Rational Attention and Adaptive Coding

理性注意力和自适应编码

• 批准号: 1357877
• 负责人: Aldo Rustichini
• 金额: $ 26.21万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-04-15 至 2017-03-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1357877&HistoricalAwards=false
• 关键词: Rational; Attention; Adaptive; Coding

Rational Attention and Adaptive Coding are two theories, developed in very different disciplines that try to explain very similar phenomena. Rational Attention is crucial in explaining important factors in the economic life of a country: for instance, in explaining how people respond to changes in prices, and why and how monetary policy is effective. Rational Attention is currently based on an ad hoc, ultimately weak and unsatisfactory theory of human decision making. Thus, we may be relying on a faulty theory to formulate essential policy decisions like monetary policy. Our proposal wants to provide a sound basis to the theory explaining what people pay attention to. To do so, it will apply to economic decision making the theory of Adaptive Coding, developed in Neuroscience, which explains how the brain adjusts to the environment to maximize the perception accuracy. For example, when one goes from a very bright outside into a dark room, his retina and brain adapts to the new intensity. Rational Attention and Adaptive Coding pose a puzzle. The adaptation to the environment has the functional reason of using optimally information processing, but the adjustment introduces the possibility that environment biases choice. Three questions are the core of our proposed research: 1.What is the trade- off between the improvement in decision making and the bias? 2. If a bias is introduced, how is it corrected? 3. If correction requires time, what is the speed of the correcting mechanism? Our proposal will provide answers to the three questions, using a new and transformative method in the analysis of economic decision making, aimed at producing a fully mechanistic model of choice. In a mechanistic model there is a one-to-one correspondence between the variables in the model and the variables we measure in experimental analysis.

理性注意和适应性编码是两个理论，在非常不同的学科中发展，试图解释非常相似的现象。理性注意力在解释一个国家经济生活中的重要因素时至关重要：例如，解释人们如何对价格变化做出反应，以及货币政策为什么有效以及如何有效。理性注意力目前是基于一个临时的，最终薄弱和不令人满意的人类决策理论。因此，我们可能依赖于一个错误的理论来制定货币政策等基本政策决定。我们的建议希望为解释人们关注什么的理论提供一个坚实的基础。为了做到这一点，它将适用于经济决策的自适应编码理论，在神经科学，它解释了大脑如何适应环境，以最大限度地提高感知准确性。例如，当一个人从一个非常明亮的室外进入一个黑暗的房间时，他的视网膜和大脑会适应新的强度。 理性注意力和适应性编码提出了一个难题。对环境的适应具有最佳利用信息加工的功能原因，但这种调整引入了环境偏向选择的可能性。本研究的核心问题有三个：1.决策改善与偏见之间的权衡是什么？2.如果出现偏差，如何纠正？3.如果纠正需要时间，那么纠正机制的速度是多少？我们的建议将提供三个问题的答案，使用一种新的和变革性的方法在经济决策的分析，旨在产生一个完全机械的选择模型。在机械模型中，模型中的变量与我们在实验分析中测量的变量之间存在一一对应关系。