Stochastic Theory, Method and Experimental Design in the Identification of Mental Architecture

这项研究将开发人们如何执行简单的心理任务的数学模型，这些任务涉及同时呈现的一种以上类型的信息。心理学中许多能够对实验做出明确预测的理论和模型都是非常具体和特殊的。在合同中，这个项目将继续汤森对重要心理学问题的一般数学表示的发展。然后，这些一般性的数学表示将被用来对这些问题提出的关键问题进行决定性的实验测试。这种方法的普遍性意味着，这些实验将支持或证伪行为的最广泛的潜在“规律”，而不是这种自然规律的非常具体和有限的方面。原则和问题的严格数学化意味着预测严格遵循理论中的假设，实验测试真正评估声称评估的东西，这在口头陈述的理论中并不总是正确的。这项研究将超越汤森最近的数学理论，这些理论基于这样的假设，即精神努力的每个连续阶段必须在下一个阶段开始之前完成。虽然这一假设在许多情况下很可能是正确的，比如在心算中，但在其他任务中，它可能是错误的，因为在其他任务中，信息可能会自由地在大脑的相关部分自由流动。例如，在识别熟悉的面孔时，当感觉信息进入时，它很可能同时流向视觉长期记忆中心、情感记忆区域，甚至是导致动作早期部分的区域。因此，可能不需要首先接收和处理感觉，然后以离散的阶段将完成的感觉信息传递到识别设备等。这项研究将开发允许认知系统不同重要部分之间持续流动的模型。第二个关键假设是认知的独立加工阶段。同样，这在某些情况下可能是合理的，但在其他情况下则不合理。这项研究将开始探索当这种独立性被放弃时会发生什么的问题，不同的子过程可能会形成严重的相互作用。这类研究的结果对于构建关于人们如何感知和思考的更准确和更复杂的理论至关重要。这些信息将有助于了解教育过程中的正常精神功能，以及了解精神残疾或功能障碍情况下的精神功能。

This research will develop mathematical models of how people carry out simple mental tasks that involve more than one type of information presented simultaneously. Many theories and models in psychology that are able to make definite predictions for experiments are very specific and particularistic. In contract, this project will continue Townsend's development of general mathematical representations of important psychological issues. These general mathematical representations will then be used to construct decisive experimental tests of the critical questions posed by such issues. The generality of the approach implies that these experiments will support or falsify the broadest possible potential "laws" of behavior, rather than quite specific and rather limited aspects of such natural regularities. Rigorous mathematization of the principles and issues implies that the predictions follow rigorously from the assumptions embedded in the theory, and that the experimental tests really assess what is claimed to be assessed, which is not always true in verbally stated theories. This research will move beyond Townsend's recent mathematical theories, which are based on the assumption that each successive stage of the mental endeavor must be completed before the next begins. Although this assumption is very likely correct in a number of situations, such as in mental arithmetic, it is probably wrong in other tasks where the information may be free to flow continuously throughout the involved parts of the brain. For instance, in the recognition of a familiar face, it may well be that as sensory information comes in, it simultaneously flows to visual-long-term-memory centers, to emotional-memory areas, and even to areas causing early portions of an action. Thus, it may not be necessary to first receive and process the sensation, then pass on the completed sensory information to a recognition device, and so on, in discrete stages. This research will develop models that permit such continuous flow among the various important parts of the cognitive system. A second critical assumption has been an independence of the separate processing stages in cognition. Again, this may be reasonable in certain contexts but not in others. The research will begin to probe the question of what happens when this independence is given up and the different subprocesses can form serious interactions. The outcome of such research is vitally needed to construct more accurate and sophisticated theories of how people perceive and think. This information will be useful for understanding of normal mental functioning within the education process as well as for understanding mental functioning in cases of mental disabilitiy or disfunction.