How do users mentally model dialogue agent partners?
Exploring what dimensions matter when people form impressions of machine dialogue partners—and whether competence, human-likeness, and flexibility all play equal roles in shaping user expectations and behavior.
The Partner Modelling Questionnaire (PMQ) validates a three-factor structure for how users perceive machine dialogue partners. The concept originates in psycholinguistics: people form mental representations of their dialogue partner's communicative and social capabilities, and these representations guide what they say, how they say it, and what tasks they entrust to that partner.
Factor 1 — Communicative competence and dependability (49% variance, α=0.88): Strongest items: competent/incompetent, dependable/unreliable, capable/incapable. This is the largest factor — nearly half the variance in how users model a dialogue agent is about whether it can do the job reliably.
Factor 2 — Human-likeness in communication (32% variance, α=0.80): Strongest items: human-like/machine-like, life-like/tool-like, warm/cold. Humans act as the archetype for evaluating communication partners. Even when using machines, people evaluate against a human standard.
Factor 3 — Communicative flexibility (19% variance, α=0.72): Items: flexible/inflexible, interactive/stop-start, interpretive/literal, spontaneous/predetermined. This factor captures whether the agent feels like a living conversation or a scripted interaction.
The definition of partner models — "an interlocutor's cognitive representation of beliefs about their dialogue partner's communicative ability, multidimensional, initially informed by experience and stereotypes, dynamically updated during dialogue" — positions this as the HCI equivalent of theory of mind. Users build, maintain, and update these models continuously.
The practical implication: designing for perceived competence matters most (49% of variance), but human-likeness and flexibility are not negligible. An agent that is reliable but inflexible and machine-like will be perceived very differently from one that is reliable, warm, and spontaneous.
Inquiring lines that use this note as a source 66
This note is a source for these synthesized inquiries. Follow a line forward into its question, or open it to trace back to all of its sources.
- What cognitive capabilities do agents need to internalize social feedback?
- How does outcome feedback change beliefs about AI versus human partner reliability?
- What interpretive work must humans perform to experience AI as a conversation partner?
- What social patterns from human training data activate in agent context?
- What role does user interface framing play in consciousness perception?
- What would co-constructed identity between human and model dialogue look like?
- What makes synthetic user data transfer to real conversational systems?
- How does non-human origin of personas affect team willingness to critique them?
- Can structured empathy measurement frameworks predict persona effectiveness?
- Why does human interaction remain the hardest failure mode for agents?
- Which alignment dimensions matter most in educational conversation design?
- What latent dimensions matter most for content creators?
- How does user overreliance on model confidence differ between chat and deployed agents?
- What individual differences predict who benefits from AI partnership?
- How do user expectations change as chatbots remember more interactions?
- How does the expectation ratchet affect long-term chatbot satisfaction?
- Why do conventional mental models fail when applied to AI interaction?
- Does conversational AI personalization increase behavioral expectations too much?
- How do humans learn to prefer AI partners over humans?
- How does intrinsic motivation drive conversational agents beyond passive responsiveness?
- Why do passive conversational agents fail at collaborative decision-making?
- How does theory of mind predict success in human-AI partnerships?
- Do dialogue agents have authentic voice agency or beliefs of their own?
- Do people treat conversational AI as social actors without conscious awareness?
- Does social presence from robots drive adherence better than conversational AI interfaces?
- Can users accurately recall their role versus the system's role in production?
- Can models optimized for solo capability support productive human collaboration?
- Can agent social framing change how humans apply collaborative social scripts?
- Do people with lower cognitive complexity prefer simpler machine communication goals?
- Do different levels of machine agency activate different interaction scripts?
- Why might media-specific scripts actually work better than human conversation mimicry?
- How do users develop different interaction scripts specifically for machines versus humans?
- Can users detect and correct an AI's mental model of their preferences?
- What competitive advantages does the ENFJ default create in human-AI interactions?
- Does perceived machine competence matter more than warmth in dialogue?
- Can dialogue agents be reliable but still feel inflexible or cold?
- How do users update their partner models during ongoing conversation?
- Can AI systems infer user personality without knowing the interaction context?
- What makes proactive conversational agents feel intrusive versus helpful to users?
- What social boundaries must proactive agents respect during conversation?
- How can agents learn to estimate user satisfaction in real-time during conversation?
- How do conversational agents overcome structural passivity and goal awareness gaps?
- Does proactive agent design improve conversation efficiency or create user frustration?
- What role does uncertainty reduction play in personalized agent interaction?
- Can general chatbot skill predict how well models roleplay adversarial personas?
- How can dialogue structure and trajectory predict social agent performance?
- Does persona-level grouping systematically trigger confidence-misdirection failures in practice?
- Why do embodied agents outperform text chatbots in therapy outcomes?
- What distinguishes communicative competence from human-like dialogue ability?
- How do contextual characteristics like emotional state shape dialogue authenticity?
- Can advertising mechanisms designed for humans work on agents?
- How do expectation-management metrics differ from traditional conversational quality metrics?
- Do embodied agents outperform chatbots because of physical presence alone?
- What ecosystem conditions beyond technical capability determine whether users adopt AI features?
- How do casual conversational styles make AI seem more human?
- What happens to user expectations as AI conversation quality improves?
- Can users adapt their competencies to match how AI actually operates?
- Can role-aligned AI systems replicate an expert's sense of audience and moment?
- How do interpersonal skills reshape task importance as automation increases?
- What makes some agent benchmarks measure interaction quality better than others?
- What are the key interaction mechanisms that make human-agent collaboration work?
- How do personalization systems reshape expectations in AI relationships?
- What behavioral signals let users detect communicative flexibility in AI?
- Why do people prefer AI partners over humans once identity is disclosed?
- How do users misattribute social competence to language models in assistant roles?
- Should AI assistants align with role-specific norms rather than user preferences?
Related concepts in this collection 2
This note in its neighbourhood — explore the map, then jump to a related concept in the list below.
Click a node to walk · click center to open · click Open in graph to see this note in the full knowledge graph
-
What breaks when humans and AI models misunderstand each other?
Explores whether misalignment in mutual theory of mind between humans and AI creates only communication problems or produces material consequences in autonomous action and collaboration.
MToM framework operates at the same level as partner models but adds the AI-side modeling
-
Can AI-generated personas build genuine empathy in product teams?
This study explored whether prompt-engineered personas created in minutes could foster the same emotional and behavioral empathy as traditional user research. The findings reveal a surprising gap between understanding users and caring about their needs.
partner models may explain why cognitive empathy works: users perceive competence (Factor 1) but not warmth (Factor 2)
Related papers in this collection 8
Papers most semantically related to this note, ranked by cosine similarity in the embedding space.
- The Partner Modelling Questionnaire: A validated self-report measure of perceptions toward machines as dialogue partners
- Linguistic Alignment in Conversational AI: A Systematic Review of Cognitive-Linguistic Dimensions, Measurements, and User Outcomes (2020–2025)
- CloChat: Understanding How People Customize, Interact, and Experience Personas in Large Language Models
- UserBench: An Interactive Gym Environment for User-Centric Agents
- Through the Lens of Human-Human Collaboration: A Configurable Research Platform for Exploring Human-Agent Collaboration
- From speaking like a person to being personal: The effects of personalized, regular interactions with conversational agents
- PersLLM: A Personified Training Approach for Large Language Models
- WHEN TO ACT, WHEN TO WAIT: Modeling Structural Trajectories for Intent Triggerability in Task-Oriented Dialogue
Original note title
partner models for dialogue agents decompose into three factors — communicative competence human-likeness and communicative flexibility