The Theory of Visual Attention (TVA) typically treats processing speed as fixed within the modelled condition structure. This assumption may be too restrictive when effective processing speed varies from trial to trial. We introduce a variability-augmented extension of a behavioural TVA count model in which trial-to-trial fluctuations in processing speed are represented explicitly through a gamma-distributed speed parameter, while preserving the classical interpretation of processing speed (C), visual short-term memory capacity (K), and perceptual threshold ( t_0 ). A Monte Carlo simulation study evaluated both models, together with the trial-to-trial t_0 -variability model of Dyrholm et al. (2011), across whole-report scenarios with negligible, mild, moderate, and strong trial-to-trial variability in processing speed, and with variability in t_0 . The extended model improved model selection and parameter recovery under strong variability, with the advantage emerging when variability was large enough to induce systematic bias in the fixed-speed model’s estimates of C and t_0 ; under moderate variability the evidence was transitional, and with fewer trials per dataset the precision lost to the additional parameter could outweigh the bias it removed. The two forms of variability left different traces in the data and the two variability models were distinguishable: the t_0 -variability model did not register speed variability and left the bias in C in place, whereas the extended model absorbed only part of the threshold variability. These findings suggest that the proposed extension is best understood as a conditional complement to standard behavioural TVA, appropriate when notable trial-to-trial variability in processing speed is present, rather than as a universal replacement.
We investigate the degree to which labelling information as true or false after exposure to it can influence the degree to which it alters beliefs and test whether valence (whether the information is better or worse than expected) plays a key role in modulating this change. To investigate this, we adapt a classic belief updating paradigm by having participants estimate the likelihood of different life events occurring in the future and then varying information presented to participants along two dimensions: valence (whether the information is better or worse than thought) and accuracy (whether information is labelled true or false). By examining how much participants’ beliefs change post exposure to the information, we find a significant effect along both dimensions. Specifically, participants change beliefs more following information labelled true compared to false and following information that is better compared to worse than expected. This latter effect is consistent with past studies that have revealed a valence effect in belief updating (in response to unlabelled information). The novel finding here is that this valence effect exists in response to false information, and we show evidence–using a Bayes Factor analysis–that it is just as strong in response to false as it is in response to true information. Computational modelling analysis suggests that these patterns of updating arise from the differential use of error signals during learning. Together, the results suggest that labelling information post exposure to it as false has the potential to curb the degree to which it can change beliefs. But this is likely to be less effective in cases where that information presents a rosier view of one’s future. Belief Updating, False Information, Optimistic Update Bias
This paper presents a revision of a recently proposed ACT–R-based cue retrieval model. Memory retrieval—the process of accessing information stored in memory—is often considered crucial to real-time sentence processing. For example, upon encountering “himself” in the sentence “John introduced himself”, interpreting the reflexive requires the retrieval of its antecedent, “John”, from memory. ACT–R-based cue retrieval models assume that morphosyntactic features, such as the masculine feature of “himself”, and information about sentence structure serve as cues for memory retrieval. A recent revision of the traditional ACT–R-based cue retrieval model incorporates the influence of prominent elements (i.e., elements that stand out in the context) and successfully predicts unexpected interference effects observed in some previous studies. However, the revised model fails to predict a target mismatch effect when a distractor, the source of interference, is substantially more prominent than the target. A target mismatch effect arises when the retrieval trigger (“himself”) and the target (“the girl”) mismatch in a morphosyntactic feature, as in “*The girl introduced himself”. The model proposed here addresses this limitation by incorporating a re-retrieval mechanism into the ACT–R-based cue retrieval model. Simulation results show that the proposed model successfully predicts target mismatch effects while retaining the previous model’s ability to account for the influence of prominent elements.
The Intelligence Advanced Research Projects Activity (IARPA) Reimagining Security with Cyberpsychology-Informed Network Defenses (ReSCIND) program was a groundbreaking effort to revolutionize cyber defense using cyberpsychology; it leveraged human cognitive limitations in the cyber domain to mitigate attacker success. For 18-months, five performer teams independently explored methods to assess and influence attack behavior through cyber-themed Human Subjects Research (HSR) studies, such as Capture the Flag (CTF) competitions. Over 2300 human participants were recruited from 59 countries – including over 450 cyber experts. Participants included the general populations, cyber students, and cyber professionals. Each team was required to evaluate two mandatory cognitive vulnerabilities (CogVulns) – Loss Aversion and Representativeness Bias – and at least three additional CogVulns that could be measured using research-backed established methods. Both human data (e.g., demographics; personality; decision-making; risk-taking; and other psychometrics; one team collected physiological data) and cyber data (e.g., host and network data; metrics indicating attack success and task progression) were collected. The human and cyber data are painstakingly linked, enabling detailed exploration of attackers’ specific CogVulns and providing evidence that cyber attack behavior can be manipulated by targeted cyber defensive methods using cyberpsychology. These datasets and published findings to date are openly shared with the scientific community to help trigger further interdisciplinary research in the area.
Understanding how conflict dynamics work in conflict tasks is crucial for unraveling why we can effectively focus on target information while ignoring distractor information. One way to examine conflict dynamics between distractor and target information is to manipulate distractor-target stimulus onset asynchronies (SOAs), which influence the allocation of attention and subsequent adaptations to conflict in the next trial. Both the spatial Stroop and flanker paradigms have been used to investigate the effects of SOA manipulations, revealing distinct conflict dynamics across different SOA conditions. However, it remains unclear whether the underlying cognitive mechanisms responsible for these differing patterns of effects can be explained within a common attention shift framework. To investigate this research question, we integrated an experimental manipulation of the spatial Stroop and flanker tasks with a model-based approach using the revised diffusion model for conflict tasks Lee and Sewell (2024). Our examination of the spatial Stroop and flanker tasks revealed a shared mechanism for within-trial conflict adjustment, where evidence is extracted from the target faster when the target is preceded by a distractor. Furthermore, the results suggest that when distractor information appears first, both tasks share common attributes of conflict dynamics, including reduced urgency in attentional shifts during congruent trials. Based on the behavioral and model-based evidence, an increasing sequential modulation across SOAs was only observed in the spatial Stroop task but not in the flanker task. Notably, our modeling results suggest that a larger sequential effect when distracting information precedes target information for a longer duration is likely due to stronger attentional control following an incongruent trial. This further implies that conflict adaptation is proactive and selective with respect to the congruency of the previous trial.
