Learning to trust our senses

为了有效地与环境互动，我们必须估计我们所能感知的可靠性或准确性。例如，如果我们在漆黑的夜晚在公园里寻找我们的狗，我们可能对视觉信号的准确性缺乏信心，因此依赖其他感官（例如听觉）或采用其他策略来获取更多信息。这种对我们感官信心的在线评估取决于大脑中的元认知监测机制，该机制被认为受上下文信息的影响（Deroy, Spence, & Noppeney, 2016； Pereira等人，2020）。虽然许多科学家认为我们确实产生了这种“关于精确的信念”，但目前尚不清楚事实是否如此，以及我们对感官的信心是如何在大脑中编码的。上了年纪的成年人往往会根据感官信号做出较差的决定，初步证据表明，知觉元认知也会随着年龄的增长而下降。这就提出了一种可能性，即糟糕的决策可以用我们对感官精确度的不准确信念来解释。建立知觉元认知的基本认知和神经机制，有助于理解次优信念和行为，特别是老年人的次优信念和行为。这个博士项目将使用多种方法来描述认知和神经机制，这些机制支撑着我们对感官可靠性的元层面信念，并将研究这些机制如何随着年龄的增长而变化。首先，我们将检验观察者形成关于感知的元层面信念的假设，并且这些信念会影响感知信心。通过使用经典的运动一致性任务（Britten, Shadlen, Newsome, & Movhson, 1992）和最先进的感知和元认知表现心理物理测量（Maniscalco & Lau, 2012），我们将测试对信号强度的预期是否会影响感知信心（例如，当观察者预期强信号时，他们是否会对自己的选择更有信心？）然后，我们将测试这种信心偏差是否会转化为行为上的战略变化，例如，如果参与者对他们的决定没有信心，他们会寻求更多的信息来改进他们的选择吗？在研究了这些基本机制之后，我们将使用脑电图（EEG）来记录脑电活动，应用最先进的多元模式解码技术来确定信号强度是如何在大脑中编码的，以及这些表征是如何被关于信号强度的元认知信念所塑造的。参与者在完成同样的任务时，大脑活动会被记录下来。首先，分析将确定客观信号强度是编码在初级视觉区域还是与决策信心有关的“高级”大脑区域。一旦这些区域被确定，我们将询问模式强度是否会被自上而下的期望所改变，并将使用脑电图的优秀时间分辨率来询问预测偏差是否会在刺激出现之前出现。然后，经颅磁刺激（TMS）实验将使用非侵入性脑刺激来测试这些脑区域在前面实验中发现的影响中的因果关系。最后，我们将在这些研究中建立的机制框架的基础上，通过广泛的年龄抽样来探索元层面信念在整个生命周期中是如何变化的。这将确定在健康衰老过程中发生的这些机制的任何变化，使我们了解随着年龄的增长，我们如何监控和依赖我们的感官，以及关于感官的信念如何出错。理解这些与年龄相关的知觉元认知和自我意识的变化对心理学家来说是重要的，对更广泛的社会科学和公共政策具有重要意义。

To effectively interact with the environment we must estimate the reliability or precision of what we can perceive. For example, if we are looking for our dog in the park on a dark night, we may have low confidence in the precision of visual signals and therefore rely on other senses (e.g., hearing) or deploy other strategies to gain more information. This online assessment of confidence in our senses depends on metacognitive monitoring mechanisms in the brain which are thought to be influenced by contextual information (Deroy, Spence, & Noppeney, 2016; Pereira et al., 2020). While many scientists assume we do generate these kinds of 'beliefs about precision', it remains unclear whether this is the case and how confidence in our senses is encoded in the brain. Ageing adults tend to make poorer decisions based on sensory signals and preliminary evidence suggests that perceptual metacognition also declines with age This raises the possibility that poor decision-making could be explained by our inaccurate beliefs about the precision of our senses. Establishing the fundamental cognitive and neural mechanisms of perceptual metacognition holds promise for understanding suboptimal beliefs and behaviour, especially those seen in older adults.This PhD project will use a multi-method approach to characterise the cognitive and neural mechanisms that underpin meta-level beliefs about the reliability of our senses and will investigate how these mechanisms change as we age. Firstly, we will test the hypothesis that observers form meta-level beliefs about perception, and that these beliefs bias perceptual confidence. By using a classic motion coherence task (Britten, Shadlen, Newsome, & Movhson, 1992) and state-of-the art psychophysical measures of perceptual and metacognitive performance (Maniscalco & Lau, 2012), we will test whether expectations about signal strength bias perceptual confidence (e.g., are observers biased to be more confident in their choices when they expect strong signals?). We will then test whether such biases in confidence translate to strategic changes in behaviour e.g., if participants are not confident in their decision will they seek more information to improve their choice? After investigating these fundamental mechanisms, we will use electroencephalography (EEG) to record electrical brain activity, applying state-of-the-art multivariate pattern decoding techniques to establish how signal strength is encoded in the brain and how these representations are shaped by metacognitive beliefs about signal strength. Participants will complete the same task while brain activity is recorded. First, analyses will establish whether objective signal strength is encoded in primary visual areas or 'higher' brain regions implicated in decision confidence. Once these regions are identified, we will ask whether pattern strength is altered by top-down expectations and will use the excellent temporal resolution of EEG to ask whether predictive biases emerge even before stimuli are presented. Transcranial magnetic stimulation (TMS) experiments will then use non-invasive brain stimulation to test the causal contribution of these brain areas in the effects found in the preceding experiments.Finally, we will build on the mechanistic framework developed in these studies to explore how meta-level beliefs change across the lifespan by sampling a wide range of ages. This will identify any changes in these mechanisms that occur in healthy ageing, developing our understanding of how we monitor and rely on our senses as we age and how beliefs about the senses may go awry. Understanding these age-linked changes in perceptual metacognition and self-awareness is important for psychologists and has significant implications for the wider social sciences and public policy.