Research shows why not everyone learns from their mistakes

Some people do not learn from their mistakes because of the way their brain works, according to research led by an academic at Goldsmiths, University of London.

The research, led by Professor Joydeep Bhattacharya in the Department of Psychology at Goldsmiths, examined what it is about the brain that defines someone as a ‘good learner’ from those who do not learn from their mistakes.

Professor Bhattacharya said: “We are always told how important it is to learn from our errors, our experiences, but is this true? If so, then why do we all not learn from our experiences in the same way? It seems some people rarely do, even when they were informed of their errors in repeated attempts.

"This study presents a first tantalising insight into how our brain processes the performance feedback and what it does with this information, whether to learn from it or to brush it aside."

The study, published in a recent issue of the Journal of Neuroscience, investigated brainwave patterns of 36 healthy human volunteers performing a simple time estimation task. Researchers asked the participants to estimate a time interval of 1.7 seconds and provided feedback on their errors. The participants were then measured to see whether they incorporated the feedback to improve their future performances.

'Good learners', who were successful in incorporating the feedback information in adjusting their future performance, presented increased brain responses as fast as 200 milliseconds after the feedback on their performance was presented on a computer screen.

This brain response was weaker in the poor learners who did not learn the task well and who showed decreased responses to their performance errors. The researchers further found that the good learners showed increased communication between brain areas involved with performance monitoring and sensorimotor processes.

Caroline Di Bernardi Luft, one of the research paper’s co-authors from the Federal University of Santa Catarina, commented: “Good learners used the feedback not only to check their past performance, but also to adjust their next performance accordingly.”

The brain responses correlated highly with how well the volunteers learned this simple task over the course of the experiment, and how good they were at maintaining the learned skill without any guiding feedback.

"Though these results are very encouraging in establishing a correlation between brains responses and learning performance, future studies are needed to identify a causal role of these effects," Professor Bhattacharya added.

Homer prevents stress-induced cognitive deficits

Before examinations and in critical situations, we need to be particularly receptive and capable of learning. However, acute exam stress and stage fright causes learning blockades and reduced memory function. Scientists from the Max Planck Institute of Psychiatry in Munich have now discovered a mechanism responsible for these cognitive deficits, which functions independently of stress hormones. In animal studies, the researchers show that social stress reduces the volume of Homer-1 in the hippocampus – a region of the brain that plays a central role in learning. This specific protein deficiency leads to altered neuronal activity followed by deterioration in the animals’ learning performance. In the experiments, it was possible to prevent the cognitive deficit by administering additional volumes of the protein to the mice. This suggests that Homer-1 could provide a key molecule for the development of drugs for the treatment of stress-induced cognitive deficits.

Klaus Wagner, a scientist at the Max Planck Institute of Psychiatry, studied the learning behaviour of mice that had been subjected to severe stress. He exposed the animals to social stress – a pressure also frequently experienced by humans today. A male mouse was placed in the cage of an aggressive member of the same species for five minutes. The latter tried to banish the “intruder” by attacking it. Unlike in nature, the test mouse was unable to flee from the cage and was under severe stress, as substantiated by measurements of the stress hormones in its blood.

Following a period of eight hours in which the animal was able to recover in its own cage, its behaviour was examined. While the mouse’s motivation, activity and sensory functions were not impaired at this time, it displayed clear deficits in its learning behaviour. A single five-minute situation of social stress was sufficient, therefore, to impair the animal’s learning performance hours later.

The researchers at the Max Planck Institute then tried to establish which mechanisms were responsible for these cognitive deficits. They identified the protein Homer-1, the concentration of which declines specifically in the hippocampus after exposure to stress. Through its interaction with the neuronal messenger substance glutamate and its receptors, Homer-1 modulates the communication in the neuronal synapses. When the volume of Homer-1 in the hippocampus falls after exposure to stress, the natural receptor activity is severely disrupted and learning capacity declines. The researchers were able to prevent this effect by increasing the Homer-1 concentration again.

Mathias Schmidt, Research Group Leader at the Max Planck Institute of Psychiatry interprets the results as follows: “With our study, we demonstrated the regulation of glutamate-mediated communication in the hippocampus, which directly controls learning behaviour. This mechanism functions independently of stress hormones for the most part. The molecule Homer-1 assumes a key role in this process and will hopefully provide new possibilities in future for targeted pharmaceutical intervention for the avoidance of cognitive deficits.”

Linking insulin to learning: Important insights in research with worms

Recent work by Harvard researchers demonstrates how the signaling pathway of insulin and insulinlike peptides plays a critical role in helping to regulate learning and memory.

The research, led by Yun Zhang, associate professor of organismic and evolutionary biology, is described in a Feb. 6 paper in Neuron.

“People think of insulin and diabetes, but many metabolic syndromes are associated with some types of cognitive defects and behavioral disorders, like depression or dementia,” Zhang said. “That suggests that insulin and insulinlike peptides may play an important role in neural function, but it’s been very difficult to nail down the underlying mechanism, because these peptides do not have to function through synapses that connect different neurons in the brain.”

To get at that mechanism, Zhang and colleagues turned to an organism whose genome and nervous system are well described and highly accessible by genetics: C. elegans.

Using genetic tools, researchers altered the transparent worms by removing their ability to create individual insulinlike compounds. These new “mutant” worms were then tested to see whether they would learn to avoid eating a particular type of bacteria that is known to infect the worms. Tests showed that although some worms did learn to steer clear of the bacteria, others didn’t — suggesting that removing a specific insulinlike compound halted the worms’ ability to learn.

Researchers were surprised to find, however, that it wasn’t just removing the molecules that could make the animals lose the ability to learn — some peptides were found to inhibit learning.

“We hadn’t predicted that we would find both positive and negative regulators from these peptides,” Zhang said. “Why does the animal need this bidirectional regulation of learning? One possibility is that learning depends on context. There are certain things you want to learn — for example, the worms in these experiments wanted to learn that they shouldn’t eat this type of infectious bacteria. That’s a positive regulation of the learning. But if they needed to eat, even if it is a bad food, to survive, they would need a way to suppress this type of learning.”

Even more surprising for Zhang and her colleagues was evidence that the various insulinlike molecules could regulate each other.

“Many animals, including humans, have multiple insulinlike molecules, and it appears that these molecules can act like a network,” she said. “Each of them may play a slightly different role in the nervous system, and they function together to coordinate the signaling related to learning and memory. By changing the way the molecules interact, the brain can fine-tune learning in a host of different ways.”

Blueprint for an artificial brain

Scientists have long been dreaming about building a computer that would work like a brain. This is because a brain is far more energy-saving than a computer, it can learn by itself, and it doesn’t need any programming. Privatdozent [senior lecturer] Dr. Andy Thomas from Bielefeld University’s Faculty of Physics is experimenting with memristors – electronic microcomponents that imitate natural nerves. Thomas and his colleagues proved that they could do this a year ago. They constructed a memristor that is capable of learning. Andy Thomas is now using his memristors as key components in a blueprint for an artificial brain. He will be presenting his results at the beginning of March in the print edition of the prestigious Journal of Physics published by the Institute of Physics in London.

Memristors are made of fine nanolayers and can be used to connect electric circuits. For several years now, the memristor has been considered to be the electronic equivalent of the synapse. Synapses are, so to speak, the bridges across which nerve cells (neurons) contact each other. Their connections increase in strength the more often they are used. Usually, one nerve cell is connected to other nerve cells across thousands of synapses.

Like synapses, memristors learn from earlier impulses. In their case, these are electrical impulses that (as yet) do not come from nerve cells but from the electric circuits to which they are connected. The amount of current a memristor allows to pass depends on how strong the current was that flowed through it in the past and how long it was exposed to it.

Andy Thomas explains that because of their similarity to synapses, memristors are particularly suitable for building an artificial brain – a new generation of computers. ‘They allow us to construct extremely energy-efficient and robust processors that are able to learn by themselves.’ Based on his own experiments and research findings from biology and physics, his article is the first to summarize which principles taken from nature need to be transferred to technological systems if such a neuromorphic (nerve like) computer is to function. Such principles are that memristors, just like synapses, have to ‘note’ earlier impulses, and that neurons react to an impulse only when it passes a certain threshold.

Thanks to these properties, synapses can be used to reconstruct the brain process responsible for learning, says Andy Thomas.

Scientists make older adults less forgetful in memory tests

Scientists at Baycrest Health Sciences’ Rotman Research Institute (RRI) and the University of Toronto’s Psychology Department have found compelling evidence that older adults can eliminate forgetfulness and perform as well as younger adults on memory tests.

Scientists used a distraction learning strategy to help older adults overcome age-related forgetting and boost their performance to that of younger adults. Distraction learning sounds like an oxymoron, but a growing body of science is showing that older brains are adept at processing irrelevant and relevant information in the environment, without conscious effort, to aid memory performance.

“Older brains may be be doing something very adaptive with distraction to compensate for weakening memory,” said Renée Biss, lead investigator and PhD student. “In our study we asked whether distraction can be used to foster memory-boosting rehearsal for older adults. The answer is yes!”

“To eliminate age-related forgetfulness across three consecutive memory experiments and help older adults perform like younger adults is dramatic and to our knowledge a totally unique finding,” said Lynn Hasher, senior scientist on the study and a leading authority in attention and inhibitory functioning in younger and older adults. “Poor regulation of attention by older adults may actually have some benefits for memory.”

The findings, published online in Psychological Science, ahead of print publication, have intriguing implications for designing learning strategies for the mature, older student and equipping senior-housing with relevant visual distraction cues throughout the living environment that would serve as rehearsal opportunities to remember things like an upcoming appointment or medications to take, even if the cues aren’t consciously paid attention to.

Memory appears susceptible to eradication of fear responses

Fear responses can only be erased when people learn something new while retrieving the fear memory. This is the conclusion of a study conducted by scientists from the University of Amsterdam (UvA) and published in the leading journal Science.

Researchers Dieuwke Sevenster MSc, Dr Tom Beckers and Prof. Merel Kindt have developed a method to determine whether an acquired fear response is susceptible to modification. By doing so, they have revealed the circumstances under which an acquired fear response can be eradicated. In order to measure whether a person actually learnt something new, the researchers used a measure for Prediction Error – in other words, the discrepancy between a person’s anticipation of what is going to happen and what actually happens.

No fear response

Cognitive Behavioural Therapy is currently the most common and effective type of treatment for people suffering from anxiety disorders. However, the effects are often short-lived and the fear returns in many patients. One major finding of Van Kindt’s research lab is that when participants were given propranolol, a beta blocker, while retrieving a specific fear memory, the acquired fear response was shown to be totally erased a day or month later. The researchers repeatedly found that the fear did not come back, despite the use of techniques specifically aimed to make it return. This indicates that the fear memory was either fully eradicated, or could no longer be accessed. One crucial finding was that while participants could still remember the association with the fear, that particular memory no longer triggered the former fear response.

Fear conditioning

For their study the researchers used a fear conditioning procedure in which a specific picture was followed by a nasty painful stimulus. While the participants viewed the pictures, the researchers measured the anticipation of the painful stimulus as well as the more autonomous fear response on the basis of the startle reflex.

The current findings will contribute to the further development of more effective and efficient therapies for patients suffering from excessive anxiety disorders, such as trauma victims. There was no independent measure to indicate whether the memory is susceptible to modification up until now. The researchers have shown that the fear response can be eradicated completely, provided that the person concerned actually learns something new while retrieving the fear memory.

(Image: iStock)

Research team discovers: brain does not process sensory information sufficiently

The reason why some people are worse at learning than others has been revealed by a research team from Berlin, Bochum, and Leipzig, operating within the framework of the Germany-wide network “Bernstein Focus State Dependencies of Learning”. They have discovered that the main problem is not that learning processes are inefficient per se, but that the brain insufficiently processes the information to be learned. The scientists trained the subjects’ sense of touch to be more sensitive. In subjects who responded well to the training, the EEG revealed characteristic changes in brain activity, more specifically in the alpha waves. These alpha waves show, among other things, how effectively the brain exploits the sensory information needed for learning. “An exciting question now is to what extent the alpha activity can be deliberately influenced with biofeedback”, says PD Dr. Hubert Dinse from the Neural Plasticity Lab of the Ruhr-Universität Bochum. “This could have enormous implications for therapy after brain injury or, quite generally, for the understanding of learning processes.” The research team from the Ruhr-Universität, the Humboldt Universität zu Berlin, Charité – Universitätsmedizin Berlin and the Max Planck Institute (MPI) for Human Cognitive and Brain Sciences reported their findings in the Journal of Neuroscience.

Learning without attention: passive training of the sense of touch

How well we learn depends on genetic aspects, the individual brain anatomy, and, not least, on attention. “In recent years we have established a procedure with which we trigger learning processes in people that do not require attention”, says Hubert Dinse. The researchers were, therefore, able to exclude attention as a factor. They repeatedly stimulated the participants’ sense of touch for 30 minutes by electrically stimulating the skin of the hand. Before and after this passive training, they tested the so-called “two-point discrimination threshold”, a measure of the sensitivity of touch. For this, they applied gentle pressure to the hand with two needles and determined the smallest distance between the needles at which the patient still perceived them as separate stimuli. On average, the passive training improved the discrimination threshold by twelve percent—but not in all of the 26 participants. Using EEG, the team studied why some people learned better than others.

Imaging the brain state using EEG: the alpha waves are decisive

The cooperation partners from Berlin and Leipzig, PD Dr. Petra Ritter, Dr. Frank Freyer, and Dr. Robert Becker recorded the subjects’ spontaneous EEG before and during passive training. They then identified the components of the brain activity related to improvement in the discrimination test. The alpha activity was decisive, i.e., the brain activity was in the frequency range 8 to 12 hertz. The higher the alpha activity before the passive training, the better the people learned. In addition, the more the alpha activity decreased during passive training, the more easily they learned. These effects occurred in the somatosensory cortex, that is, where the sense of touch is located in the brain.

Researchers seek new methods for therapy

“How the alpha rhythm manages to affect learning is something we investigate with computer models”, says PD Dr. Petra Ritter, Head of the Working Group “Brain Modes” at the MPI Leipzig and the Berlin Charité. “Only when we understand the complex information processing in the brain, can we intervene specifically in the processes to help disorders”, adds Petra Ritter. New therapies are the aim of the cooperation network, which Ritter coordinates, the international “Virtual Brain” project, which her team collaborates on, and the “Neural Plasticity Lab”, chaired by Hubert Dinse at the RUB.

Learning is dependent on access to sensory information

A high level of alpha activity counts as a marker of the readiness of the brain to exploit new incoming information. Conversely, a strong decrease of alpha activity during sensory stimulation counts as an indicator that the brain processes stimuli particularly efficiently. The results, therefore, suggest that perception-based learning is highly dependent on how accessible the sensory information is. The alpha activity, as a marker of constantly changing brain states, modulates this accessibility.

Finding the way to memory

Our ability to learn and form new memories is fully dependent on the brain’s ability to be plastic – that is to change and adapt according to new experiences and environments. A new study from the Montreal Neurological Institute – The Neuro, McGill University, reveals that DCC, the receptor for a crucial protein in the nervous system known as netrin, plays a key role in regulating the plasticity of nerve cell connections in the brain. The absence of DCC leads to the type of memory loss experienced by Dr. Brenda Milner’s famous subject HM.  Although HM’s memory loss resulted from the removal of an entire brain structure, this study shows that just removing DCC causes the same type of memory deficit. The finding published in this week’s issue of Cell Reports, extends Dr. Milner’s seminal finding to another level, revealing a key part of the molecular basis for learning and memory.

Although both netrin and DCC are essential for normal development (in terms of guiding nerve cell growth) until now their function in the adult brain was not known. Dr. Tim Kennedy, lead researcher and neuroscientist at The Neuro, contributed to the discovery of netrins as a young post-doctoral fellow. This new study reveals the answer to the question that drove him to first start a lab. “I remember that exact moment when I knew I could run a research lab, it was 1993 and I was studying the developing nervous system and I was amazed to spot netrins in the adult brain - raising the important question, ‘what are they doing there?’ 20 years of dedicated research later the answer provides an important piece of the puzzle for understanding our nervous system and neurological disorders.

“The power of this study is that it looks at the animal on all levels, molecular, structural, and behavioural. We show that the netrin receptor DCC is a critical component of synapses between neurons in the adult brain, and is required for synapses to function properly. To demonstrate this, we selectively removed DCC from a specific subset of neurons in the adult mouse brain. This results in progressive degeneration of synapses, leading to defects in synaptic plasticity and memory. The synapses continue to function in that they still communicate but, the synapses cannot adjust or change in response to new experiences. Therefore, you can’t learn anymore.”

Furthermore, DCC deletion from mature neurons results in changes in the shape of specialized protrusions called dendritic spines, and alters the NMDA receptor, a critical trigger for mechanisms that make changes in synaptic strength. Therefore the study reveals that DCC is required to maintain proper synapse morphology or shape, and to regulate the ability of the NMDA receptor to switch on, which ensures activity-dependent synaptic plasticity.

With Evolved Brains, Robots Creep Closer To Animal-Like Learning

The most nightmare-inducing characteristic of Big Dog, DARPA’s robotic military mule, might be the way it moves so stiffly, yet unrelentingly, over treacherous battleground. Turns out the repetitive mechanical gait that calls to mind some coming robopocalypse is also a huge headache for Big Dog’s makers—and lots of the big thinkers behind walking bots envisioned for everyday domestic use.

Units like Big Dog move so awkwardly because of their rudimentary brains, which require pre-programming for every little action. A four-legged walking bot could jump smoothly over rocks or weave through trees with the fluid grace and reflexes of a cheetah—if it only had a better brain. One that was more animal-like. Thanks to breakthroughs in understanding how biological brains evolve, a team of robotic researchers say they’re close.

“We are working on evolving brains that can be downloaded onto a robot, wake up, and begin exploring their environment to figure out how to accomplish the high-level objectives we give them (e.g. avoid getting damaged, find recharging stations, locate survivors, pick up trash, etc.),” says Jeffrey Clune, Assistant Professor of Computer Science at the University of Wyoming, who is part of the robotics team.

Continue reading