Cognitive Control Capacity as an Integrative Hub: A Multimethod Validation Through Network Analysis and Lasso Regression

Abstract: Cognitive control capacity (CCC) has been proposed as a compact index of how effectively individuals guide thought and behaviour when contingencies shift. To examine its validity, 198 adults completed a multi-component battery encompassing working-memory (N-back, AX-CPT), response inhibition (Go/No-Go, Stop-Signal), cognitive flexibility (task switching), and three attentional functions (altering, orienting and conflict). CCC was computed from entropy-time functions derived across majority function task. Higher CCC predicted greater N-back accuracy, fewer commission errors on inhibition trials, and reduced switch costs, but showed negligible ties to alerting and orienting scores. Network analysis situated the 3-back node at the graph centre with strong connections to inhibition and switching indices, whereas attentional nodes occupied peripheral positions. A least absolute shrinkage and selection operator (Lasso) regression prioritized 3-back accuracy as the significant predictor of CCC, with other subcomponent metrics (e.g., Go/No-Go errors, switching cost, and AX-CPT sensitivity) eliminated through regularization. Together, these results indicate that CCC consolidates the shared variance of updating, restraint, and flexibility without capturing stimulus-driven orienting. While the composite score proves useful for rapid screening and longitudinal tracking, the modest unique variance of each subcomponent cautions against replacing detailed process measures with a single metric. The study fills a methodological gap by integrating network topology with sparsity-based modelling and offers a framework for targeted interventions that bolster executive performance through the channels most relevant to CCC.