Waves of Control: Surfing Behavioral and EEG Uni/Multivariate Signatures to Map Proactive and Reactive Control

The overarching aim of this thesis was to explore cognitive control, a multifaceted construct. As outlined in the introduction (Chapter 1), this pursuit was not devoid of challenges, given the inherent complexity of cognitive control. Our research was guided by the Dual Mechanisms of Control (DMC) model, which highlights qualitative differences in control mechanisms rather than treating control as a single entity. Within this framework, we explored proactive and reactive control, for which evidence is still controversial about both their separability and signatures. To achieve this aim, we first laid the foundation of our work, emphasizing the importance of methodological (and statistical) rigor to effectively investigate cognitive control using the influential and widely used Stroop task. This not only marked the inception and fundamental tool for pursuing our goal but also emerged as a key aspect of the entire thesis, potentially extending its contribution beyond its main aim. We indeed started our research journey by performing a methodological review focused on the measurement validity of the Stroop effects, providing researchers using this conflict task with clear methodological criteria (Chapter 2). This served us to design spatial Stroop tasks satisfying the criteria for yielding valid measures of Stroop performance, which were tested and validated in a behavioral study using multilevel modeling to obtain more precise and reliable Stroop measures (Chapter 3). These methodological and statistical novelties, in turn, formed the foundation for our exploration of proactive and reactive control using a multimethod approach, combining behavioral and electrophysiological (EEG) investigations and then trying to bridge behavioral and EEG evidence. We thus proposed a novel approach to manipulate the proxies of proactive and reactive control (LWPC and ISPC) simultaneously and at the trial level. The results of a two-experiment behavioral study suggested that our approach was indeed effective to explore more directly the DMC assumptions, overcoming limitations of previous approaches (Chapter 4). We then turned to the EEG evidence for proactive and reactive control, using multiple complementary analytical approaches to scrutinize them from various angles, yielding a richer perspective on their neural dynamics. Using both univariate and multivariate analyses, we could indeed gain insights into both control processes and representations, respectively. Moreover, within each analytical approach we employed more than one analysis. As such, on the univariate side, we explored not only the temporal (ERP) but also the spectral (ERSP) aspects of control processes, enriching the understanding of how control processes are implemented (Chapter 5). On the multivariate side, our bidirectional exploration using both representational similarity analysis and ridge regression offered insights into how control-related representations are encoded and whether their content can be decoded, respectively, shedding initial light on the informational patterns on which cognitive control relies (Chapter 6). Finally, we concluded this thesis by returning to its origin, with two pre-registered behavioral experiments aimed to decompose the Stroop effect into its fundamental components (Chapter 7), which are only the first step of the next stage of our research journey. The present thesis thus contributed to shedding further light on the existence of multiple control mechanisms, providing evidence for behavioral and EEG differences between proactive and (two forms of) reactive control. Nonetheless, the discussion of our findings (Chapter 8) unmistakably confirms the intricacies inherent in cognitive control, unveiling unresolved issues that warrant future studies and opening the door to numerous exciting avenues for further exploration.