Evidence Accumulation

The sense of agency (SoA) is the experience of control over actions and outcomes. Though SoA is a fundamental human experience, there is no consensus on the mechanism or theoretical explanation behind it. Additionally, there are few formal models. One proposed explanation of SoA is that it can be explained as a function of an individual's goals: if something (internal or external) prevents the individual from making progress on their goals, their sense of agency should decrease (Saad et al., 2024). We extend this explanation and posit that a participant’s SoA is the result of a continuous process by which the participant is evaluating evidence about their actions and outcomes in reference to their goal. To test this explanation, we implemented a version of a classic evidence-accumulation modeling framework, the drift diffusion model (Ratcliff, 1978) to infer parameter values (e.g., drift rate) from data from two online behavioral experiments where participants’ sense of agency was manipulated. Specifically, we test a two-stage model in which evidence accumulation (i.e., drift rate) is determined by two expectation evaluations. In stage one, the evaluation is between the perception of actions and prior expectations. In stage two, the evaluation is between the achieved outcome and the expected goal. SoA is then a combination of the accumulated evidence over the course of these two stages. We discuss limitations, implications, and directions for future work.

The decline in accuracy under time pressure is a well-established property of human cognition. For simple decisions, this effect is typically attributed to reduced evidence accumulation when quick responses are required. We propose an alternative model in which evidence accumulation becomes self-excitatory when speed is emphasized. Analyzing six existing datasets, we demonstrate that this self-excitation model provides a better account of the speed-accuracy trade-off than the traditional evidence-threshold model. We identify a key behavioral phenomenon that the self-excitation account captures more effectively and show that it is more parsimonious than a collapsing-threshold model, which can also explain the data pattern. We suggest that self-excitation arises from a confirmation bias in attention allocation, where initially promising evidence is preferentially attended when fast decisions are required. Finally, we outline a series of behavioral experiments manipulating both time-varying evidence and speed-accuracy conditions to further investigate the role of self-excitation in decision-making under time pressure.

Sequential sampling models are widely used to account for response time (RT) and choice probability data in behavioral decision tasks. The models in this domain assume that noisy evidence is accumulated over time, and they have different assumptions about how this is accomplished. For example, the standard Decision Diffusion model assumes that evidence accumulation for two choice alternatives is one single process, which means that the evidence that favors one alternative is evidence against the other alternative, while the Dual Diffusion model assumes that evidence accumulates separately for the alternatives in two parallel processes. However, these models can exhibit model mimicry, where they perform similarly well in predicting behavioral data despite their different model structures and underlying theoretical interpretations. In this study, we employ both data-informed and model-informed cross-fitting methods to assess model mimicry, and to explore parameter spaces of the standard diffusion decision model and the dual diffusion model. First, we fit the Dual Diffusion model on direct predictions from known Diffusion model parameter space. Second, we fit the two models to simulated data from two experimental paradigms. Third, we apply the cross-fitting procedure to empirical data of a memory recognition task.

Perceptual decision-making provides a framework for understanding how organisms translate sensory evidence into actions, but traditional models face challenges in explaining choice phenomena and motor integration. Despite evidence of both covert and overt motor processes during deliberation, most frameworks treat movement as merely implementing a completed decision. We explore the relationship between action and decision making by extending a proposed framework for embodied choice and independently varying the influence of motor feedback on internal choice variables and the contribution of evidence to action. This new model, Degenerate Embodied Choice (DEC), arbitrates between parallel and embodied theories of choice. We demonstrate that DEC replicates the speed-accuracy trade-off (SAT) degenerately, with embodiment proving both necessary and unique for trading speed and accuracy across urgent and accuracy-emphasised tasks. DEC emulates empirical data both qualitatively and quantitatively, with model-fitted parameters falling exclusively within the embodied set and producing congruent predictive SAT values within a narrow band. We then introduce the Optimality Framework for Embodied Choice (OFEC) as a lens for examining embodied choice through optimality principles. Our findings suggest that complex decision behaviours can emerge from simple underlying principles, whether through geometric properties of decision boundaries or motor-cognitive integration.