The impact of probabilistic feature cueing depends on the level of cue abstraction
The impact of probabilistic feature cueing depends on the level of cue abstraction
复制标题
DOI：
10.1007/s00221-015-4487-2
复制
发表时间：
2016-03-01
影响因子：
2
通讯作者：
Vossel, Simone
中科院分区：
文献类型：
--
作者：
Dombert, Pascasie L.;Fink, Gereon R.;Vossel, Simone
注意力资源的分配取决于对事件可能性的预测。虽然这种效果已被广泛研究，在空间注意域中的目标刺激的位置是预先提示，较少被称为线索的刺激功能，如颜色的行为相关的目标。此外，关于哪种类型的颜色线索对偏置注意有效，也存在分歧。在这里，我们调查的概率上下文（线索有效性的百分比，%CV）的影响，不同层次的线索抽象，以阐明如何基于特征的搜索信息的处理和使用直接注意。通过呈现知觉颜色、颜色词或两个字母的缩写来提示目标的颜色。%CV，即，提示正确指示颜色的概率在50%、70%和90%之间不可预测地变化。在60个数据集中记录每个%CV条件下有效和无效试验的响应时间（RT），并进行方差分析。结果表明，所有的线索与可比的RT成本无效提示后。RT成本的调制概率，但是，取决于线索抽象的水平和时间的任务：虽然一个强大的，直接的影响%CV被发现为两个字母的线索，效果只观察到在下半年的实验知觉和文字线索。这些结果表明，概率特征为基础的信息处理不同层次的线索抽象。此外，环境统计的调节作用不同地依赖于不同特征线索的任务时间。
Allocation of attentional resources rests on predictions about the likelihood of events. While this effect has been extensively studied in the spatial attention domain where the location of a target stimulus is pre-cued, less is known about the cueing of stimulus features such as the color of a behaviorally relevant target. Moreover, there is disagreement about which types of color cues are effective for biasing attention. Here we investigated the effects of probabilistic context (percentage of cue validity, %CV) for different levels of cue abstraction to elucidate how feature-based search information is processed and used to direct attention. The color of a target was cued by presenting the perceptual color, the color word, or two-letter abbreviations. %CV, i.e., the probability that the cue indicated the color correctly, changed unpredictably between 50, 70, and 90 %. Response times (RTs) for valid and invalid trials in each %CV condition were recorded in 60 datasets and analyzed with analyses of variance. The results showed that all cues were associated with comparable RT costs after invalid cueing. The modulation of RT costs by probabilities, however, depended upon level of cue abstraction and time on task: While a strong, immediate impact of %CV was found for two-letter cueing, the effect was solely observed in the second half of the experiment for perceptual and word cues. These results demonstrate that probabilistic feature-based information is processed differently for different levels of cue abstraction. Moreover, the modulatory effect of the environmental statistics differentially depends on the time on task for different feature cues.