A Memory-and-Retrieval Model of Evaluative Learning

人们喜欢什么和不喜欢什么会预测他们的行为。大多数这样的偏好是在我们的生活过程中习得的。先前的研究表明，仅仅观察一个中性刺激(例如一个不知名的人)和一个积极或消极的刺激(例如一个喜欢或不喜欢的人)就会改变我们喜欢或不喜欢中性刺激的程度。这种现象称为评价性条件反射(EC)。这个项目调查了EC背后的学习和记忆过程以及相关的获得偏好的形式。关于EC，可以问几个问题：当人们不记得配对时，它会发生吗？新的偏好能持续多久？它是在心理实验室之外发生的吗？在艾美奖-诺伊特组中，这些问题和其他问题都是从记忆的角度来探讨的。这一观点，特别是我作为这个项目的一部分开发的“评价性学习的陈述性记忆模型”(DMM)，区分了经历配对的学习阶段、评估或表达新偏好的测量阶段，以及学习和测量之间的保持间隔。该模型提供了对导致偏好的过程的动态和经验基础的理解，它指定了一些新的可检验的假设，它允许更好地理解在实际生活中何时可以预期EC效应，这些情景比实验室中的控制要少。在这个项目的第一阶段，对DMM的几个预测进行了测试，并在很大程度上得到了证实：EC效应表明如果配对被记住了。他们也受到影响记忆的因素的类似影响--在学习过程中，在测量过程中，以及在两者之间。为了检验模型的局限性，我们还用其他理论方法测试了假设，这些方法提出了在没有记忆的情况下学习态度的条件，例如，用恐惧或厌恶唤起刺激或嗅觉和味觉配对进行条件反射。然而，在所有这些条件下，我们都没有找到没有记忆的EC的证据。到目前为止，该项目的另一个焦点是研究不足的问题，即EC效应是否也在现实生活中表现出来。当测试人们对日常物品的态度以及他们以前使用它们的经历时，我们确实发现了DMM预测的证据。然而，我们试图在演讲厅中寻找EC效应，但基本上失败了。在扩建项目中，我们计划主要跟进两个问题。第一个问题是，是否存在在忘记配对后出现EC效应的条件。我们测试了两个关于这一点的假设，一个来自DMM，另一个来自之前一个意想不到的发现。其次，我们想要跟进的问题是，EC效应在现实生活中是否以及在什么条件下表现出来。我们提出了两个新的范式，使我们能够在保持实验控制的同时，挑出现实生活中潜在的相关因素。

What people like and what they dislike predicts their behavior. Most of such preferences are learned over the course of our lives. Previous research has shown that simply observing a neutral stimulus (for example an unknown person) together with a positive or negative stimulus (for example a liked or disliked person) changes how much we like or dislike the neutral stimulus. This phenomenon is called evaluative conditioning (EC). This project investigates the learning and memory processes that underlie EC and related forms of acquiring preferences.Several questions can be asked about EC: Does it occur when people do not remember the pairings? How long lasting are the new preferences? Does it occur outside the psychological lab? In the Emmy Noether-Group these and other questions are approached from a memory perspective. This perspective, and specifically the “Declarative Memory Model of Evaluative Learning” (DMM), which I developed as part of this project, distinguishes between a learning phase during which the pairings are experienced, a measurement phase, during which the new preferences are assessed or expressed, and a retention interval between learning and measurement. The model offers a dynamic and empirically founded understanding of the processes that result in preferences, it specifies a number of new testable hypotheses, and it allows a better understanding of when EC effects can be expected in real-life situations that are less controlled than in the lab.In the first phases of this project, several predictions of the DMM were tested and largely confirmed: EC effects show if pairings are remembered. They are influenced in a similar way by factors that influence memory – during learning, during measurement, and in between. To test the limits of the model, we also tested hypotheses by other theoretical approaches that propose conditions for attitude learning without memory, for example conditioning with fear or disgust arousing stimuli or the pairing of smell and taste. Under none of these conditions, however, we found evidence for EC without memory. Another focus of the project so far was the understudied question of whether EC effects also show in real-life settings. We did find evidence for a prediction of the DMM when testing people’s attitudes about everyday objects and their previous experiences with them. Our attempts to find EC effects in a lecture hall, however, largely failed. In the extension project, we plan to follow up on mainly two question. The first is whether there are conditions under which EC effects show after pairings are forgotten. We test two hypotheses on this, one derived from the DMM and one from an unexpected previous finding. Second, we want to follow-up on the question of whether and under which conditions EC effects show in real life. We propose two new paradigms that allow us to single out potentially relevant factors in real-life situations while maintaining experimental control.