How to correctly put the "subsequent" in subsequent search miss errors
How to correctly put the "subsequent" in subsequent search miss errors
复制标题
DOI：
10.3758/s13414-019-01802-8
复制
发表时间：
2019-11-01
影响因子：
1.7
通讯作者：
Mitroff, Stephen R.
中科院分区：
文献类型：
--
作者：
Adamo, Stephen H.;Cox, Patrick H.;Mitroff, Stephen R.
视觉搜索，即在干扰物中寻找目标，在理论上很有趣，在实践中也很重要，因为它涉及许多认知能力，对几个关键行业(如放射学、行李检查)至关重要。不幸的是，当显示器中存在一个以上的目标时，搜索尤其容易出错(这种现象被称为搜索效果的满足或随后的搜索未命中效果)。通常的影响是，如果已经检测到第一个目标，观察者更有可能错过第二个目标。拆解底层机制需要在分析和设计中的两个关键方面。首先，为了说明效应的“后续”性质，分析必须在发现第一个目标后，将单目标试验的性能与双目标显示器中第二个目标的性能进行比较。其次，设计必须包括在难度上与每个双目标显示器相匹配的单目标显示器，以实现公平的比较。然而，尚不清楚之前的研究是否同时达到了这两个标准。学术放射学的工作主要使用了匹配良好的单目标和双目标试验的设计，但大多数使用的分析并不只关注第一个目标被检测到后的性能。认知心理学的研究大体上进行了正确的分析，但依赖于不匹配的单目标和双目标试验，引入了可能扭曲结果的混淆。在这篇论文中，我们用经验数据论证了这种混乱的影响，并为适当的研究设计和分析提供了路线图。
Visual search, finding targets among distractors, is theoretically interesting and practically important as it involves many cognitive abilities and is vital for several critical industries (e.g., radiology, baggage screening). Unfortunately, search is especially error prone when more than one target is present in a display (a phenomenon termed the satisfaction of search effect or the subsequent search miss effect). The general effect is that observers are more likely to miss a second target if a first was already detected. Unpacking the underlying mechanisms requires two key aspects in analysis and design. First, to speak to the "subsequent" nature of the effect, the analyses must compare performance on single-target trials to performance for a second target in dual-target displays after a first has been found. Second, the design must include single-target displays that are matched in difficulty to each dual-target display to enable fair comparisons. However, it is not clear that prior research has met these two standards simultaneously. Work from academic radiology has primarily used designs with well-matched single- and dual-target trials, but most employed analyses that do not focus solely on performance after a first target has been detected. Work from cognitive psychology has generally performed the correct analyses, but relied on unmatched single- and dual-target trials, introducing a confound that could distort the results. In the current paper, we demonstrate the impact of this confound in empirical data and provide a roadmap for proper study design and analyses.