摘要
通过2(外在锚类型:高锚VS低锚)×2(内在锚类型:有VS无)被试间设计,考察不同锚定信息来源:由外部世界提供的外在锚与个体自身内部产生的内在锚信息对锚定效应及其加工机制的影响。结果发现:(1)当内在锚不存在时,外在高低锚组的估计值有显著差异,当内在锚存在时此种差异变得不显著;(2)当内在锚存在时,外在高低锚组被试的答题反应时有显著差异,内在锚与外在锚一致时反应时比不一致时更快,当内在锚不存在时,此种差异变得不显著。实验结果表明,当锚定调整机制与选择通达机制同时存在时,前者更占优势;锚定信息一致性会影响不同加工机制的启动,一致的信息会激活选择通达机制,不一致的信息则会激活锚定调整机制。
The anchoring effect, one of the most common heuristics, refers to a phenomenon that the assimilation of a quantitative estimates a previously presented irrelative number under uncertain circumstances. Since the anchoring effect was proposed by Tversky and Kahneman in 1974 , lots of research has been done about it. Now there are two main models while account for anchoring effects：the anchoring and adjustment heuristic and the selective accessibility model. But most of researches just focused on the anchoring and adjustment heuristic or the selective accessibility model separately. In recent years, some scholars have pointed out that even if participants have a self-generated anchor, when there is an experimenter-provided anchor showing up at the same time, the processing mechanism of the anchoring effect is not only anchoring and adjustment heuristic, but also may happen to the selective accessibility model at the same time. But no research has been available to prove it yet. Therefore, this study is to try to explore the relationship between anchoring and adjustment heuristic and the selective accessibility model with the self-generated anchor and the experimenter-provided anchor showing up at the same time. A total of 64 undergraduate or graduate students were recruited and paid for their participation. The participants, were placed in separate cubicles in the laboratory, each containing a computer used to present the information and to register the dependent measures. The participants were told that the experiment was about how people made estimates. First, they had to estimate if there was predominance of capital or lower case letters in several letter combinations. After that, they had to make some probability estimates. The letter combination task was used to conceal the true purpose of our experiment. In this task, the participants had to estimate if a letter combination of six letters （e.g. MJFqRe） contained more capital letters or more lower case letters. After the last letter combination, a screen was presented for 5s with the announcement that this was the last letter combination and that they had to make the probability estimate. Immediately after the presentation of this screen, the participants had to estimate：whether （1） the normal boiling point of alcohol was higher than 89 degrees Celsius or below 89 degrees Celsius. Or （2） the normal boiling point of alcohol was higher than 243 degrees Celsius or below 243 degrees Celsius. After the participants’ answer, he/she was told that the numerical value （89/243） was just selected randomly by the computer;it had no value for reference. Then they had to estimate the real degrees Celsius of the boiling point of alcohol. Finally, the participants had to answer whether he/she considered the normal boiling point of water was 100 degrees Celsius, and used it as reference when they were answering what was the normal boiling point of alcohol. The results show that （1） only without self-generated anchors is there a significant difference in participants’ estimates between those exposed to the high anchor and those exposed to the low anchor. （2） Only with self-generated anchors is there a significant difference in participants’ estimated response latencies between those exposed to the high anchor and those exposed to the low anchor. The results of our study suggest that the anchoring and adjustment heuristic are more dominant than the selective accessibility model when they show up at the same time. The consistency of anchor information will affect different processing mechanisms’ activation, anchor-consistent information will activate the selective accessibility model, and anchor-inconsistent information will activate anchoring and adjustment heuristic, which makes the participants’ estimated reaction time longer than the former.
出处
《心理科学》
CSSCI
CSCD
北大核心
2012年第1期171-176,共6页
Journal of Psychological Science
关键词
锚定效应
内在锚
外在锚
锚定调整机制
选择通达机制
双加工机制
anchoring effects experimenter provided anchor self-generated anchor anchoring and adjustment heuristic selective accessibility model dual-processing model
