The Symbiosis of MathPsych, ICCM, and Human Factors Psychology

My talk will describe my approach to cognitive computational modeling for human factors applications. This is quite distinct from a more frequent approach, based upon discrete event simulations of the cognitive processing, like ACT-R. In contrast, I focus on computing the performance product, as yielded by simple linear equations. These two approaches are complementary. Generically, performance (P) can be predicted by [P = aA + bB = cC …..] where A,B,C… are different cognitive or environmental influences and a,b,c… are different weights. The influences are derived from cognitive task analysis, and the weights are typically derived from meta-analyses of effect sizes of those influences. Validation is accomplished by correlating predictions with obtained behavior across a set of conditions. I will describe this approach in the context of models I and my colleagues have developed for five models of different aspects of attention: of visual attention, the supervisory scanning (SEEV) model of display layout, the change blindness and noticing (NSEEV) model, and the scan-clutter tradeoff model of display design, of divided attention or multi-tasking, the multiple resource model of concurrent time-sharing, and the STOM model of sequential multi- tasking or task switching. I will conclude by presenting the framework of such a model for influences on human-automation interaction.

This is an in-person presentation on July 18, 2026 (09:00 ~ 09:20 EDT).

Promising research is often stranded in the ‘valley of death’, where efficacy is lost before solutions can be implemented in complex operational environments. This symposium presentation attempts to bridge the gap from research to practice by demonstrating how traditional models of human performance can be leveraged to understand performance degraders and make predictions on future performance in naval operations. I present three use-cases that leverage the intersection of mathematical psychology and human factors: 1) quantifying the impacts of minimal manning on watchstander workload, (2) identifying key factors that affect risk perception on the flight deck, and (3) evaluate how sailors adopt, adapt, and integrate new technology. Successfully translating research from the lab to the deckplate requires overcoming significant hurdles. I conclude by discussing these challenges including the difficulty of defining operational performance metrics, accounting for complex socio-environmental influences, and navigating the slow and incremental changes inherent in long-term capability development.

This is an in-person presentation on July 18, 2026 (09:20 ~ 09:40 EDT).

The objective of this research is to develop an interface-based, human-in-the-loop monitoring mechanism that supports AI-assisted situational awareness by estimating a real-time metric of deadline-compliance risk in highly constrained procedural tasks. Risk is operationalized as one-sided lag, where predicted progress exceeds observed progress for the currently required benchmark obligations. Participants complete time-structured sections composed of distinct perceptual-cognitive modality-path subsections motivated by Multiple Resource Theory (Wickens, 2008) and multitasking-performance modeling (Fox, Houpt, & Tsang, 2021), while avoiding any a priori assumption that processing architecture follows modality boundaries. Instead, subsection and alert–subsection interaction architecture are inferred during calibration using Systems Factorial Technology diagnostics under calibration conditions intended to isolate subprocesses and satisfy selective influence constraints. Calibration also yields capacity functions associated with each task category. Architecture labels and capacity functions are then held fixed for the metric testing condition.The interface constrains task-relevant information to cursor-contingent display regions, enabling high-resolution benchmarking via cursor-region timestamps and discrete completion events. Observed progress is defined as the fraction of completed obligations among the currently required set, yielding a stepwise trajectory with denominator changes when alerts insert new obligations. Predicted progress is defined as the expected fraction completed based on calibration-conditioned event-time models, incorporating capacity through an accelerated failure time (AFT) time-warp. Real-time risk is computed using a one-sided lag discrepancy that accumulates only when predicted progress exceeds observed progress, supporting online detection of elevated deadline-miss risk under a hard deadline.

This is an in-person presentation on July 18, 2026 (09:40 ~ 10:00 EDT).

Many mathematical and computational frameworks are developed with an aim of generality or flexibility in order to enable their application over a range of human research domains. These types of efforts can open doors to integration with highly applied scientific fields such as human factors engineering which stand to benefit greatly from the advances. The combining of approaches in mathematical and computational psychology to the various realms of contemporary and practical application however can come with complexities that make for unique and interesting challenges. Medical devices and surgical settings for example involve the complex and dynamic environments of surgical operating rooms (ORs), where performance depends not only on individual cognition but on multi-faceted interactions among diverse team roles and an atmosphere of multiple and ever-changing factors. This presentation highlights such applications exploring how quantitative approaches can provide meaningful insights, underscoring the value and ongoing challenge of leveraging mathematical models to address critical questions in medical safety, performance, and innovation. It will discuss the unique challenges faced when bridging abstract models with real-world complexity, including the variability in human behavior, team dynamics, and contextual influences inherent in healthcare environments.

This is an in-person presentation on July 18, 2026 (10:00 ~ 10:20 EDT).