Ayaan Goswami

Implicit & Adaptive Game Tutorials

Problem: Many students struggle to stay engaged and learn effectively in traditional classrooms because lessons often feel boring, confusing, or not tailored to individual needs. While gamified educational tools aim to address this, many rely on surface-level game elements rather than the deeper design strategies that make games effective learning environments.

Research Focus: Video games frequently use tutorials to teach players mechanics and strategies. Some are explicit and instructional; others are implicit, allowing players to learn through experimentation and feedback. Adaptive tutorials further adjust guidance or difficulty based on player performance. This research examines how implicit and adaptive video game tutorials influence player engagement and learning outcomes.

Approach: By analyzing tutorial design patterns in successful video games and their impact on player behavior, this research explores how engagement, motivation, and learning are shaped by interactive and adaptive systems.

Why It Matters: My younger brother quickly masters complex video games but becomes disengaged with traditional learning methods. By studying how games teach effectively, I hope to identify strategies that improve gamified learning tools and make education more engaging and accessible for a wider range of learners.

Human vs. AI: Emotional Distress Recognition

Problem: Many people experience stress, anxiety, or sadness, yet access to empathetic and timely support is limited. While AI systems are increasingly used in emotional-support contexts, they often rely on keywords rather than tone, context, or lived experience — leading to misinterpretation or bias.

Research Focus: This project compares how AI models and humans identify emotional distress in written text, and how effective their responses are in reducing short-term negative emotion. The research examines differences in nuance recognition, contextual understanding, and perceived empathy.

Key Hypotheses

• Humans will better identify subtle emotional distress using context and tone.
• AI systems will perform well on explicit emotional language but struggle with nuance and cultural context.
• AI-generated responses may offer limited short-term relief but will be perceived as less empathetic than human responses.

Why It Matters: As AI becomes embedded in mental health, education, and communication platforms, understanding its limitations is critical. This research explores ethical AI use, bias mitigation, and human-centered design in emotionally sensitive systems.