Cognitive control capacity as a selective capacity constraint across cognitive control processes: a multimethod examination using network analysis and LASSO regression.

Cognitive control is characterized by severe capacity limitations, yet it remains unclear whether performance constraints across different cognitive control processes reflect independent limitations or shared capacity-related constraints. To investigate this question, we adopted an individual-differences approach and multimethod analyses to examine how cognitive control capacity (CCC), defined as the maximum rate at which the cognitive control system can successfully process task-relevant information, is associated with performance across multiple cognitive-control processes. A sample of 197 healthy adults completed a modified backward-masking majority function task (MFTM-A) that adaptively estimated each participant's exposure-time threshold at a fixed information load, from which CCC was derived (bits per second). These participants also completed a multi-component battery assessing domain-specific functions including working memory (N-back), inhibition (Go/No-Go, Stop-Signal), flexibility (task switching), proactive/reactive control (AX-CPT) and attentional functions (alerting, orienting and conflict). Results showed that CCC was significantly associated with several cognitive-control measures: higher CCC was associated with greater N-back accuracy, better AX-CPT discriminability, higher No-Go accuracy, and reduced switch costs. Network analysis indicated that proactive control and response inhibition occupied the most central positions in the network, whereas CCC showed selective connections with high-load control measures. A least absolute shrinkage and selection operator (LASSO) regression identified 3-back accuracy as the most robust unique predictor of CCC. These findings suggest that CCC reflects a selective throughput-related constraint that is closely tied to high-load working memory updating and extends to selected control processes requiring sustained and flexible information processing. Methodologically, this study illustrates how network analysis and sparsity-based modelling can be combined to clarify both the organization of cognitive-control measures and the behavioral predictors of CCC.