PSYCH OpenIR  > 社会与工程心理学研究室
低事件率效应的机制与对策
其他题名The mechanism and countermeasures of low prevalence effect
王陵宇
学位类型博士
导师孙向红
2015-05
学位授予单位中国科学院研究生院
学位授予地点北京
学位专业心理学
关键词低事件率效应 模拟 X光安检任务 视觉搜索 反应方式 奖惩措施
摘要在视觉搜索任务中,极少出现的目标,其漏报率显著增多的现象,称为低事件率效应。X光行李安检具有事件率极低的特点,低事件率效应可能是这种任务中漏报率较高的原因之一。因而,阐明低事件率效应的产生机制对于提高 X 光行李安检工作绩效具有重要的意义。
目前,关于低事件率效应产生的原因,主要有两种观点,一是优势反应抑制失败的观点,另一种是多重决策模型的观点。从信息加工的角度看,这两种观点并不是互斥的,而是从信息加工的不同阶段对低事件率效应做出解释。尽管已有研究已经探讨了多种因素对低事件率效应的影响,但有些实验结果并不一致。另外有一些因素在以往的研究中还未涉及,需要进一步考察它们的作用。因此,本文在前人研究的基础上,继续深入探讨低事件率效应的产生机制,并探索应对其不良影响的有效措施。
本文依据人类信息加工模型和视觉搜索阶段模型,分别采用行为学方法和眼动追踪技术,对模拟 X 光安检任务中的低事件率效应的产生机制展开研究。在研究一中,依据人类信息的加工模型,将视觉搜索任务视为一个连续的多阶段的信息加工过程。针对影响这一过程的各种因素,如时间压力、显示形式、目标类型、反应方式和奖惩措施等,分别探讨了它们对低事件率效应的影响。研究二通过眼动追踪技术,记录眼动数据,依据 Kundel 等人提出的视觉搜索的四阶段模型,将漏报错误进行归类分析,揭示漏报错误产生的原因。
结果表明:(1)高时间压力和双目标搜索会显著增加模拟 X 光安检任务中的漏报错误率;没有目标不按键的反应方式可以显著降低甚至消除低事件率效应;在没有时间压力的条件下,奖惩措施可以显著减小低事件率效应;(2)漏报错误的眼动分析发现,低事件率条件下,扫描错误导致的漏报错误显著多于高事件率条件。
基于以上结果,本文认为低事件率效应产生于视觉搜索信息加工的多个阶段,一方面是源自优势反应抑制失败导致的动作错误,另一方面则源于低事件率条件下判断标准和终止搜索阈限的变化。扫描错误是低事件率条件下漏报错误显著多于高时间率时的主要原因。
其他摘要In visual search, rare targets are missed disproportionately often. This phenomenon is known as the low prevalence effect. Rare target is one of the characteristics of X-ray baggage screening, which may lead to the high miss error rate in this task. Therefore, it is of importance to clarify the mechanism of low prevalence effect  for improving the performance in X-ray luggage screening.  
At present, there are two main  accounts  for the  low prevalence  effect: one is  the failure of  inhibition of  dominant  response; the other is the viewpoint of multiple decision model. From the perspective of information processing, these two views are not mutually exclusive, but from the different stages of information processing to explain the low prevalence effect. Although some studies have discussed the effect of many factors on the low prevalence effect, some results are not inconsistent. Besides, there are other factors  which  are not involved in the past research  and need  to examine their effects. Therefore, based on the previous  research, this paper will continue to  examine  the mechanism of low  prevalence  effect, and explore the effective measures to counteract the adverse effects.
According to the model of human information processing and visual search stage model respectively,  the mechanism  of  low  prevalence effect  in  the  simulated X-ray luggage screening  task  was  studied  by  behavioral approach  and eye tracking technology. In Study 1, according to the human information processing model, visual search task was considered as a continuous multi-stage of information processing. The effects of various factors affecting the process on the low prevalence effect, such as time pressure,  display  pattern, target type, reaction mode and  payoff matrix,  were examined  respectively.  In Study 2,  the eye tracking technology was used to analyze the cause of miss errors in the low prevalence condition. The eye movement data were recorded. According to the four stages model of visual search, proposed by Kundel et al. (1978), miss errors were classified to the corresponding stage of visual search.
The results showed that: (1) high time pressure and dual-target search significantly increased  the  miss error rate in the  simulated  X-ray  luggage screening  task;  the response mode of no target no response markedly reduced or even eliminated the low prevalence effect; in the no time pressure condition, the reward significantly reduced the low event rate effect; (2) according to the eye movement analysis of miss errors, the miss errors resulting from scanning errors in low prevalence condition were significantly more than those in high prevalence condition.
Based on the above results, this paper argues that low prevalence effect occurs  in multiple stages of information processing in visual search. One  is  resulted  from  the motor errors which are caused by the failure of inhibition of dominant response, on the other hand, is derived from  the criterion shift and quitting search threshold drop. The scanning errors are the main reason for the more miss errors in low prevalence condition than those in high prevalence condition.
学科领域应用心理学
语种中文
文献类型学位论文
条目标识符http://ir.psych.ac.cn/handle/311026/19713
专题社会与工程心理学研究室
作者单位中国科学院心理研究所
推荐引用方式
GB/T 7714
王陵宇. 低事件率效应的机制与对策[D]. 北京. 中国科学院研究生院,2015.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
王陵宇-博士学位论文.pdf(3525KB)学位论文 开放获取CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[王陵宇]的文章
百度学术
百度学术中相似的文章
[王陵宇]的文章
必应学术
必应学术中相似的文章
[王陵宇]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。