Newly-minted Dr. Audrey Morrow publishes new study examining how spatial attention influences neural processing associated with perceptual decisions

Whether or not attending to a stimulus changes how the brain processes that information at a very early sensory level remains a long-standing debate in the cognitive neurosciences. Studies in humans have shown both evidence against and in favor of attention having an effect on the earliest cortical response to a visual stimulus (the C1 ERP component). We speculate that difficulty in resolving this issue comes from the fact that typical paradigms elicit rather small C1 responses, making their modulation by attention difficult to discern. In Audrey's paper, specially-designed stimuli elicited C1 responses several times greater than most prior work, yet this signal was not modulated by participant's spatial focus of attention, suggesting it reflects a largely feed-forward driven response. Instead, spatial attention was found to speed up perceptual decisions and modulate a brain response thought to index the accumulation of sensory evidence. This study suggests that the first cortical response to a visual stimulus is not readily modulated by spatial attention but later, decision-related processing is.

 And HUGE CONGRATS to Dr. Audrey Morrow for successfully defending her PhD and landing a stellar post-doc at UCSF!

Wei Dou's new paper showing that EEG signals that reflect evidence accumulation dynamics uniquely predict perceptual confidence ratings is now out in Psychological Science

While making perceptual decisions, people often feel a sense of confidence in their choice. This metacognitive capacity is critical for making adaptive behaviors, such as learning from mistakes or optimizing the allocation of limited cognitive resources. In this research, we investigated how the brain computes confidence by recording electroencephalograms (EEGs) while participants were performing motion-perception tasks and rating confidence. We found a neural signal around 300 ms after viewing the motion stimulus and 250 ms prior to making a decision that indexed participants’ confidence regardless of their decision accuracy, response time (RT), and the evidence quality of the stimulus. Our findings indicate that confidence computation in the brain is more complex than just evaluating whether the decision is right or wrong, how quickly people make decisions, and how good the evidence contained in a given stimulus is.

April Pilipenko's new paper gets featured in the Journal of Neuroscience! We show how alpha waves and pupil fluctuations independently shape vision

We found that synchronous brain waves in the alpha-band (around 10 Hz) lead to inhibited visual perception regardless of the stimulus intensity level, even during stimulus absent conditions. This type of "additive" mechanism implies that alpha waves may boost or inhibit visual detection regardless of whether there is a stimulus, leading to false percepts (or hallucinations). Simultaneously, changes in people's pupil size, which reflect arousal levels, revealed a boost in perception for strong but not weak stimuli, suggesting that high arousal leads to more accurate perception (with fewer false percepts). Thus, alpha waves and pupil-linked arousal have different effects on perception, but both contribute to variability in our experience of the same stimulus. Congrats to April on publishing your first-year project!

Special Focus in the Journal of Cognitive Neuroscience: Alpha-band dynamics and temporal processing.

Huge thanks to all the contributors who wrestled with the question of if and how alpha-band oscillations relate to temporal processing. The contributions span new empirical work, meta-analyses, and opinion pieces motivated, in part, by recent null findings in our field. Into to the Special Focus here: 

Alpha-band Brain Dynamics and Temporal Processing: An Introduction to the Special Focus

J Samaha, V Romei Journal of Cognitive Neuroscience

New pre-print on how alpha oscillations and pupil fluctuations jointly shape visual detection and confidence

In a project led by PhD student April Pilipenko, we found that states of weak ongoing alpha oscillations lead to a boost in visual detection and confidence that was independent of stimulus strength. In other words, spontaneous alpha amplitude  fluctuations exerted additive gain on perception, boosting detection reports even for absent stimuli and increasing confidence for strong stimuli and decreasing confidence for weak stimuli. Conversely, pupil-fluctuations occurring contemporaneously with alpha changes exerted  multiplicative gain on stimulus detection and confidence. Large pupil states (thought to reflect high arousal) boosted confidence and detection more for strong than weak stimuli. Importantly, pupil-independent fluctuations in alpha were still found to induce an overall additive factor to detection, suggesting that alpha and pupil-linked arousal have dissociable influences on perception and that alpha activity alters perception via an additive shift in excitability that is independent of stimulus strength. 

Three new lab publications

PhD Student Audrey Morrow led a project reviewing evidence for the 'functional inhibition' theory of alpha oscillations by summarizing the literature linking attention modulation of pre-stimulus alpha power and attention modulation of subsequent stimulus-evoked brain responses. Audrey's review finds that most studies that examine attention-related changes in alpha do not correlate those changes with attention effects on stimulus-evoked responses, and those that do find mixed evidence for whether pre-stimulus alpha modulations predict changes in sensory- of decision-related brain responses.

In a collaboration with researchers at Georgia Tech, led by postdoc Johan Nakuci, we discovered two distinct sub-types of brain activity that participants switch between during perceptual decision making. These two sub-types reflect different decision strategies corresponding to changes in thresholds required to terminate sensory evidence accumulation.

This project investigated the impact of feedback on serial dependence (the finding that perceptual reports are biased towards recently seen stimuli). We found that providing feedback virtually abolishes serial dependence in a 3D motion perception task. This implies that serial dependence may be partly post-perceptual since simply telling subjects if they were correct or not is not expected to rapidly shape perception.

New paper shedding doubt on the idea that alpha frequency modulates the latency of early visual processing

PhD students Audrey Morrow and Wei Dou analyzed data from 42 participants who underwent visual stimulation using stimuli designed to produce "C1" ERP components (electrical activity thought to index afferent visual cortex responses). They correlated the latency of each participants C1 with their individual alpha frequency to test the hypothesis that individuals with faster alpha frequencies have earlier visual cortex activation. Using several different methods of quantfying C1 latency, they found virtually no correlation between the two, suggesting that, although alpha activity may reduce the amplitude of sensory responses, alpha frequency does not seem to modulate their latency.

The paper appears in a special issue in Frontiers in Neuroscience alongside other great papers investigating peak oscillation frequencies.

Individual alpha frequency appears unrelated to the latency of early visual responses

Lab eats (and presents) at the Cognitive Neuroscience Society (CNS) meeting!

Check out some of the ongoing work we shared in poster form!

New paper on the influence of expectations on serial dependence.

MCD biology PhD student Stefan Abreo helped lead a project investigating how our recent perceptual experience influences current perception and how this is impacted by expectations. We found that in the absence of expectations, people report stimuli as being more similar to the past than they really were. However, when stimuli were expected they were reported as being further away from the past than they really were. When stimuli violated expectations they were once again reported as being more similar to the past. This shows that expectations can reverse serial dependence from an attractive to a repulsive bias and that serial biases do not depend on continuity in the environment - explicitly violating expectations can still produce strong serial effects.

New paper on the role of PFC in perceptual metacognition

There is growing debate around the role of the prefrontal cortex (PFC) in sensory metacognition. In this paper, lead by UCSB prof. Regina Lapate, we add new evidence in support of a causal role of PFC to human metacognition by perturbing the dorsal-lateral PFC with MRI-guided transcranial magnetic stimulation (TMS). TMS to dlPFC was specifically found to impair metacognitive awareness (the extent to which subjective visibility ratings discriminate between task performance) without changing task performance. This was all done in the context of a face discrimination task, thereby extending prior results into the domain of higher-level perception of socially-relevant stimuli. Open access link below!

Perceptual metacognition of human faces is causally supported by function of the lateral prefrontal cortex

New preprint: The positive evidence bias in perceptual confidence is not post-decisional

We investigated at what stage of processing the "positive evidence bias" (PEB) in human confidence occurs. We find that the bias is not explained by post-decisional mechanisms (e.g., by assuming that additional evidence accumulates after the perceptual decision is made). This suggest the PEB emerges in perception itself or during the decision process.

biorXiv link / data and code

"Confidence Database" paper accepted in Nature Human Behavior

We are thrilled to be part of a new resource for the community working on confidence and decision making. As part of a multi-site initiative led by Doby Rahnev, a new paper describing a database of confidence experiments with over 8,700 participants has just been accepted at Nature Human Behavior. We contributed some of our datasets on perceptual estimation and confidence in perception and short-term memory. 

Data and code are here (https://osf.io/s46pr/)

Pre-print is here (https://psyarxiv.com/h8tju)

The database is still accepting experiments. Consider contributing!

STEM Diversity Poster Session

Undergraduate researchers Lleymi Martinez (right) and Olenka Graham Castaneda (left) present their work in the lab looking at neural correlates of subjective confidence in motion perception. We used a novel paradigm that controls for objective accuracy in motion direction discrimination but leads to changes in subjective confidence. Under these conditions, it seems like late components of the event-related potential - which have previously been associated with confidence under non-performance matched conditions - still tracks subjective reports of confidence!