version of a questionnaire on psychological ownership (for virtual objects). Although several questionnaires on psychological ownership already exist, they are often tailored to specific scenarios and ignore the fundamental routes that, according to theory, are necessary for developing psychological ownership. The questionnaire was validated with approximately 80 people and already shows good internal consistency, and the subscales based on the routes also appear to have been chosen appropriately. The contribution of the questionnaire lies particularly in its proximity to theory and its general adaptability to a wide variety of scenarios.

The thesis explores whether and how the embodiment of a person (i.e., one's own representation in VR through a virtual body or avatar) affects the development of psychological ownership of virtual objects. Psychological ownership refers to the claim of ownership that people can develop toward objects or concepts. In this project, special attention is paid to the effect of avatar personalization based on psychological ownership theory, personalized embodiment could enhance the effect. To this end, we are conducting a study in which participants perform a task with a virtual object that is designed to generate ownership of the object.

What if a virtual therapist could instantly respond to a patient’s emotional state based on voice cues? This project builds a modular speech analysis tool that captures fundamental frequency (F₀) in real time, enabling a dynamic adaptation of the virtual therapist’s behavior in Virtual Reality Exposure Therapy (VRET). By integrating these vocal insights into an AI-driven environment, therapy sessions become more personalized, empathetic, and potentially more effective.

As artificial intelligence (AI)-driven virtual agents become more common in work environments, trust is key to user acceptance. But what if users only think they have control? This thesis explores how the Illusion of Control (IoC) affects trust in AI agents within virtual reality (VR). We examine whether static placebo settings, interactive configuration, combining illusion with actual control, and repeated control interactions enhance trust. Findings will provide insights into perceived control in AI trust and the ethical balance between empowerment and deception.

Wouldn't you like to be able to ask for guidance when facing a complex problem or when the solution is not obvious? Powerful AI and a voice-based natural language interface could one day make this a reality. This thesis investigates if the presence of a smart speaker causes an observer effect (i.e. social facilitation | inhibition effect) and if such an effect extends to creative problem-solving.

The bachelor thesis evaluates different coping and mood regulation strategies presented by a conversational agent (built with BOTfriends X). It investigates the impact of the agent's characteristics and personality traits on change behavior while giving the users advice on coping with the challenging times of the ongoing covid-19 pandemic.

Recent research suggests that non-experts have incomplete mental models about smart speakers which can lead to privacy and security threats. Addressing their shortcomings using a competence learning program might support users in their decision-making process regarding privacy and data sovereignty.

Source credibility is an important part of every journalistic report. The MAIN-Model of Sundar (2008) proposes that credibility can be enhanced by increasing different affordances like agency or interactivity. In this work interactive elements, like info-hotspots, were added to a journalistic 360° video with the goal to create a more credible contribution.

Special trust in such systems is required for the retrieval of personal data such as illnesses by voice assistants. In cooperation with the startup vuemondo, the project aims to find out how language assistants need to be designed to inspire this trust.

In this thesis, the social interaction with robots in an industrial context is investigated under consideration of the media equation, whereby the influence and the emergence of "social presence" are particularly relevant. The impact of social interaction on employee productivity and work climate is also considered.

The aim of this project is to explore whether embodying an avatar with increased attractiveness influences self-esteem. To increase the avatar's attractiveness, the participants can modify the body weight of their avatar to match their ideal. We implement a mirror exposure in virtual reality (VR), where participants will embody personalized photorealistic avatars configured to their self-defined ideal body weight.

This project investigates the comparability of eye-tracking data in the real world, as well as in virtual reality in the context of visual body perception. Highlighting the differences in gaze behavior between both fields is important for the use of eye-tracking in future virtual reality applications.

Research goal of this project is to examine whether the extension of interactivity in a 360 degree video has a positive effect on credibility, story recall and story sharing intention. Interactive Elements such as info hotspots and scene switch buttons were designed following the MAIN Model by Sundar (2008), which proposes a heuristic approach of understanding credibility of digital media.