Modeling Eye Movements and Response Times in Consumer Choice
Modeling Eye Movements and Response Times in Consumer Choice
复制标题
对消费者选择中的眼球运动和响应时间进行建模
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Stephanie M. Smith
中科院分区：
文献类型：
--
作者：
I. Krajbich;Stephanie M. Smith
人们的选择不是瞬间的，也不是完全自我一致的。虽然这两个事实乍看起来似乎毫无关联，但它们实际上是密不可分的。决策科学家习惯于使用Logit和Probit模型来解释他们选择的数据中的“噪音”。但是，这些行为不一致背后的驱动力是什么呢？随机效用理论(RUT)在这方面提供的指导作用很小。虽然RUT为处理随机选择提供了一个数学基础，但它不知道噪声是由于决策者和/或选择环境的未观察到的特征所致，还是由于实际的“错误”所致。这种区别很重要，因为前者意味着从决策者的角度来看，她的选择是完全一致的，而后者意味着决策者自己可能会对她的一系列选择感到惊讶。在这里，我们认为非选择(“过程”)数据强烈支持后一种解释。我们不认为选择是存储偏好的瞬间实现，而是将选择概念化为信息积累和比较的动态过程。采用从认知心理学到经济选择的“序贯抽样模型”，我们说明了选择和反应时间数据之间令人惊讶的复杂关系。最后，我们回顾了最近的数据，这些数据展示了如何将眼球跟踪和神经记录等其他过程测量方法整合到这种建模方法中，从而对选择过程产生了进一步的见解。
Abstract Peoples’ choices are not instantaneous, nor are they perfectly self consistent. While these two facts may at first seem unrelated, they are in fact inextricably linked. Decision scientists are accustomed to using logit and probit models to account for “noise” in their choice data. But what is the driving force behind these behavioral inconsistencies? Random utility theory (RUT) provides little guidance in this respect. While providing a mathematical basis for dealing with stochastic choice, RUT is agnostic about whether the noise is due to unobserved characteristics of the decision maker and/or the choice environment, or due to actual “mistakes.” The distinction is important because the former implies that from the point of view of the decision maker, her choices are perfectly consistent, while the latter implies that the decision maker herself may be surprised by her set of choices. Here we argue that non-choice (“process”) data strongly favors the latter explanation. Rather than thinking of choice as an instantaneous realization of stored preferences, we instead conceptualize choice as a dynamical process of information accumulation and comparison. Adapting “sequential sampling models” from cognitive psychology to economic choice, we illustrate the surprisingly complex relationship between choice and response-time data. Finally, we review recent data demonstrating how other process measures such as eye-tracking and neural recordings can be incorporated into this modeling approach, yielding further insights into the choice process.
影响因子：
2.5
作者：
Shadlen, MN;Newsome, WT
通讯作者：
Newsome, WT
影响因子：
3.6
作者：
Chabris CF;Laibson D;Morris CL;Schuldt JP;Taubinsky D
通讯作者：
Taubinsky D