Preprint · EdArXiv 2026 · Under journal review
The Absent Cognitive Baseline: Theorizing a Structural Gap in AI‑Native College Students' Academic Self‑Assessment
Yizhen Jia & Jiarui Xu · College of Education, Westcliff University
01 · Abstract
Students entering college in the 2026–2027 academic year are among the first cohorts to complete most of their secondary schooling under pervasive generative AI availability. Generative AI can substitute core cognitive operations involved in academic work, including reasoning, synthesis, evaluation, and articulation. This raises questions about what unassisted cognitive experience these learners accumulate, and what internal reference points they draw on when assessing their own academic competence.
This paper proposes the Absent Cognitive Baseline (ACB) as a conceptual framework to name this condition. Drawing on metacognitive theory, self-regulated learning research, epistemic development scholarship, and post-2022 empirical research, the paper argues that academic self-assessment depends on experiential reference points formed during adolescence, that these points are often unevenly developed even under traditional conditions, and that substitutive AI use during the formative high-school window may reduce the independent cognitive encounters on which they depend.
A conceptual model identifies three groups of moderating variables shaping whether ACB emerges. The central argument is that generative AI may change not only what students can do academically, but the conditions under which they come to know what they can do. Interpretative phenomenological analysis is proposed as a methodologically coherent path into the subjective layer the framework theorizes.
02 · The Three Dimensions
Three ways the gap shows up
- IUnknowabilityNot being able to tell where your own understanding ends, because there is no memory of doing this kind of work without help.
- IIFalse calibrationFeeling that you understand because the AI's answer reads so smoothly.
- IIIDe-normalization of struggleLosing the sense that getting stuck is a normal part of learning.
03 · Model & Method
What decides whether it happens
The paper says three things shape whether this gap opens: how the AI was used, who was using it, and whether the setting pushed students to let the tool do the work. To study it, the paper proposes close, one-on-one interviews with a small number of students about their own experience, a method called interpretative phenomenological analysis.
04 · Read & Cite
Preprint v5, live on EdArXiv
→ Read on EdArXiv05 · Related
The sibling construct
ACB describes the case where a student has no way to judge their own understanding at all. Its sister paper, Metacognitively Discordant Completion, describes the case where the student does judge, knows they do not understand, and hands the work in anyway.
→ MDC paper