[Paper Review] Trust and Cognitive Load During Human-Robot Interaction

Muneeb Ahmad, Jasmin Bernotat+2 more|arXiv (Cornell University)|Sep 11, 2019

Social Robot Interaction and HRIPsychology34 references21 citations

TL;DR

This study investigates the interplay between cognitive load, trust, and anthropomorphism in human-robot interaction using a collaborative 'Matching the Pairs' game with Husky (non-humanoid) and Pepper (humanoid) robots. Results show an inverse relationship between cognitive load and trust, with a triple interaction where trust preferences shift based on robot type and error rate—Pepper was trusted more under high error rates, while Husky was trusted more under low error rates.

ABSTRACT

This paper presents an exploratory study to understand the relationship between a humans' cognitive load, trust, and anthropomorphism during human-robot interaction. To understand the relationship, we created a \say{Matching the Pair} game that participants could play collaboratively with one of two robot types, Husky or Pepper. The goal was to understand if humans would trust the robot as a teammate while being in the game-playing situation that demanded a high level of cognitive load. Using a humanoid vs. a technical robot, we also investigated the impact of physical anthropomorphism and we furthermore tested the impact of robot error rate on subsequent judgments and behavior. Our results showed that there was an inversely proportional relationship between trust and cognitive load, suggesting that as the amount of cognitive load increased in the participants, their ratings of trust decreased. We also found a triple interaction impact between robot-type, error-rate and participant's ratings of trust. We found that participants perceived Pepper to be more trustworthy in comparison with the Husky robot after playing the game with both robots under high error-rate condition. On the contrary, Husky was perceived as more trustworthy than Pepper when it was depicted as featuring a low error-rate. Our results are interesting and call further investigation of the impact of physical anthropomorphism in combination with variable error-rates of the robot.

Motivation & Objective

• To investigate the relationship between cognitive load and trust in human-robot interaction (HRI) under high-cognitive-load task conditions.
• To examine how anthropomorphism—via humanoid (Pepper) versus non-humanoid (Husky) robot designs—affects trust perceptions during collaboration.
• To assess the impact of robot error rate on trust ratings and behavioral responses in collaborative task settings.
• To explore whether collaborative gameplay can enhance trust and reduce perceived cognitive load in HRI.
• To identify moderating effects of robot type and reliability on trust dynamics under varying cognitive demands.

Proposed method

• Designed a 'Matching the Pairs' game requiring memory and coordination, simulating a high-cognitive-load collaborative task.
• Used two distinct robot types: Husky (non-humanoid, technical robot) and Pepper (humanoid, socially expressive robot) to manipulate anthropomorphism.
• Induced cognitive load by requiring participants to memorize card positions and coordinate selections with the robot.
• Measured cognitive load via pupil dilation using an online peak detection system, validated under controlled lighting.
• Collected pre- and post-interaction trust ratings using standardized questionnaires to assess changes in trust perception.
• Systematically varied robot error rates (low vs. high) to evaluate their impact on trust and perceived reliability.

Experimental results

Research questions

• RQ1 How does cognitive load affect trust in human-robot interaction during a collaborative task?
• RQ2 Does anthropomorphism (via humanoid vs. non-humanoid robot design) influence participants' trust in a collaborative setting?
• RQ3 How does robot error rate moderate trust perceptions in human-robot teams?
• RQ4 Is there a significant change in trust ratings after interacting with a robot in a collaborative game?
• RQ5 What is the interaction effect between robot type, error rate, and trust perception in HRI?

Key findings

• There was a significant inverse relationship between cognitive load and trust, such that higher cognitive load led to lower trust ratings.
• Participants trusted the Husky robot more than Pepper when the robot had a low error rate, indicating higher cognitive trust in reliable non-humanoid robots.
• Participants trusted the Pepper robot more than Husky when the robot had a high error rate, suggesting that anthropomorphic robots may be perceived as more forgiving or resilient under failure.
• A triple interaction effect was found between robot type, error rate, and trust ratings, indicating that trust is not solely determined by anthropomorphism or reliability alone.
• Trust ratings increased after HRI, suggesting that interaction itself may enhance trust, though this effect was not significantly different between robot types.
• Individual differences in interaction style were observed, with some participants relying heavily on the robot, others using it as a consultant, and others ignoring it entirely, highlighting the need for personalized HRI design.