What are some topics for a final year dissertation relating to UX and technology

Explanation: As conversational agents (chatbots and voice assistants) become more widely used, users often need to understand how and why these systems behave as they do — especially when the agent makes recommendations, interprets ambiguous inputs, or takes actions on the user’s behalf. Designing explainable user experiences (XUX) focuses on creating interactions that make the agent’s reasoning, limitations, and consequences transparent, intelligible, and actionable for diverse users. Key points to cover: - Purpose: Improve user trust, decision-making, error recovery, and perceived control by providing concise, context-sensitive explanations about the agent’s inputs, processes, confidence, and outputs. - Types of explanations: Procedural (what the agent did), evidential (what data or signals led to a response), confidence indicators (certainty/ambiguity), and corrective guidance (how users can rephrase or provide missing information). - Interaction modalities: Tailor explanations to modality constraints—brief, spoken explanations for voice assistants vs. richer visual/textual affordances for chatbots or multimodal interfaces. - Timing and granularity: Balance interruption cost and cognitive load—offer lightweight inline cues with optional granular explanations on demand (progressive disclosure). - Personalization and user models: Adapt explanations to users’ expertise, goals, cultural expectations, and privacy concerns. - Ethical and practical considerations: Avoid exposing sensitive data or complex internal models that confuse users; disclose limitations and potential biases; ensure explanations do not create unjustified overtrust. - Evaluation: Combine qualitative (think-aloud, interviews), quantitative (task success, trust/confidence ratings), and behavioral measures (help-seeking, correction rates) to assess effectiveness. Why this is a strong dissertation topic: - Interdisciplinary: Sits at the intersection of UX, HCI, AI ethics, and NLP—allowing literature from multiple fields. - Practical relevance: Industry demand as conversational agents proliferate in customer service, healthcare, finance, and smart homes. - Research gaps: Need for principled design patterns, guidelines for spoken explanations, measurable evaluation methods, and approaches that respect privacy while remaining useful. - Deliverables: Usable prototypes, design guidelines, and empirical evaluation studies are achievable within a final-year project. Relevant starting references: - Ehsan, U., et al. (2021). "Towards Practical Explanations for AI: A Survey and Research Agenda." (for explainability concepts) - Liao, Q. V., et al. (2020). "Questioning the AI: Informing Design of Explanations." (for design implications) - Kocielnik, R., et al. (2019). "Towards a Design Space for Explainable AI Interfaces." (for interface patterns) You can narrow this topic by focusing on a domain (e.g., healthcare assistant), modality (voice-only), or user group (older adults) for a more manageable scope.