Decision heuristics in contexts exploiting intrinsic skill

Abstract

Heuristics can inform human decision making in complex environments through a reduction of computational requirements and a robustness to overparameterisation. However, tasks capturing the efficiency of reduced decision dimensionality typically ignore action proficiency in determining rewards. The value of movement parameterisation in sensorimotor control questions whether heuristics preserve efficiency when actions are non-trivial. We developed a novel selection-execution task requiring joint optimisation of action selection and spatio-temporal skill. Optimal choices could be determined by either a spatio-temporal forward simulation or a simpler spatial heuristic. Sequential-sampling models of action-selection response times parsimoniously distinguished human participants who adopted either strategy. Heuristics preserved broad decisional advantages over forward simulations. In addition, heuristics aligned with greater action proficiency, though predominantly through the core feature (spatial) shaping their decision policy. We accordingly reveal evidence that the dimensionality of information guiding action selection might be yoked to the granularity of plasticity in the motor system.

Cardiac-sympathetic contractility and neural alpha-band power: cross-modal collaboration during approach-avoidance conflict

“…Third, conflict arising from integrated reward and cost recruits a broad array of frontal cortical brain networks 4,8,9,[12][13][14][15][16] , in addition to increased cardiac-sympathetic responses 4,5 , suggesting the paradigm drives qualitative shifts both in behavior and across a complex physiological cascade, allowing us to test cross-modal relations. To decompose behavior even further, and extract fine-grained assays of behavioral perturbations to correlate with physiology signals, we fitted the drift-diffusion model [17][18][19][20][21][22][23][24][25] (DDM) to choice and RT data (Figure 1D).…”

“…Initially considered in perceptual contexts [17][18][19] , parameters from the DDM are an increasingly useful tool for disambiguating the underlying reasons for lengthier RT in more complex valuebased contexts [20][21][22][23][24][25] (such as people incorporating more information into decisions vs accumulating information less efficiently -Figure 1D).…”