Data Foraging with Human-Robot Teams

In the same way human and non-human animals forage for food, expert field scientists forage for data. Data foraging decisions require consideration of an uncertain information gradient, i.e., what areas have the highest data value, and what are the costs required to find and extract data? This line of research focuses on understanding how expert field scientists make data foraging decisions, and aims to use scientists’ spatial knowledge of the world to (1) better understand the basic principles of the human mind that evolved to search and explore the world, and (2) identify specific ways in which aerial and terrestrial robots can aid humans in spatial search problems.

We rely on field observations of earth scientists to study naturalistic data foraging, and have developed new web-based decision tasks inspired by our field experiences. In collaboration with earth scientists and robotics engineers, we are working to build robotic field assistants that can present data to human scientists in a manner that improves their processing fluency, while also providing decision guidance.

This research is currently funded by the NASA through the Planetary Science and Technology from Analog Research program.

Risky Decision Metascience

The science of scientific decision making under uncertainty

As humans, we must conduct science with a mind that is imperfectly adapted to optimal scientific practice. How is this good? How is this bad? How should scientific workflows change to accommodate this? How could the mind change? This line of research addresses these questions, and aims to use expert scientific decision making to improve science, while also illuminating how non-scientists ask questions, collect information, and reason about their environment.

We conduct cognitive-ethnographic work with Earth scientists in the field to study scientific decisions in highly uncertain conditions — where the isolation of geologic processes is challenging, because multiple processes have combined to transform the Earth's surface, and where direct observation is impossible due to the large time spans of processes that leave evidence eroded or buried. We have also developed web-based surveys and decision tasks to reveal the structure of how scientific decisions are made in experts, scientists-in-training, and novices.

Control Dynamics Underlying Decision Making

Good decision making hinges on a person's ability to strike an optimal balance between stability — maintenance of behavioral plans even in the face of distracting influences from the environment — and flexibility — adapting behavior and goals whenever significant changes in the environment or in internal states occur. One of the central challenges in cognitive psychology is determining factors that regulate the dynamic adaptive balance between these complementary control states. This line of research focuses on testing theories of dynamic cognitive control in complex and uncertain decision making.

We use individual and situational factors to test dynamic control theories in laboratory decision tasks, where task-relevant information must be acquired through experience and must be updated depending on changing events. The goal of this work is to better understand the influence of the flexibility-stability balance on decision making, and to improve dynamic control theories to better predict failures of decision making inside and outside the laboratory.