The effect of stimulus strength on the speed and accuracy of a perceptual decision
The effect of stimulus strength on the speed and accuracy of a perceptual decision
复制标题
DOI：
10.1167/5.5.1
复制
发表时间：
2005-01-01
影响因子：
1.8
通讯作者：
Shadlen, MN
中科院分区：
文献类型：
--
作者：
Palmer, J;Huk, AC;Shadlen, MN
知觉判断的速度和准确性都取决于感官刺激的强度。刺激强度高时，准确度高，反应时间快；当刺激强度较低时，准确度较低且反应时间较慢。尽管心理测量函数已被广泛认为是分析准确性和刺激强度之间关系的工具，但响应时间和刺激强度之间关系的相应计时函数尚未得到太多考虑。在本文中，我们描述了基于扩散模型的感知决策理论。其中，决策是基于感官证据随着时间的推移累积到一定程度的。结合简单的缩放假设，比例率和功率率扩散模型可以预测计时函数和心理测量函数的简单分析表达式。在一系列心理物理学实验中，我们表明该理论将响应时间和准确性解释为刺激强度和速度准确性指令的函数。特别是，结果证明了响应时间和准确性之间的紧密耦合。该理论还被证明包含了皮隆定律的预测，即响应时间对刺激强度的幂函数依赖性。该理论的解析计时函数允许人们将精度理论扩展到响应时间。
Both the speed and the accuracy of a perceptual judgment depend on the strength of the sensory stimulation. When stimulus strength is high, accuracy is high and response time is fast; when stimulus strength is low, accuracy is low and response time is slow. Although the psychometric function is well established as a tool for analyzing the relationship between accuracy and stimulus strength, the corresponding chronometric function for the relationship between response time and stimulus strength has not received as much consideration. In this article, we describe a theory of perceptual decision making based on a diffusion model. In it, a decision is based on the additive accumulation of sensory evidence over time to a bound. Combined with simple scaling assumptions, the proportional-rate and power-rate diffusion models predict simple analytic expressions for both the chronometric and psychometric functions. In a series of psychophysical experiments, we show that this theory accounts for response time and accuracy as a function of both stimulus strength and speed-accuracy instructions. In particular, the results demonstrate a close coupling between response time and accuracy. The theory is also shown to subsume the predictions of Pieron's Law, a power function dependence of response time on stimulus strength. The theory's analytic chronometric function allows one to extend theories of accuracy to response time.