Graduation Year


Document Type




Degree Name

Master of Arts (M.A.)

Degree Granting Department


Major Professor

Edelyn Verona, Ph.D.

Committee Member

Jonathan Rottenberg, Ph.D.

Committee Member

Geoffrey Potts, Ph.D.


Cognitive Control, Disinhibition, ERP, Predictability, Stress


When a threat is detected, brain networks associated with threat processing are activated while other processes are deprioritized. While this resource allocation is adaptive, it makes it especially difficult to effortfully direct thoughts, emotions, and behaviors (use cognitive control) during situations of high stress. Further, this threat response is most efficient in response to short-term or predictable stressors (“threats”) but loses its efficiency for ambiguous or unpredictable threats. Despite research that suggests that threat induces psychological states associated with breakdown in cognitive control processes, no study has directly examined how predictability of threat impacts neurocognitive indicators of cognitive control processes. Thus, the current study sought to fill this gap by examining whether threat is associated with alterations in cognitive control, and if these threat-related alterations relate to disinhibited and impulsive behaviors (emotion-based rash action, problematic alcohol and drug use, self-harm, and aggressive behavior).

The present study used ERPs to examine threat-related alterations in cognitive control and associations with disinhibited behaviors in a diverse, community sample (N = 143). Participants had their brain activity recorded while completing a flanker task under conditions of predictable, unpredictable, and no threat of shock. Disinhibited behavior was quantified using a combination of self-report measures and semi-structured interviews; and exploratory factor analysis was used to derive a latent disinhibition factor, representing shared variance among the different types of disinhibited behaviors. To determine whether these behaviors relate to ERP indices, each type of disinhibited behavior as well as the shared disinhibition factor were entered separately into regression models.

Findings from the overall sample indicated greater early engagement with flanker stimuli during predictable threat (enhanced frontocentral N2 for predictable vs. unpredictable threat) and reduced overall later-stage processing under conditions of threat, especially predictable threat (blunted parietal P3 for threat vs. no threat and predictable vs. unpredictable threats). This suggests a tradeoff between early vs. later stage attention to flanker during predictable threat blocks. Furthermore, relatively predictable vs. unpredictable threat improved accuracy on the task by reducing accuracy decrements for incongruent trials. Conflict processing (N2 or P3 amplitude to incongruent vs. congruent trials) did not vary as a function of threat condition in the overall sample. Contrary to our predictions, associations with disinhibited behavior revealed a pattern of facilitated processing and improvements in accuracy on more difficult incongruent trials under conditions of stress for those scoring higher vs. lower on real-life disinhibited behaviors.

This research expanded on what is known about threat processing and linked it to high risk behaviors with high societal burden. Previous literature suggests that stress disrupts cognitive control, especially for those prone to engaging in disinhibited behaviors. However, our study suggests a more nuanced relationship, whereby stress influences behavior via reallocation of cognitive processing resources. Depending on the predictability of threat and individual differences in disinhibition, this could actually temporally enhance performance. Our findings provide useful evidence to advance theories of cognitive processing under conditions of threat and disinhibition.