摘要：时距是指两个相继事件的时间长度。Fraisse将2-3秒以上的时局信息加工称为时间估计。研究采用预期和回溯两种实验范式，考察了计数顺序对时间估计的影响。研究发现，在两种范式下，正数和倒数对30s的时间估计有显著差异，但是对50s和70s的时间估计差异不显著。范式间存在差异：在预期范式下,30s和70s水平上，倒数情况下的时间估计更短。这一结果符合注意闸门控制理论。在回溯范式下，两种计数顺序的时间估计差值更大。在三种时间水平上，倒数的时间估计均更短。通过两种范式间的比较，研究者认为在简单任务中，两种范式可能涉及注意和记忆等不同的认知过程。

关键词：时间估计，计数顺序，预期范式，回溯范式

Abstract：Time interval is the events between two consecutive lengths of time. Duration estimation refers to the information processing time which is more than 2-3s. Using prospective paradigm and retrospective paradigm, the purpose of the present two experiments was to examine the effects of counting sequence on duration estimation. The results showed that: (1)When the period is 30s counting order significantly influenced duration estimation in both prospective and retrospective paradigm. (2)There is significant difference in two paradigms. In prospective paradigm, counting backwards is shorter then forward in 30 and 70s, which is consistent with the attentional-gate model. In retrospective paradigm, compared with prospective paradigm, difference between two counting orders is greater. The results indicated that, if the task is very simple, different process may affect prospective and retrospective duration judgments.

Key words: duration estimation, counting sequence, prospective paradigm, retrospective paradigm

　　　计数顺序对时间估计有显著影响。在倒数情况下，时间的估计更短。这种影响在预期范式和回溯范式中都存在，但在回溯范式下影响更大。两个范式下都发现了计数顺序和时间长度之间的交互作用。其中在较短时距（30s）上，这种计数顺序的效应最显著。在较长时距（70s）上，计数顺序效应虽然不明显，但存在着一定差异。这提示我们以后在上传或者估计时间的任务中，要关注提示顺序对人们时间估计的影响。如果上传时间比较短，倒数可以帮助人们降低对时间的估计。如果在生活中需要立即对某一时间进行估计，那么正数更有利于人们准确的估计时间。