Evaluating Cognitive Processes During Decision-Making Tasks using Mobile Phones, Computer Mice, and Keyboards: Theory, Method, and Application

Abstract

Trace Data, characterized by fine-grained metadata collected at millisecond precision, has become increasingly popular in Human-Computer Interaction studies to investigate various cognitive states of individuals. However, a significant challenge these studies face is the considerable variation in their theoretical foundation, leading to a diverse array of disconnected empirical findings. This dissertation proposes a unified theoretical framework, Decision Evaluation with Cognitive Observations using Device Engagement (DECODE), grounded in Dual Process Theory, a widely accepted theory that posits that human's thinking is an interplay between intuitive and deliberate processes and Cognitive Control. This widely accepted construct encompasses attention regulation, decision-making, and goal-directed behavior. The DECODE framework proposes that the Cognitive Task Requirements (CTR), which signifies the cognitive demand imposed on an individual by a particular task, and task-specific expertise, influence Cognitive Control (CC) and individuals' engagement in intuitive and deliberate thinking. This dissertation empirically validates that the differences in behaviors proposed by the DECODE framework can be captured and analyzed with trace data from various Human-Computer Interaction devices, including computer mice, keyboards, and mobile devices. 'Essay One' examines how CTR and individual task-specific expertise influence mouse movement speed. 'Essay Two' further investigates the role of CTR and an individual's task-specific expertise in typing fluency using the keyboard. 'Essay Three' extends previous studies' findings by demonstrating how removing CTR speeds up decision-making using mobile devices. The overall results from all three essays suggest that increasing CTR significantly slows the device usage speed while developing task-specific expertise increases the device usage speed.