Research case · Applied psychology × UX 研究案例 · 应用心理学 × UX

Human-AI Interaction Design in High AI-Literacy Workplaces

高 AI 素养职场环境中的人机交互设计

This case translates a literature review into practical interaction principles for enterprise IM AI assistants, with a focus on the boundary conditions of anthropomorphic cues.

本案例将文献综述转译为企业 IM AI 助手可执行的交互原则，重点讨论拟人化线索的效用边界。

Role角色 Researcher / Product Designer研究者 / 产品设计师

Context场景 Enterprise IM + AI assistant企业 IM + AI 助手

Method方法 Literature synthesis文献整合

Status状态 Draft with reserved placeholders草稿版（含预留占位）

01. Research Question

01. 研究问题

In high AI-literacy teams, does stronger anthropomorphic design always improve human-AI collaboration, or can it reduce efficiency, autonomy, and judgement quality in complex work?

在高 AI 素养团队中，更强的拟人化设计是否一定提升人机协作？还是会在复杂任务中损害效率、自主感与判断质量？

Read more: why this question matters in enterprise IM展开：这个问题为什么对企业 IM 关键

In enterprise IM, interaction cadence is high and decision cycles are short. If anthropomorphic framing is overused, users may experience hidden social pressure, slower verification, and reduced confidence in high-stakes contexts.

企业 IM 的交互频率高、决策节奏快。若拟人化表达过强，用户在高风险任务中可能出现隐性社交压力、核验变慢与决策信心下降。

02. Key Findings (Condensed)

02. 关键结论（精简版）

In low-literacy contexts, moderate anthropomorphic cues can reduce anxiety and support onboarding.在低素养场景，适度拟人化可缓解焦虑并支持上手。
In high-literacy contexts, excessive social cues become cognitive noise and slow expert decision cycles.在高素养场景，过度社交线索会转化为认知噪声并拖慢专家决策循环。
Tool-like transparency, explicit control, and response predictability better match expert mental models.工具化的透明性、显式控制与可预期反馈，更匹配专家心智模型。

Read more: theoretical interpretation展开：理论解释

This pattern aligns with Cognitive Load Theory (extraneous load), Self-Determination Theory (autonomy), and Social Facilitation effects in complex tasks. As user expertise rises, low-noise interfaces and controllable parameters outperform social mimicry.

这一模式与认知负荷理论（外在负荷）、自我决定理论（自主感）及复杂任务中的社会促进/抑制效应一致。随着用户经验提升，低噪声界面与可控参数优于社交拟态。

03. Design Translation for IM AI Assistant

03. 面向 IM AI 助手的设计转译

Adaptive anthropomorphism: lighter social cues as user AI literacy rises.渐进式拟人化：随用户 AI 素养提升，逐步减弱社交线索。
Priority on control surfaces: structured parameters over vague conversational persuasion.优先控制界面：结构化参数优先于模糊说服式对话。
Professional boundary by default: assistant as instrument unless social framing is explicitly needed.默认专业边界：除非确有必要，否则助手保持“工具”定位。

Read more: design implications by layer展开：分层设计含义

Interaction layer: reduce affective fillers and keep response structure explicit. Control layer: expose key parameters and confidence cues. System layer: tune anthropomorphic intensity based on user literacy and task criticality.

交互层：减少情绪性填充语，保持响应结构显式。控制层：暴露关键参数与置信线索。系统层：根据用户素养与任务关键性动态调节拟人化强度。

04. Reserved Components (To Be Filled)

04. 预留模块（待补充）

Method Snapshot (Placeholder) 方法速览（占位）

TODO: sample scope, source selection criteria, synthesis method, and known limitations.

TODO：补充样本范围、文献筛选标准、整合方法与研究限制。

Quantitative Outcomes (Placeholder) 量化结果（占位）

TODO: task time, error rate, trust/autonomy metrics, and baseline comparison.

TODO：补充任务时长、错误率、信任/自主感指标及基线对比。

Validation Plan (Placeholder) 验证计划（占位）

TODO: expert review protocol, pilot test plan, and release criteria.

TODO：补充专家评审方案、试点验证计划与发布门槛。

Research Memo Download (Placeholder) 研究备忘下载（占位）

TODO: attach final PDF memo and citation appendix.

TODO：补充最终 PDF 研究备忘与参考文献附录。

Citation note (draft): Liu et al. (2025); Carrasco-Farré & Hakobjanyan (2026); plus CLT / SDT / Social Facilitation / Organizational Learning frameworks.

引用说明（草稿）：Liu 等（2025）、Carrasco-Farré & Hakobjanyan（2026），及认知负荷、自我决定、社会促进、组织学习等理论框架。

Read more: how to complete placeholders展开：占位内容如何补齐

Recommended order: (1) finalize method snapshot, (2) add at least one measurable outcome pair (before/after), (3) document validation criteria, (4) attach memo PDF and citation appendix for reviewer traceability.

建议补齐顺序：1）先完成方法速览；2）补至少一组可量化前后对比；3）明确验证门槛；4）附上研究备忘 PDF 与参考文献附录，便于评审追溯。