Beyond process model complexity: a multi-granular investigation across time and space based on eye-tracking

Sorg, Thierry; Abbad-Andaloussi, Amine; Kindler, Ekkart; Weber, Barbara · 2026 · Crossref

DOI: 10.1007/s44311-025-00035-3

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the relationship between process model complexity and the cognitive load experienced by users, addressing a gap in Business Process Management (BPM) where existing metrics’ ability to predict mental effort remains unclear. The authors distinguish between essential complexity (inherent logic) and accidental complexity (layout and presentation), hypothesizing that both influence intrinsic and extraneous cognitive load, respectively. The research aims to determine if complexity metrics align with multimodal cognitive load indicators and how task characteristics and visual behavior dynamics affect this alignment. The researchers conducted a controlled eye-tracking experiment with 27 participants. The study utilized process models that systematically varied in essential and accidental complexity, including advanced models that embedded both simple and complex regions within a single diagram. Task complexity was explicitly manipulated so that comprehension tasks targeted regions of differing complexity. The experimental design employed a multi-granular analysis: a coarse-grained level assessed the overall relationship between model/task complexity and cognitive load, while a fine-grained level segmented eye-tracking data spatially (task-relevant vs. non-task-relevant regions) and temporally (discovery vs. post-discovery phases). Cognitive load was measured using subjective self-reports, performance metrics (accuracy and time), physiological indicators (pupil activity via the Low/High Index of Pupil Activity), and behavioral measures (fixations and Areas of Interest run counts). The results demonstrate that complexity metrics align with cognitive load when both essential and accidental complexities are comprehensively captured. Task characteristics significantly influence this relationship, confirming that the specific region of the model targeted by a task affects perceived difficulty. Fine-grained analysis revealed distinct phases in visual behavior during comprehension, specifically a "discovery" phase followed by a "post-discovery" phase. Crucially, the study found that non-task-relevant regions of process models contribute to cognitive load in a non-uniform manner, particularly during the early stages of a task. This indicates that irrelevant complex parts of a model impose extraneous cognitive load even when they are not directly required for the task at hand. These findings strengthen measurement practices in BPM by establishing an empirical correspondence between comprehensive metric suites and multimodal cognitive load indicators. The study validates the utility of separating essential and accidental complexity in metric design. Furthermore, the identification of temporal phases and the impact of non-task-relevant regions suggests that adaptive modeling tools could improve user experience by helping focus attention on task-relevant parts of process models, thereby reducing extraneous cognitive load. This work extends prior research by integrating fine-grained eye-tracking analysis with complexity metrics, offering a deeper understanding of the temporal and spatial dynamics of model comprehension.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• mental demand