Effects of Uncertainty and Cognitive Load on User Trust in Predictive Decision Making

  title={Effects of Uncertainty and Cognitive Load on User Trust in Predictive Decision Making},
  author={Jianlong Zhou and Syed Arshad and Simon Luo and Fang Chen},
Rapid increase of data in different fields has been resulting in wide applications of Machine Learning (ML) based intelligent systems in predictive decision making scenarios. [] Key Result Presentation of uncertainty under high load conditions (when cognitive resources were short in supply) leads to a decrease of trust in the system and its recommendations.
When Does Uncertainty Matter?: Understanding the Impact of Predictive Uncertainty in ML Assisted Decision Making
This work carries out user studies to systematically assess how people respond to different types of predictive uncertainty i.e., posterior predictive distributions with different shapes and variances, in the context of ML assisted decision making, and demonstrates that uncertainty is an effective tool for persuading humans to agree with model predictions.
Should I Follow this Model? The Effect of Uncertainty Visualization on the Acceptance of Time Series Forecasts
Time series forecasts are ubiquitous, ranging from daily weather forecasts to projections of pandemics such as COVID-19. Communicating the uncertainty associated with such forecasts is important,
Updates in Human-AI Teams: Understanding and Addressing the Performance/Compatibility Tradeoff
It is shown that updates that increase AI performance may actually hurt team performance, and a re-training objective is proposed to improve the compatibility of an update by penalizing new errors.
Trust in Prediction Models: a Mixed-Methods Pilot Study on the Impact of Domain Expertise
The results underline the multi-faceted nature of trust, and suggest that domain expertise alone cannot fully predict people’s trust perceptions.
Assessing the communication gap between AI models and healthcare professionals: explainability, utility and trust in AI-driven clinical decision-making
This paper contributes with a pragmatic evaluation framework for explainable Machine Learning (ML) models for clinical decision support. The study revealed a more nuanced role for ML explanation
Algorithmic versus Human Advice: Does Presenting Prediction Performance Matter for Algorithm Appreciation?
A theoretical model based on the judge-advisor system (JAS) is proposed and empirically examined how algorithmic advice, compared to identical advice from humans, influences human judgment; it is shown that individuals largely exhibit algorithm appreciation.
A Case for Backward Compatibility for Human-AI Teams
The notion of the compatibility of an AI update with prior user experience is introduced and methods for studying the role of compatibility in human-AI teams are presented.
Towards Trustworthy Human-AI Teaming under Uncertainty
A novel framework of human and AI teaming for trustworthy AI under uncertainty for trustworthy decisions under uncertainty is proposed to incorporate uncertainty and performance into AI-advised human decision making to boost user trust in AI.
Why Am I Not Seeing It? Understanding Users’ Needs for Counterfactual Explanations in Everyday Recommendations
This paper investigates whether and when to provide counterfactual explanations to support people’s decision-making with everyday recommendations through a question-driven approach and reveals that the utility of decision is a primary factor that may affect theircounterfactual information needs.
A User Study on the Effect of Aggregating Explanations for Interpreting Machine Learning Models
This work-in-progress paper explores the effectiveness of providing instance-level explanations in aggregate, by demonstrating that such aggregated explanations have a significant impact on users’ ability to detect biases in data.


Investigating User Confidence for Uncertainty Presentation in Predictive Decision Making
It is found that all 26 subjects rated higher uncertainty task to be most difficult and had lowest user confidence in predictive decisions of the same, which agrees with the expectation that increased uncertainty would reduce userconfidence in predictive decision making.
Decision Making for Risk Management: A Comparison of Graphical Methods for Presenting Quantitative Uncertainty
Two experiments examined the ability of a comprehensive set of commonly used probabilistic display methods to effectively communicate critical information to a decision maker and influence confidence in decision making and investigated the performance of these methods under time pressure.
The Effect of Cognitive Load on Decision Making with Graphically Displayed Uncertainty Information
Overall, the research suggests that interpreting basic characteristics of uncertainty data is unharmed under conditions of limited cognitive resources, whereas more deliberative processing is negatively affected.
Effects of Information Availability on Command-and-Control Decision Making
The results indicate that increasing the volume of information, even when it is accurate and task relevant, is not necessarily beneficial to decision-making performance and may even be detrimental to SA and trust among team members.
Trust in Automation: Designing for Appropriate Reliance
This review considers trust from the organizational, sociological, interpersonal, psychological, and neurological perspectives, and considers how the context, automation characteristics, and cognitive processes affect the appropriateness of trust.
Making machine learning useable
This paper investigates making ML useable from the point of view of how human-computer interaction HCI techniques benefit ML in order to simplify the interface between users and ML algorithms.
Trust in Automation
A three-layered trust model provides a new lens for conceptualizing the variability of trust in automation and can be applied to help guide future research and develop training interventions and design procedures that encourage appropriate trust.
Decision Making Under Uncertain and Risky Situations
Decision Making is certainly the most important task of a manager and it is often a very difficult one. The domain of decision analysis models falls between two extreme cases. This depends upon the
The Effect of Cognitive Load on Economic Decision Making