The effects of increasing target prevalence on information processing during visual search
The effects of increasing target prevalence on information processing during visual search
复制标题
DOI：
10.3758/s13423-014-0686-2
复制
发表时间：
2015-04-01
影响因子：
3.5
通讯作者：
Donnelly, Nick
中科院分区：
文献类型：
--
作者：
Godwin, Hayward J.;Menneer, Tamaryn;Donnelly, Nick
在视觉搜索过程中出现目标的试验比例（称为目标流行率）会影响目标被检测到的概率。随着患病率的增加，参与者会倾向于报告目标存在。这种偏差导致目标检测率增加，同时误报的可能性也增加。先前的研究表明，随着患病率的增加，参与者花费越来越多的时间来搜索目标缺失的试验。本研究的目的是确定随着患病率增加，在寻找目标缺失试验所花费的额外时间期间的信息处理。我们记录了参与者在进行低流行率（25% 目标存在试验）、中流行率（50%）或高流行率（75%）搜索时的眼动行为。普及率的增加主要通过增加检查显示器中的对象所花费的时间来影响搜索，而不是增加每个显示器中检查的对象的比例。此外，在高流行率的目标缺失试验中花费额外的时间来检查物体是重新访问物体的结果。我们讨论了这些结果对当前搜索模型的影响以及减轻流行效应的持续努力。
The proportion of trials on which a target is presented (referred to as the target prevalence) during visual search influences the probability that the target will be detected. As prevalence increases, participants become biased toward reporting that the target is present. This bias results in an increase in detection rates for the target, coupled with an increased likelihood of making a false alarm. Previous work has demonstrated that, as prevalence increases, participants spend an increasing period of time searching on target-absent trials. The goal of the present study was to determine the information processing during the additional time spent searching on target-absent trials as prevalence increased. We recorded participants' eye movement behavior as they were engaged in low-prevalence (25% target-present trials), medium-prevalence (50%), or high-prevalence (75%) search. Increased prevalence primarily influenced search by increasing the time spent examining objects in the display, rather than by increasing the proportion of objects examined in each display. In addition, the additional time spent examining objects in high-prevalence target-absent trials was the result of revisiting objects. We discuss the implications of these results in relation to current models of search as well as ongoing efforts to alleviate the prevalence effect.