Rouder J N
Department of Psychology, Northwestern University, USA.
J Exp Psychol Hum Percept Perform. 2000 Feb;26(1):359-78.
Many models of perceptual processing assume that participants integrate stimulus evidence over time, for example, random walk models. This class of models is tested in a luminance discrimination paradigm in which the onsets of the stimuli are either instantaneous (stepped) or slowly ramped. The ramped portion of ramped stimuli occurs prior to the stepped stimuli onsets. Consequently, there is more luminance energy in ramped stimuli. Therefore, if participants integrate luminance energy, they should perform better to ramped stimuli. This did not occur in 4 experiments. Participants performed better to stepped stimuli than ramped stimuli in earlier foreperiods and the reverse in later foreperiods. A new model is proposed in which participants monitor both integrated luminance energy and quick temporal changes in luminance, but they do so in a serial fashion. First, participants monitor temporal changes in luminance; later, they monitor integrated luminance energy.
许多感知处理模型假设参与者会随着时间整合刺激证据，例如随机游走模型。这类模型在亮度辨别范式中进行测试，其中刺激的起始要么是瞬间的（阶梯式），要么是缓慢上升的。上升式刺激的上升部分发生在阶梯式刺激起始之前。因此，上升式刺激中有更多的亮度能量。所以，如果参与者整合亮度能量，他们对上升式刺激的表现应该更好。在4个实验中并未出现这种情况。在较早的预备期，参与者对阶梯式刺激的表现比对上升式刺激更好，而在较晚的预备期则相反。提出了一个新模型，其中参与者同时监测整合的亮度能量和亮度的快速时间变化，但他们是以串行方式进行的。首先，参与者监测亮度的时间变化；之后，他们监测整合的亮度能量。
