Our interdisciplinary longitudinal study investigates the evolving dynamics of human-AI interaction over six waves spanning one year. By examining individual, behavioral, and task-related variables, the project aims to uncover how users' trust in, perceptions of, self-efficacy, and willingness to engage with AI systems develop and interrelate over time. The insights gained from this research are essential for better understanding human-machine interaction, a critical foundation for fostering effective collaboration between users and AI systems. This knowledge will inform user-centered AI design and guide the ethical integration of these technologies into various aspects of everyday life.
How do citizens form beliefs about politicized topics science such as climate change, COVID-19 or vaccinations? In this project, we illuminate the role of metacognition, the insight that citizens have into the reliability and fallibility of their own knowledge and reasoning.
Multimodal large language models (LLMs) generate texts based on image inputs. This makes them attractive for a wide range of applications where a large amount of image data needs to be processed. One of these applications is the cataloguing of archival images. ArchiveGPT thus focuses on applying a multimodal LLM to archaeological photo material provided by the Leibniz-Zentrum für Archäologie (LEIZA) in Mainz.
Belief in climate change misinformation such as denying its anthropogenicity has been shown to confuse the electorate, and stall political action. While the extent of the public’s incorrect beliefs is comparatively well-understood, its causes are subject to an ongoing debate. One of the most prominent questions centers around the role of social media: Does social media use drive widespread belief in science-related misinformation?
How should we communicate politiczed science such as climate change or COVID-19? We explore how metacognition, the insight that citizens have into the reliability and fallibility of their own knowledge, interacts with science communication to shape behavior and beliefs, and how metacognition is shaped back by science communication.
The comprehension of different types of narratives, such as text, pictures, or comics, is important for societal participation. This dissertation project investigates how narrative comprehension changes with age and which factors contribute positively or negatively to narrative comprehension. These findings are essential for the development of interventions for different age groups and thus for greater societal participation.
It is generally believed that humans prefer information that confirms their attitudes and avoid information that represents opposing views. Striving for confirmation and congeniality are also held responsible for a number of toxic phenomena on the Internet, such as the emergence of echo chambers and filter bubbles, the polarization of society, or the dissemination of misinformation. The present project investigates how people deal with opposing opinions – are they really ignored?
Narratives communicate information in many ways, for example in books, audio dramas, films, or visual narrations like comics. While there is extensive research on text or film comprehension, relatively little is known about comic comprehension. Visual narratives, however, offer many possibilities in formal and informal education settings. This project therefore addresses the question how we comprehend and process visual narratives like comics.
While historically, the aim of propaganda was to convince citizens of a certain agenda, novel forms of disinformation come with a different goal in mind: To confuse, rather than convince. Or, as former president Trump’s advisor Steve Bannon put it: “The Democrats don’t matter. The real opposition is the media. And the way to deal with them is to flood the zone with shit”. Although this zone-flooding strategy poses a serious threat to democratic functioning, it currently lacks empirical investigation that maps out its effects on citizens. We conduct a rigorous, pre-registered investigation into the effects of zone-flooding that harnesses state of the art-methods from Signal Detection Theory and metacognition to illuminate pressing questions: Does zone-flooding affect citizens’ ability to distinguish truth from falsehood? Does it affect their insight into the accuracy of this distinction? Does it render citizens more skeptical or more gullible? And are these effects politically symmetrical?
In various research areas and topics such as climate change or testimonies it has already been demonstrated that mental representations are influenced by true and false information. Problematically, it becomes increasingly difficult to identify false information in our daily lives. Furthermore, new technologies simplify the creation of realistic-looking false messages in media. This dissertation project, therefore, addresses the question of how discriminability of information influences mental representations.
Event perception and cognition theories assume dynamic events are segmented into meaningful chunks of sub-actions with partonomic relationships. This allows viewers to process streaming information in units and predict future states of action based on their expectations and event knowledge. Event models store relevant information for events and guide perception using schemas (or scripts). While event models hold immediately accessible representations stored in long-term memory, working event models process perceptual representations of unfolding activity throughout the event.
How do people perceive dynamic media such as educational videos, movies, or soccer broadcasts? Human information processing is specialized in processing dynamic information. It distinguishes relevant, and thus informative, information from irrelevant information. This project follows two research lines, bridging the gap between cognitive psychological theories of event cognition and typical situations of media reception. On the one hand, we investigate the perceptual and psychological foundations of dynamic event perception by specifying, for example, the processes of encoding and the properties of mental representations of natural action sequences. On the other hand, we use cinematic stylistic devices (e.g., different camera perspectives, film editing) and new cinematographic film techniques (e.g., 3D films) to explain basic psychological processes, such as the experience of spatial presence or the experience of suspense.
The human environment is populated by other agents. Other humans make up an important part of this, but animals and increasingly autonomously acting machines, i.e., artificial agents, also act without explicit prompting. As a result, their actions are sometimes incomprehensible. This hinders joint work, which is meant to be facilitated by automation. An important aspect of any interaction with other agents is understanding the other. This includes assessing the capabilities of the other person. In a cooperative situation, it is important to be able to assess how much you can rely on your partner. In a competitive situation, you need to assess in which aspects you are superior.
As the world becomes increasingly technologically forward, the presence of artificial agents in day-to-day life also becomes more apparent. Studying the interaction between humans and artificial agents, such as robots, has long been in the research spotlight. While research in this field is traditionally focused on how robots can improve our lives, this PhD project aims at flipping the focus on humans helping robots.
In societal discourse, Artificial Intelligence (AI) is strongly tied with both opportunities and risks. In this project, it is investigated how humans perceive the risks of AI and how their risk assessments are associated with psychological factors like prior knowledge and judgmental confidence. The behavioral consequences of risk perception are investigated, as well as intervention methods aimed at raising an awareness of AI risks.
Generative Artificial Intelligence is capable of generating texts or images based on verbal prompts. With its universal range of application fields and the human-like output quality the interaction with generative AI becomes increasingly similar to the interaction with other humans. How does interaction with generative AI impact human behavior, understanding, and trust, and how can these insights be used to optimize human-machine collaboration?
How can we protect biodiversity and the climate while also ensuring stable and resilient food supplies? LL-SUSTAIN is a research project aiming to answer this vital question. Recognizing that these goals can sometimes conflict, the project seeks solutions that respect our planet’s limited resources. By uniting experts from various fields, LL SUSTAIN aims to identify gaps in our understanding of sustainability related to biodiversity, climate, agriculture, and food. This interdisciplinary approach helps create a comprehensive view of these interconnected areas. A key focus of LL-SUSTAIN is connecting scientific research with public discussions. The project prioritizes integrating knowledge and finding transformative solutions that can be applied at local, regional, and international levels. By fostering innovation and supporting informed decision-making, LL-SUSTAIN aims to contribute to societal changes toward overall sustainability. An important part of the project involves developing effective ways to communicate scientific information to the public. This includes addressing misinformation about sustainability. LL-SUSTAIN reviews and synthesizes different methods of science communication, develops educational materials tailored to various audiences, and creates training programs focused on biodiversity, climate, agriculture, and food as interconnected fields. The project also explores new methods for sharing knowledge and technology. Consideration would be given to tools that make complex information more accessible and engaging, as well as support both classroom learning and individual education.
In nearly all educational settings (schools, universities, further education), videos play an increasingly large role. On video portals learners can deepen and broaden their acquired knowledge and while watching they leave traces like pauses or skips. This cooperation project investigates how such usage data in conjunction with videos automatically prepared under pedagogical and psychological considerations can be harnessed to make video learning adaptive and effective.
This PhD project explores the role of video structure in influencing learning outcomes as well as popularity. By studying structural elements of videos through cognitive psychology, event segmentation theory, and basic shapes of narratives, the project examines how these impact measures of learning and video popularity. The research combines traditional psychological methods with big data and machine learning techniques to gain comprehensive insights into video impact.
Do people know about their own knowledge? And how does this relate to their experience and behavior? In this project, we investigate how insight into ones’ own cognition relates to information selection and processing, as well as opinion and judgment formation.
Video-SRS is an interdisciplinary project that focuses on supporting video learning in mathematics by exploring and improving self-regulation. It combines insights from cognitive and educational psychology, mathematics education, and computer science to develop innovative approaches to this type of learning. The use of responsible machine learning algorithms plays a significant role in this process.
