Source Monitoring and Response Bias in Human Memory

我提出了两条研究路线，旨在对人类记忆的基本机制进行突破性发现。1.我将继续我长达数十年的来源监控研究，这是一种假设的机制，通过这种机制，思想和图像被归因于特定的来源(例如，记忆与推理；对目击事件的记忆与建议)。三条支线将推动这一研究项目朝着全新和创新的方向发展：(1)在一组共同见证的记忆研究中，我们将采用一种突破性的技术，使用偏振滤光片，使两个人能够并排坐在一起观看相同的背投屏幕，并观看不同版本的视频。当两名证人被邀请一起谈论视频事件时，这为共同证人的污染创造了丰富的条件，使我们能够研究影响一名证人将另一名证人的报告的部分内容纳入他们自己对事件的记忆的可能性的条件。(1B)在一系列研究中，我和我的学生将探索“建议的隐形”，通过这些研究，受试者被引导报告他们没有看到他们实际上看到的特定的目击细节。(1C)在关于源头监测的第三个子线研究中，我将扩展之前与玛丽安·加里(来自新西兰惠灵顿的维多利亚大学)的工作，测试带有非证明性照片的声明可以推动人们相信声明的想法；拟议的新研究线测试了照片可以对受试者对未来的预测产生平行影响的假设。2.提出的第二个主要研究方向旨在对再认记忆中反应偏差的理论解释做出贡献。接受认知测试的受试者可能会表现出自由主义倾向(即，在应该说“不”的时候说“是”的时候犯了更多的错误)，或者保守主义的偏见(更多的是在应该说“是”的时候说了“不”)，或者没有偏见(同样经常犯这两种错误)。偏见极大地影响行为(例如，如果一名士兵错误地将一名平民视为战斗人员，则可能会发生悲剧)。(2A)我将与当时的学生贾斯汀·坎特纳(Justin Kantner)(现在是助手)继续我最近的发现。加州州立大学北岭分校教授)认为，在反应偏差上存在稳定的个体差异：一些人倾向于自由派，另一些人倾向于保守，另一些人则是中立的。我的目标是揭示这种个体差异背后的机制。(2B)我的目标也是阐明刺激材料对反应偏差的影响的基础机制。在单词列表中，回答偏差平均倾向于中性(即使一些单独的主题是保守的，而另一些是自由的)。但我们最近发现，在某些类型的材料(例如，画作扫描)中，偏见是非常保守的。关于个体和材料之间的偏差如何以及为什么不同，之前的研究很少，这项工作将极大地有助于反应偏差模型的建立。

I propose two lines of research aimed at breakthrough discoveries concerning basic mechanisms of human memory. 1. I will continue my decades-long study of source monitoring, the hypothesized mechanisms by which thoughts and images are attributed to particular sources (e.g., memory vs. inference; memory of a witnessed event vs. of suggestions). Three sub-lines will advance this program of research in fundamentally new and innovative directions: (1a) In a set of co-witness memory studies, we will employ a ground-breaking technology that uses polarized filters to enable two people to sit side by side watching the same back-projection screen and see different versions of a video. This creates rich conditions for co-witness contamination when the two witnesses are invited to talk together about the video event, enabling us to study the conditions that modulate the likelihood that one witness will incorporate portions of another's report into their own memories of the event. (1b) In a series of studies my students and I will explore “suggested invisibility,” through which subjects are led to report that they had NOT seen particular witnessed details that they had in fact seen. (1c) In a third sub-line of research on source monitoring, I will extend prior work with Maryanne Garry (Victoria U of Wellington, NZ) testing the idea that accompanying claims with non-probative photos can nudge people toward believing the claims (“truthiness”); the new line of proposed studies tests the hypothesis that photos can have parallel effects on subjects' predictions for the future. 2. The second proposed major line of research aims to contribute to theoretical accounts of response bias in recognition memory. Subjects tested on recognition can exhibit liberal bias (i.e., more often err by saying "Yes" when they should say "No") or conservative bias (more often err by saying "No" when they should say "Yes") or no bias (make the two kinds of errors equally often). Bias greatly affects behaviour (e.g., if a soldier erroneously recognizes a civilian as a combatant a tragedy may ensue). (2a) I will follow up my recent discovery, with then-student Justin Kantner (now Assist. Prof. at Cal State Northridge) that there are stable individual differences in response bias: Some people tend to be liberal, others conservative, others neutral. I aim to uncover the mechanisms that underlie this individual difference. (2b) I also aim to illuminate the mechanisms that underlie effects of stimulus materials on response bias. With word lists, response bias on average tends to be neutral (even though some individual subjects are conservative and others are liberal). But we recently discovered that with certain kinds of materials (e.g., scans of paintings) bias is strikingly conservative. There is very little prior research on how and why bias differs across individuals and materials and this work will greatly contribute to models of response bias.