Dissociable contributions of goal-relevant evidence and goal-irrelevant familiarity to individual and developmental differences in conflict recognition

Weigard, Alexander; Suzuki, Takakuni; Skalaban, Lena J.; Conley, May I.; Cohen, Alexandra O.; Garavan, Hugh; Heitzeg, Mary M.; Casey, B.J.; Sripada, Chandra; Heathcote, Andrew · 2023 · Unknown

DOI: 10.31234/osf.io/nvh98

archive: archived pipeline: cataloged

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

Summary

This paper addresses a discrepancy in cognitive control research where the diffusion decision model (DDM) typically identifies a general "efficiency of evidence accumulation" (EEA) factor driving performance across tasks, yet EEA estimates from the n-back "conflict recognition" paradigm in the Adolescent Brain Cognitive Development (ABCD) Study showed inconsistent results. Specifically, drift rates from "lure" trials (familiar but non-target stimuli) exhibited atypical v-shaped bivariate distributions, suggesting that many participants relied on goal-irrelevant stimulus familiarity rather than goal-relevant evidence. The authors propose a new formal model of evidence integration that dissociates EEA for goal-relevant evidence from the use of goal-irrelevant familiarity. The study utilized baseline and year-2 n-back data from 8,824 and 6,539 ABCD participants, respectively, along with an independent age-diverse sample of 150 individuals aged 8–30. The authors developed a modified DDM that decomposes drift rates into three parameters: efficiency of goal-relevant evidence accumulation ($e$), strength of goal-irrelevant familiarity evidence ($f_s$), and a familiarity evidence criterion shift ($f_c$). This model was estimated using Bayesian methods with informative priors derived from an independent subsample. The authors then examined the developmental trajectories, test-retest reliability, and associations of these parameters with general cognitive ability, working memory, and clinical measures of attention problems. Results indicated that both EEA ($e$) and familiarity strength ($f_s$) were task-general dimensions, showing strong correlations across different levels of n-back load. However, only EEA demonstrated large, robust developmental differences in both the ABCD sample and the independent age-diverse sample. EEA also exhibited higher test-retest reliability and meaningful associations with clinically relevant dimensions, including general cognitive ability and parent/teacher-reported attention problems. In contrast, the familiarity-based parameters did not show the same robust developmental or clinical correlates. The v-shaped distributions observed in standard DDM analyses were explained by the mixture of individuals relying on familiarity versus goal-relevant evidence. The significance of this work lies in establishing a principled modeling framework for characterizing conflict recognition mechanisms. By distinguishing between goal-relevant and goal-irrelevant evidence, the study clarifies why standard factor analyses of n-back performance can be misleading. It highlights that individual differences in cognitive control are not solely driven by a single general efficiency factor but are also influenced by strategic reliance on familiarity. These findings have broader implications for understanding developmental trajectories of cognitive control and identifying specific cognitive deficits in psychiatric conditions such as ADHD, where inattention may manifest as an over-reliance on familiarity rather than a lack of goal-directed evidence accumulation.

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. Discovered via author_sweep_intake on 2026-05-28.

Summary generated by qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

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

• cognitive capacity variation