Philip L. Smith1 and Roger Ratcliff2
1Department of Psychology, University of Melbourne, Victoria, 3010, Australia
2Department of Psychology, Ohio State University, 1885 Neil Avenue, Columbus, OH 43210, USA
Patterns of neural firing linked to eye movement decisions show that behavioral decisions are predicted by the differential firing rates of cells coding selected and nonselected stimulus alternatives. These results can be interpreted using models developed in mathematical psychology to model behavioral decisions. Current models assume that decisions are made by accumulating noisy stimulus information until sufficient information for a response is obtained. Here, the models, and the techniques used to test them against response-time distribution and accuracy data, are described. Such models provide a quantitative link between the time-course of behavioral decisions and the growth of stimulus information in neural firing data. The question of how two-alternative decisions are made is an important one for neuroscience and psychology alike because of the pivotal role played by decision making in translating perception and cognition into action. This translation brings encoded stimulus information into contact with the behavioral intention of the decision maker to produce a goal-directed act. Psychology has a long history of decision-making research that has resulted in detailed mathematical models of underlying processes [1,2] but only recently has it become possible to observe the neural correlates of these processes directly in awake behaving monkeys. To study processes involved in simple two-choice decisions, neuroscientists have used an analog of the two-choice response-time (RT) task from psychology, in which monkeys make saccadic eye movements to indicate their decisions about visual stimuli. Recordings from cells in premotor areas of the frontal lobe and the posterior parietal cortex have shown that the time-course of activity in these cells corresponds well with that of behavioral eye movement decisions [3–7]. This article describes how these developments are leading psychologists and neuroscientists to converge at a common view of the underlying mechanisms. As a result, it could soon be possible to explain behavioral data and single-cell firing data with the same class of mathematical models.
Neural correlates of simple two-choice decisions
Neural activity linked to eye movement decisions has been recorded in several visual tasks (Figure 1) from oculomotor areas including the middle temporal area (MT), the lateral interparietal area (LIP) in extrastriate cortex , the frontal eye field (FEF) [9–11], and the superior colliculus (SC) [12–14]. These structures are part of the circuit that controls saccadic eye movements to behaviorally salient targets .
From neurons to sequential-sampling models
The picture that emerges from these findings is strikingly consistent with statistical decision models that have been developed during the past 40 years in mathematical psychology. Two broad classes of model have been developed that apply to different kinds of decisions. One class, of sequential-sampling models, applies to speeded decisions in perceptual and memory tasks [1,21]. These decisions are typically made within a second or so. A second class, based on economic concepts of expected utility, applies to complex decisions among differently valued alternatives . Both have been linked to recent neurobiological findings but only the former is discussed here. The link between neurobiology and utility-based decision theories is discussed in Refs [7,22].