Curriculum vitae · August 2026Trust, mediated interaction, and experiments
Hao Ling
I study what happens to the conditions for trust when economic interaction moves onto a new medium.
My training in behavioral and experimental economics lets me compare physical, virtual-reality, and AI-mediated settings with incentive-compatible designs and causal identification.
Two 2×2 factorial studies test how explainability and anthropomorphism jointly shape trust in delegated decisions: an online experiment (N=900) and a pre-registered eye-tracking lab study (N=57). In the low-engagement online setting, adding anthropomorphism to an explainable agent reduced trust; in the high-engagement lab setting the same combination was complementary. Eye tracking shows the mechanism — explanation acts as a cognitive activator that redirects gaze toward the avatar (icon dwell 1.22% → 15.99%) — so trust depends on cognitive engagement moderating how social cues are processed.
AI delegationexplainabilityanthropomorphismeye tracking
A three-group design compares an avatar-free VR baseline, brief exposure to a standardized counterpart avatar, and a physical decision condition (N=180). The baseline and physical conditions are practically equivalent for trust under a stated smallest effect of interest; equivalence for trustworthiness is not established. Brief avatar exposure is associated with a positive but imprecisely estimated change in trust. The study supports a bounded methodological use of high-immersion VR, not general superiority over conventional laboratory experiments.
Pre-registered at the AEA RCT Registry as AEARCTR-0013935 (28 Aug 2024). The results above report Stage 1 of the registered design.
high-immersion VRBerg Trust Gamehuman presencepre-registered
The three-condition design distinguishes interactive, embodied VR from passive video and separates that effect from added emotional intensification. Both VR conditions produced greater empathy gains than passive video, while Emotional VR added no measurable empathy gain over Non-Emotional VR. Non-Emotional VR yielded a higher proportion of auditory-focused solutions and the strongest user-need alignment, without greater measurable empathy: emotional intensity is not a reliable proxy for downstream user-need alignment.
Virtual Realityempathycognitive loadinclusive design
A developing framework for understanding and evaluating VR as a scientific instrument.
VR as instrumentresearch design
A pilot study on how multi-user virtual reality may shape different levels of learning in economics education.
multi-user VReconomics education
No. 2025A1515010549 · Experimental Investigation of Human Trust and Cooperation in VR
Role: Key Student Researcher — experimental design, VR development, data analysis. Output: one accepted CHI paper and one working paper.
No. HKUST(GZ)-ROP2025040 · featured in the HKUST(GZ) Opening Ceremony President's Lecture, 2025
Role: Key Student Researcher — VR experiment development, course design, data analysis. Output: one working paper.