Measuring Real-Time Cognitive Load and Metabolic Strain as a Function of Communication Signal Ambiguity

Abstract

This document presents a turnkey experimental protocol designed to quantify real-time cognitive load and autonomic strain as a function of communication signal ambiguity. Moving beyond retrospective self-report questionnaires, the protocol operationalizes social communication through an information-theoretic framework. The paradigm evaluates a crossover interaction between autistic adults and neurotypical controls across two tightly controlled conditions: deterministic, high-fidelity logical propositions versus lossy, under-constrained natural language containing dropped antecedents and ambiguous syntax. The protocol details full hardware specifications and event-locked synchronization (TTL/LSL) across a multimodal physiological recording stack, including task-evoked pupillometry (locus coeruleus activity), prefrontal functional near-infrared spectroscopy (fNIRS hemodynamic expenditure), electrodermal activity (phasic GSR), and cardiac heart rate variability (parasympathetic vagal tone). Analytical models are specified using linear mixed-effects models (LMM) and multinomial logistic regression. By isolating signal fidelity as an independent variable, this paradigm enables laboratories to directly measure the metabolic cost of resolving ambiguous input, reframing communication breakdown as an engineering mismatch between high-precision cognitive architectures and lossy signals.