Attentional control has become a central yet debated construct in individual-differences research. Traditional measures—usually difference scores based on congruency effects—have been questioned because they show weak correlations across tasks (Rey-Mermet et al., 2018) and divergent delta-function patterns across conflict paradigms (Pratte et al., 2010), calling into doubt the assumption of a unitary construct. To address these psychometric limitations and gain deeper insight into the underlying cognitive mechanisms, we applied variants of the Drift Diffusion Model with time-varying drift rates to data from confound-minimized flanker tasks. Participants completed four-alternative forced-choice versions of arrow, letter, and number flanker tasks, as well as a matrices-based measure of fluid intelligence (HeiQ; Pallentin et al., 2023). Using Amortized Bayesian Inference, we estimated individual-level model parameters and examined shared variance across tasks using structural equation modeling. Results revealed that attentional-control-related parameters formed coherent latent factors across flanker variants, and their association with a latent fluid intelligence factor exceeded those observed for traditional difference scores. However, correlations remained modest, and high-conflict conditions following congruent trials did not differ meaningfully from those following incongruent trials. These findings suggest that computational modeling captures individual differences in attentional control more effectively than standard behavioral measures, while also indicating that attentional-control processes in the flanker task and fluid intelligence may share only limited overlap.