Published Abstracts

2021

eLife

Separable neural signatures of confidence during perceptual decisions

Balsdon, T., Mamassian, P.*, and Wyart, V.*
*equal contributors

Perceptual confidence is an evaluation of the validity of perceptual decisions. While there is behavioural evidence that confidence evaluation differs from perceptual decision-making, disentangling these two processes remains a challenge at the neural level. Here, we examined the electrical brain activity of human participants in a protracted perceptual decision-making task where observers tend to commit to perceptual decisions early whilst continuing to monitor sensory evidence for evaluating confidence. Premature decision commitments were revealed by patterns of spectral power overlying motor cortex, followed by an attenuation of the neural representation of perceptual decision evidence. A distinct neural representation was associated with the computation of confidence, with sources localised in the superior parietal and orbitofrontal cortices. In agreement with a dissociation between perception and confidence, these neural resources were recruited even after observers committed to their perceptual decisions, and thus delineate an integral neural circuit for evaluating perceptual decision confidence.

2020

Nature Communications

Confidence controls perceptual evidence accumulation.

Balsdon, T., Wyart, V.*, and Mamassian, P.*
*equal contributors

Perceptual decisions are accompanied by feelings of confidence that reflect the likelihood that the decision was correct. Here we aim to clarify the relationship between perception and confidence by studying the same perceptual task across three different confidence contexts. Human observers were asked to categorise the source of sequentially presented visual stimuli. Each additional stimulus provided evidence for making more accurate perceptual decisions, and better confidence judgements. We show that observers' ability to set appropriate evidence accumulation bounds for perceptual decisions is strongly predictive of their ability to make accurate confidence judgements. When observers were not permitted to control their exposure to evidence, they imposed covert bounds on their perceptual decisions but not on their confidence decisions. This partial dissociation between decision processes is reflected in behaviour and pupil dilation. Together, these findings suggest a confidence-regulated accumulation-to-bound process that controls perceptual decision-making even in the absence of explicit speed-accuracy trade-offs.


2019

Vision Science Society

Graded, multidimensional representations of sensory evidence allow for dissociable performance in second-choice and confidence judgments.

Balsdon, T., Wyart, V.*, and Mamassian, P.*
*equal contributors

Perceptual decisions are made on the basis of complex and often ambiguous sensory evidence. Whilst the outcome of sensory evidence accumulation may reflect the mere crossing of a decision criterion, it is well known that human observers have access to a far richer representation of their sensory environment. Indeed, it is this rich representation that allows observers to assess the accuracy of their perceptual decisions, as in confidence judgements, and to offer a second choice judgement when deciding between more than two options. In this experiment, we examine human observers' ability to use this graded, multidimensional representation of sensory evidence to make second-choice judgements, and how this second-choice process relates to their metacognitive ability. On each trial we presented observers with a series of oriented Gabors, drawn from one of three categories, defined by circular Gaussian distributions centred on -60°, 0° and 60° relative to vertical. Observers' task was to decide which distribution the Gabors were drawn from, by accumulating the evidence for each category over the stimuli presented to them. After making their first choice, observers then provided a second choice - which is the next most likely category? Over pairs of trials, observers also chose which trial they were more confident that they made a correct first choice. A computation model was fit to each observer's first choice decisions to determine their sensory internal noise. An ideal observer that was only limited by this sensory noise was then defined to predict performance in the second choice and confidence judgments. Observers underperformed in the second choice task, but overperformed in the metacognitive decision, relative to the ideal observer. This suggests that these two ways in which observers access the sensory evidence are dissociable, and thus, that these decisions target different aspects of the sensory evidence representation, utilizing distinct computations.


European Conference of Visual Perception

Meta-Meta-Meta Perception

Recht, S., Jovanovic, L., Balsdon, T., and Mamassian, P.

It is often stated that every decision comes with a feeling of confidence. Thus, confidence decisions should themselves come with a feeling of confidence. To test this prediction, participants performed four consecutive tasks on each trial. First, they performed a 2AFC spatial frequency discrimination task, and rated confidence in their performance (low/high). After two consecutive trials, they evaluated which of the two confidence ratings better reflected their performance (confidence forced-choice judgement). This is a "meta-confidence" judgement rather than another confidence judgment on the initial task. Finally, participants rated their performance (low/high) in their confidence forced-choice judgment ("confidence in meta-confidence"). We compared performance in the spatial discrimination task for the three levels of meta-perceptive judgments (eight categories). Spatial frequency discrimination performance was different between trials rated low vs. high at the first level. This difference was greater for trials chosen as having more accurate confidence ratings (second level), and even greater for chosen trials with a high-confidence rating (third level). Therefore, participants performed each meta-judgement with above chance accuracy. Comparison with an ideal observer showed no evidence for a systematic decrease in sensitivity of confidence judgments across the levels. In conclusion, every judgement comes with a reliable feeling of confidence.


European Conference of Visual Perception

A Reverse Hierarchy for Perceptual Confidence?

Balsdon, T., Wyart, V.*, and Mamassian, P.*
* equal contributors

The reverse hierarchy theory (Hochstein and Ahissar, 2002) proposes that conscious, deliberate perception proceeds from top-down, accessing higher levels of visual processing and only tuning down to low-level representations as required. A natural prediction from the reverse hierarchy theory is therefore that metacognitive monitoring will have greater access to high-level compared to low-level visual information. We tested this prediction using two perceptual discrimination tasks on the same visual stimuli: grey-scale faces in which the eye direction and contrast of the irises was manipulated. In the low-level task, observers discriminated whether the left or right eye was higher contrast, a decision relying on information processed in early visual cortex. In the high-level task they discriminated whether the gaze direction was looking to the left or the right of them, a decision that arguably relies on information processed in the Superior Temporal Sulcus (Perrett et al., 1992), high in the visual processing hierarchy. Metacognitive performance was measured in a confidence forced-choice task (Mamassian, 2016): every two trials observers chose which trial they were more likely to be correct. There was weak evidence for better metacognitive efficiency for the high-level compared to the low-level task. 


Computational Systems Neuroscience (CoSyNe)

The accumulation of sensory evidence for decisions and confidence

Balsdon, T., Mamassian, P.*, and Wyart, V.*
*equal contributors

In order to resolve complex and ambiguous sensory input, the observer must accumulate multiple samples of sensory evidence. Accumulation-to-bound models have provided an adequate description of perceptual decisions in terms of behavior and neural correlates. However, these decisions are also accompanied by a feeling of confidence, and a consensus has not yet been reached on how best to computationally describe observers' metacognitive confidence judgments. In a two-session psychophysical experiment, human participants (N = 20) were asked to categorize the source of up to 40 noisy visual stimuli presented sequentially. In the first session, participants could end the presentation of stimuli at any point in the sequence to match target levels of decision accuracy. Bayesian accumulation-to-bound models were fitted to estimate participants' ability to adjust their decision bounds according to target levels of accuracy. In the second session, the same participants first provided their decisions whenever they felt ready, which could be fitted in terms of a 'default' decision bound. They were then cued to provide decisions and confidence ratings at different points relative to this fitted bound. We found that participants' ability to adjust their decisions bounds in the first session predicted the sensitivity of their confidence ratings to the accuracy of their decisions in the second session. Despite this interrelation between decision-making and confidence, we further found that the accumulation of evidence for perceptual decisions and confidence could be dissociated at different points relative to the observers 'default' decision bound. Together, these findings offer to reconcile seemingly antagonistic effects through a confidence-regulated accumulation-to-bound process that controls decision-making even in the absence of any speed-accuracy trade-off.

2018

European Conference of Visual Perception

Metacognitive control of sensory evidence accumulation

Balsdon, T., Wyart, V.*, and Mamassian, P.*
*equal contributors

Perceptual decisions are accompanied by a feeling of confidence as to whether the decision was correct. We asked whether this metacognitive information is used to stop collecting evidence for a decision in complex sensory environments. Observers were gradually shown evidence for an orientation discrimination decision and were asked to stop when they reached a certain level of confidence. In a later session, trials were replayed with sometimes slightly more evidence. On those trials, observers were more confident and yet they were not more accurate. These results highlight the role of confidence in metacognitive control -the selection of sensory evidence- rather than simply metacognitive monitoring -the evaluation of perceptual correctness. Thanks to computational models, we can then quantify the contribution of sensory evidence to perceptual decisions and confidence judgements, and the relationship of suboptimalities in each.


Cognitive Research: Principles and Implications 3:25

Improving face identification with specialist teams

Balsdon, T., Summersby, S., Kemp, R.L., and White, D.

People vary in their ability to identify faces, and this variability is relatively stable across repeated testing. This suggests that recruiting high performers can improve identity verification accuracy in applied settings. Here, we report the first systematic study to evaluate real-world benefits of selecting high performers based on performance in standardized face identification tests. We simulated a recruitment process for a specialist team tasked with detecting fraudulent passport applications. University students (n = 114) completed a battery of screening tests followed by a real-world face identification task that is performed routinely when issuing identity documents. Consistent with previous work, individual differences in the real-world task were relatively stable across repeated tests taken 1 week apart (r = 0.6), and accuracy scores on screening tests and the realworld task were moderately correlated. Nevertheless, performance gains achieved by selecting groups based on screening tests were surprisingly small, leading to a 7% improvement in accuracy. Statistically aggregating decisions across individuals-using a 'wisdom of crowds' approach-led to more substantial gains than selection alone. Finally, controlling for individual accuracy of team members, the performance of a team in one test predicted their performance in a subsequent test, suggesting that a 'good team' is not only defined by the individual accuracy of team members. Overall, these results underline the need to use a combination of approaches to improve face identification performance in professional settings.


Consciousness and Cognition 64 (19-31)

All is not lost: Post-saccadic contributions to the perceptual omission of intra-saccadic streaks

Balsdon, T., Schweitzer, R., Watson, T.*, and Rolfs, M.*
*equal contributors

Saccades rapidly jerk the eye into new positions, yet we rarely experience the motion streaks imposed on the retinal image. Here we examined spatial and temporal properties of post-saccadic masking-one potential explanation of this perceptual omission. Observers judged the motion direction of a target stimulus, a Gaussian blob, that moved vertically upwards or downwards and then back to its initial position, just as observers made a saccade. We manipulated the onset and offset of the target and of distractors in various spatial relations to the target, and assessed their effect on performance and subjective confidence. Although the presence of the target after the saccade caused the strongest omission, the offset of spatially distant distractor stimuli upon saccade offset also impaired performance. The temporal properties of these two separate effects suggest that, in addition to masking, an independent effect of attentional distraction further accentuates perceptual omission of intra-saccadic motion streaks.


Frontiers in Psychology 9:2491

Task dependent effects of head orientation on perceived gaze direction

Balsdon, T., and Clifford, C.W.G

The perception of gaze direction involves the integration of a number of sensory cues exterior to the eye-region. The orientation of the head is one such cue, which has an overall repulsive effect on the perceived direction of gaze. However, in a recent experiment, we found the measured effect of head orientation on perceived gaze direction differed within subjects, depending on whether a single- or two-interval task design was employed. This suggests a potential difference in the way the orientation of the head is integrated into the perception of gaze direction across tasks. Four experiments were conducted to investigate this difference. The first two experiments showed that the difference was not the result of some interaction between stimuli in the two-interval task, but rather, a difference between the types of judgment being made across tasks, where observers were making a directional (left/right) judgment in the single-interval task, and a non-directional (direct/indirect gaze) judgment in the two-interval task. A third experiment showed that this difference does not arise from observers utilizing a non-directional cue to direct gaze (the circularity of the pupil/iris) in making their non-directional judgments. The fourth experiment showed no substantial differences in the duration of evidence accumulation and processing between judgments, suggesting that observers are not integrating different sensory information across tasks. Together these experiments show that the sensory information from head orientation is flexibly weighted in the perception of gaze direction, and that the purpose of the observer, in sampling gaze information, can influence the consequent perception of gaze direction.


Royal Society Open Science 5 (8)

How wide is the cone of direct gaze?

Balsdon, T., and Clifford, C.W.G.

The cone of direct gaze refers to the range of gaze deviations an observer accepts as looking directly at them. Previous experiments have calculated the width of the cone of direct gaze using the gaze deviations actually presented to the observer, however, there is considerable evidence that observers actually perceive gaze to be systematically more deviated than actually presented. Here, we examine the width of the cone of direct gaze in units of perceived gaze deviation. In doing so, we are able to disambiguate differences in width both within and between observers that are due to differences in their perception of gaze and due to differences in what observers consider to be looking at them. We suggest that this line of inquiry can offer further insight into the perception of gaze direction, and how this perception may differ in clinical populations.


Royal Society Open Science 5 (1)

Visual processing: Conscious until proven otherwise

Balsdon, T., and Clifford, C.W.G

Unconscious perception, or perception without awareness, describes a situation where an observer's behaviour is influenced by a stimulus of which they have no phenomenal awareness. Perception without awareness is often claimed on the basis of a difference in thresholds for tasks that do and do not require awareness, for example, detecting the stimulus (requiring awareness) and making accurate judgements about the stimulus (based on unconscious processing). Although a difference in thresholds would be expected if perceptual evidence were processed without awareness, such a difference does not necessitate that this is actually occurring: a difference in thresholds can also arise from response bias, or through task differences. Here we ask instead whether the pattern of performance could be obtained if the observer were aware of the evidence used in making their decisions. A backwards masking paradigm was designed using digits as target stimuli, with difficulty controlled by the time between target and mask. Performance was measured over three tasks: detection, graphic discrimination and semantic discrimination. Despite finding significant differences in thresholds measured using proportion correct, and in observer sensitivity, modelling suggests that these differences were not the result of perception without awareness. That is, the observer was not relying solely on unconscious information to make decisions.

2017

Attention, Perception, and Psychophysics 79 (7)

Detecting and identifying offset gaze

Balsdon, T., and Clifford, C.W.G

A number of experiments have demonstrated that observers can accurately identify stimuli that they fail to detect (Rollman and Nachmias, 1972; Harris and Fahle, 1995; Allik et al. 1982, 2014). Using a 2x2AFC double judgements procedure, we demonstrated an analogous pattern of performance in making judgements about the direction of eye gaze. Participants were shown two faces in succession: one with direct gaze and one with gaze offset to the left or right.We found that they could identify the direction of gaze offset (left/right) better than they could detect which face contained the offset gaze. A simple Thurstonian model, under which the detection judgement is shown to be more computationally complex, was found to explain the empirical data. A further experiment incorporated metacognitive ratings into the double judgements procedure to measure observers' metacognitive awareness (Meta-d') across the two judgements and to assess whether observers were aware of the evidence for offset gaze when detection performance was at and below threshold. Results suggest that metacognitive awareness is tied to performance, with approximately equal Meta-d' across the two judgements, when sensitivity is taken into account. These results show that both performance and metacognitive awareness rely not only on the strength of sensory evidence but also on the computational complexity of the decision, which determines the relative distance of that evidence from the decision axes. 


Scientific Reports, 7

A bias-minimising measure of the influence of head orientation on perceived gaze direction

Balsdon, T., and Clifford, C.W.G

The orientation of the head is an important cue for gaze direction, and its role has been explained in a dual route model. The model incorporates both an attractive and a repulsive effect of head orientation, which act to support accurate gaze perception across large changes in natural stimuli. However, in all previous studies of which we are aware, measurements of the influence of head orientation on perceived gaze direction were obtained using a single-interval methodology, which may have been affected by response bias. Here we compare the single-interval methodology with a two-interval (bias-minimising) design. We find that although measures obtained using the two-interval design showed a stronger attractive effect of head orientation than previous studies, the influence of head orientation on perceived gaze direction still represents a genuine perceptual effect. Measurements obtained using the two-interval design were also shown to be more stable across sessions one week apart. These findings suggest the two-interval design should be used in future experiments, especially if comparing groups who may systematically differ in their biases, such as patients with schizophrenia or autism. 

Frontiers in Psychiatry, 8

Depression reduces accuracy while parkinsonism slows response time for processing positive feedback in patients with parkinson's disease with comorbid major depressive disorder tested on a probabilistic category-learning task

Herzallah, M.M., Khdour, H.Y., Taha, A., Elmashala, A.M., Mousa, h.N., Taha, M.B., ...Balsdon, T., ... and Gluck, M.A.

Major depressive disorder (MDD) is the most common non-motor manifestation of Parkinson's disease (PD) affecting 50% of patients. However, little is known about the cognitive correlates of MDD in PD. Using a computer-based cognitive task that dissociates learning from positive and negative feedback, we tested four groups of subjects: (1) patients with PD with comorbid MDD, (2) patients with PD without comorbid MDD, (3) matched patients with MDD alone (without PD), and (4) matched healthy control subjects. Furthermore, we used a mathematical model of decision-making to fit both choice and response time data, allowing us to detect and characterize differences between the groups that are not revealed by cognitive results. The groups did not differ in learning accuracy from negative feedback, but the MDD groups (PD patients with MDD and patients with MDD alone) exhibited a selective impairment in learning accuracy from positive feedback when compared to the non-MDD groups (PD patients without MDD and healthy subjects). However, response time in positive feedback trials in the PD groups (both with and without MDD) was significantly slower than the non-PD groups (MDD and healthy groups). While faster response time usually correlates with poor learning accuracy, it was paradoxical in PD groups, with PD patients with MDD having impaired learning accuracy and PD patients without MDD having intact learning accuracy. Mathematical modeling showed that both MDD groups (PD with MDD and MDD alone) were significantly slower than non-MDD groups in the rate of accumulation of information for stimuli trained by positive feedback, which can lead to lower response accuracy. Conversely, modeling revealed that both PD groups (PD with MDD and PD alone) required more evidence than other groups to make responses, thus leading to slower response times. These results suggest that PD patients with MDD exhibit cognitive profiles with mixed traits characteristic of both MDD and PD, furthering our understanding of both PD and MDD and their often-complex comorbidity. To the best of our knowledge, this is the first study to examine feedback-based learning in PD with MDD while controlling for the effects of PD and MDD.


European Conference of Visual Perception

Task dependent effects of head orientation on perceived gaze direction

Balsdon, T., and Clifford, C.W.G.

Accurate perception of where someone else is looking is an important social cue that must be utilised across vast changes in stimulus properties and quality. One such changing property is head rotation, which causes a repulsive effect on perceived gaze direction, as the visible part of the sclera changes in a similar way to the rotation of the eyes in the opposite direction. At the same time, the direction of the head can act as a coarse scale spatial cue to the direction of gaze, causing an attractive effect of head orientation on perceived gaze direction. These opposing effects have been accounted for in a dual-route model. Typically, the overall effect of head rotation is measured with a single-interval task, and is shown to be more repulsive. However, using a two-interval task, we found performance was close to veridical. This difference between measurements across the two tasks is weakened when the observer is shown only the eye-region, rather than the full head. This suggests that the way information from the head is incorporated into the perception of gaze direction might be dependent on the task at hand.


European Conference of Visual Perception

Transfer of object information between the periphery and fovea; an MEG study

Watson, T., Balsdon, T., Carlson, T., and Williams, M.

This study investigated the flow of visual information from the periphery to the fovea utilising discriminant cross training of neural activity recorded via MEG. Thirteen participants fixated a central point and completed a one back, object category repetition task. The objects were either 'spikeys' or 'smoothies' and presented in one of five locations; the fovea or four peripheral locations. MEG data was used to train a series of classifiers to identify the object category and location. Classification errors were examined, specifically, times the classifier identified the correct object identity but the incorrect location at test. Overall classification accuracy peaked between 100 to 150 ms post stimulus onset. The most frequent location-only errors were those when the object was presented in the periphery but the classifier identified the object as being at the fovea. These errors occurred at 120ms and 150 ms post stimulus onset, when the overall performance was very high. The classification errors toward the fovea are suggestive of an object identity and location signal occurring at these times that is sufficiently similar to that of the same object presented to the fovea, even though it is presented in the periphery. We interpret this pattern as indicative of transfer of information about the identity of the object from peripheral receptive fields to foveal receptive fields during the visual task.

2016

European Conference of Visual Perception

Effect of audio-visual source misalignment on timing performance

Cass, J., Van Der Burg, E., and Balsdon, T.

This study investigates the psychophysical effect of audio-visual source displacement on auditory timing performance. Two speakers hidden behind a screen were placed along the horizontal meridian at various separations. Each speaker produced a white noise burst using a range of onset lags and subjects reported their temporal order. Predictably, performance improved with increasing speaker separation. This allowed us to equate performance by choosing the speaker separation corresponding to 50% improvement. This provided a baseline for Experiment 2: each trial was accompanied by two synchronous disks projected horizontally onto the screen at various angles of displacement relative to the speakers. The luminance of both disks changed abruptly coincident with the first noise burst, then again with the second burst. Performance improved maximally when the audiovisual signals were aligned then deteriorated gradually with increasing disk eccentricity. Intriguingly, even small audio-visual misalignments in the direction of fixation bore no improvement in auditory TOJ performance. These results suggest that the perceived (ventriloquized) location of auditory events, rather than their physical location, limit the resolution with which humans make auditory timing judgments.


Behavioural brain research 296, 240-248

Probabilistic reward-and punishment-based learning in opioid addiction: experimental and computational data.

Myers, C.E., Sheynin, J., Balsdon, T., Luzardo, A., Beck, K.D., Hogarth, L., ... and Moustafa, A.A.

Addiction is the continuation of a habit in spite of negative consequences. A vast literature gives evidence that this poor decision-making behavior in individuals addicted to drugs also generalizes to laboratory decision making tasks, suggesting that the impairment in decision-making is not limited to decisions about taking drugs. In the current experiment, opioid-addicted individuals and matched controls with no history of illicit drug use were administered a probabilistic classification task that embeds both reward-based and punishment-based learning trials, and a computational model of decision making was applied to understand the mechanisms describing individuals' performance on the task. Although behavioral results showed that opioid-addicted individuals performed as well as controls on both reward- and punishment-based learning, the modeling results suggested subtle differences in how decisions were made between the two groups. Specifically, the opioid-addicted group showed decreased tendency to repeat prior responses, meaning that they were more likely to "chase reward" when expectancies were violated, whereas controls were more likely to stick with a previously-successful response rule, despite occasional expectancy violations. This tendency to chase short-term reward, potentially at the expense of developing rules that maximize reward over the long term, may be a contributing factor to opioid addiction. Further work is indicated to better understand whether this tendency arises as a result of brain changes in the wake of continued opioid use/abuse, or might be a pre-existing factor that may contribute to risk for addiction.

2015

Behavioural brain research 291, 147-154

Drift diffusion model of reward and punishment learning in schizophrenia: Modeling and experimental data

Moustafa, A.A., Keri, S., Somlai, Z., Balsdon, T., Frydecka, D., Misiak, B., and White, C.

In this study, we tested reward- and punishment learning performance using a probabilistic classification learning task in patients with schizophrenia (n = 37) and healthy controls (n = 48). We also fit subjects' data using a Drift Diffusion Model (DDM) of simple decisions to investigate which components of the decision process differ between patients and controls. Modeling results show between-group differences in multiple components of the decision process. Specifically, patients had slower motor/encoding time, higher response caution (favoring accuracy over speed), and a deficit in classification learning for punishment, but not reward, trials. The results suggest that patients with schizophrenia adopt a compensatory strategy of favoring accuracy over speed to improve performance, yet still show signs of a deficit in learning based on negative feedback. Our data highlights the importance of applying fitting models (particularly drift diffusion models) to behavioral data. The implications of these findings are discussed relative to theories of schizophrenia and cognitive processing.


Consciousness and Cognition 32, 79-91

Absolute and relative blindsight

Balsdon, T., and Azzopardi, P.

The concept of relative blindsight, referring to a difference in conscious awareness between conditions otherwise matched for performance, was introduced by Lau and Passingham (2006) as a way of identifying the neural correlates of consciousness (NCC) in fMRI experiments. By analogy, absolute blindsight refers to a difference between performance and awareness regardless of whether it is possible to match performance across conditions. Here, we address the question of whether relative and absolute blindsight in normal observers can be accounted for by response bias. In our replication of Lau and Passingham's experiment, the relative blindsight effect was abolished when performance was assessed by means of a bias-free 2AFC task or when the criterion for awareness was varied. Furthermore, there was no evidence of either relative or absolute blindsight when both performance and awareness were assessed with bias-free measures derived from confidence ratings using signal detection theory. This suggests that both relative and absolute blindsight in normal observers amount to no more than variations in response bias in the assessment of performance and awareness. Consideration of the properties of psychometric functions reveals a number of ways in which relative and absolute blindsight could arise trivially and elucidates a basis for the distinction between Type 1 and Type 2 blindsight.

Create your website for free! This website was made with Webnode. Create your own for free today! Get started