The log rolls on: Hybrid search with same-category targets and distractors

Journal of Vision(2021)

引用 0|浏览0
暂无评分
摘要
In a “hybrid search”, observers search for any of multiple targets stored in memory. Reaction time (RT) is typically a linear function of visual set size (VSS) and a logarithmic function of memory set size (MSS) (Wolfe, 2012). In previous experiments, targets and distractors were dissimilar objects taken from different categories, and, hence, they could be coarsely discriminated. Would logarithmic search through memory survive if targets and distractors were similar, thus requiring finer discrimination in memory? We compared search patterns when targets and distractors came from unrelated basic object categories to search when targets and distractors were different exemplars from the same category. Observers memorized 4-32 items and performed a pure memory search (Experiment 1: VSS=1, i.e., no visual search) or a hybrid search (Experiment 2: VSS=4 or 8). In both tasks, we found that search was slower and that RTxMSS functions were steeper in the “exemplar” search. However, the results were inconclusive regarding the shape of the RTxMSS function, because of insufficient VSS (Experiment 1) or a high error rate with large MSS (Experiment 2). In Experiment 3, to decrease the error rate, we employed a more engaging hybrid novelty search. On each trial, observers clicked on the one novel target among distractor objects, that had been seen at least once in previous trials (VSS=4). In this paradigm, MSS increases from trial to trial. 107 observers were tested for MSS=4-64. This task kept the error rate at a reasonable level (≤25% even at the highest MSS). Again, we found that the search among similar exemplars was slower than search among unrelated items. Critically, the shapes of the RTxMSS functions were logarithmic in both conditions. We conclude that the “logarithmic” character of memory search is robust and generalizes to the fine-resolution, same-category search.
更多
查看译文
关键词
hybrid search,distractors,targets,same-category
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要