Profiling cognitive offloading in LLM-mediated synthesis writing: Volume vs. content

2026-06-09

AI Summary

This study investigates cognitive offloading patterns in large language model (LLM)-assisted synthesizing writing and their impact on learning. Grounded in distributed cognition and text comprehension theories, offloading behaviors are characterized along two dimensions—usage frequency and content—and analyzed in conjunction with learners’ prior knowledge through clustering. Data from 97 undergraduate students interacting with a general-purpose LLM via a custom interface were collected, with prompts coded for agency (active/passive) and level of understanding (local/global). K-means clustering identified four usage-based and five content-based offloading profiles. For the first time, this research distinguishes and compares these two typologies, revealing that content-based patterns more effectively capture qualitative differences in cognitive engagement and are significantly associated with learning strategies and textual originality, thereby highlighting their critical role in elucidating the mechanisms underlying LLM-supported learning.

Abstract

This study compares two approaches to profiling how learners offload cognitive activity to LLMs during a synthesis writing task. Drawing on Salomon's distributed cognition and the Kintsch and van Dijk model of text comprehension, the study operationalises offloading to an LLM in two ways: as a volume of LLM use and as content of what is offloaded, both along with prior knowledge. Data from 97 university students interacting with a general-purpose LLM via a custom interface were analysed using k-means clustering. To capture the content of offloading, their prompts were interpreted as to who performs the activity (active or passive) and at what level of comprehension (local or global). Volume-based profiling (k=4) differentiated learners primarily by prior knowledge, with volume negatively associated with essay authorship. Content-based profiling (k=5) revealed qualitatively distinct patterns of offloading, from vocabulary clarification to active direction of structuring and generation to passive delegation of comprehension at both levels. These patterns reflect different fragmentation of the cognitive process, with differences in learning strategies, behavioural markers, and essay authorship. Combining volume and content of offloading could improve future analyses on how LLM use redistributes cognitive activity and its effects on learners.

Problem

• cognitive offloading
• LLM-mediated writing
• synthesis writing
• distributed cognition
• prior knowledge

Innovation

• cognitive offloading
• LLM-mediated writing
• content-based profiling
• distributed cognition
• k-means clustering