Last Tuesday, we had a speech by Ira Ovchinnikova, a researcher at the Laboratory for Interdisciplinary Research of Human Development at St. Petersburg State University, a research assistant at the University of Houston.
Most of her time at work, Ira explores how early childhood experiences affect language development and how this is reflected in brain activity, as well as studying developmental disorders.
We share with you the recording and transcript of the broadcast.
My name is Ira Ovchinnikova, I am a researcher at the Laboratory for Interdisciplinary Research of Human Development at St. Petersburg State University. I am also writing a dissertation at the University of Houston, so I am now in Houston, 9 hours ago from Moscow, and I am now somewhere in the middle of the day.
Q: is psychology a science?
This is a normal question that everyone asks. And students of the psychology department, and people who are engaged in psychology.
The answer is simple: everything that meets the criteria of scientific character and is labeled as science is science. In the 30s of the last century, Karl Popper introduced a criterion for the falsifiability of theories: that is, any theory can be considered scientific if it can be refuted by some facts, experiments on empirical material. The problem of psychology here is obvious: the theory will be associated with general concepts that are abstract, but the material world is here and now.
In order to meet this criterion, there is a concept of operationalization in psychology . This is when I take an abstract concept and come to a conventional opinion about how we will calculate this concept. Such an example would be the speed of response to certain stimuli.
In general, I am a cognitive psychologist (I will explain soon what this means). It can also be processing speed, response accuracy, many other operationalization concepts. So theories, thoughts, ideas that relate to psychology can be both scientific and not at all scientific. Non-scientific theories include classical psychoanalysis, for example: it simply cannot be refuted. Because, according to classical psychoanalysis, anything is possible. At the same time, from psychoanalysis for 60 years of the development of thought, theories have grown that are refutable. For example, Mary Ainsworth's attachment theory, where we are interested in how early infantile experiences affect human development.
Further, there is a cognitive direction in psychology. It can be considered the gold standard of science, it fully complies with the criteria of scientific character. The most developed are Alan Baddeley's theories of working memory; you can see many different theories about how the concept of working memory has changed, as empirical material, the results of experiments have allowed us, as a scientific direction, to improve this knowledge and make it more accurate. At the moment, the theory of working memory consists of several separate parts that have not yet found their refutation in experiments. That is, according to Popper, none of them will be true, but we say that we simply do not have enough material to refute it. And one more problem with any theory is the methodology of science as a whole. Only a new theory can replace the old theory. You can't just saythat the old one does not work, a new paradigm must be presented, an explanation for old phenomena and new paradoxes.
I am currently doing neuroimaging research in developmental disorders, although I started 12 years ago with concept formation research. I have always wondered how people form concepts, how they learn new things, and how mistakes occur in this story. For the last 4 years I have been doing neuroimaging, psychophysiology, and also doing MRI for people with developmental disorders - autism, dyslexia, a class of other disorders.
Q: what methods of neuroimaging exist, where are they applied, how do you use them?
Neuroimaging methods can be classified according to two scales. In general, these are all methods where we try to understand how the brain works as a biological substrate, how it functions and responds to different stimuli. All methods can be plotted on a graph with two axes: X is the temporal resolution (how exactly in time we capture information about the processes), and Y is the spatial resolution (how small the processes we can see).
For example, an MRI that you may be sent to check for problems with any organ is usually a structural MRI. That is, the radiologist or researcher will look at how the organ looks and is in space.
And, if this is an anatomical MRI, then I will not be able to look at how the same brain functions - only at what it consists of. Or I may be interested in, for example, diffusion tensor imaging (DTI) - how different parts of the brain are connected. And this method has a very low temporal resolution, almost zero - information is taken in one moment, there is no dynamics. But it has a high spatial resolution, I can see not only individual structures, but also the nuclear level, although I will not see individual layers.
EEG is the opposite situation. I put a hat on a person with high-resolution electrodes and receive information with millisecond precision. This is a very high temporal resolution for the human brain. But the spatial resolution will be very low because I will be guessing which parts of the brain are responding. We will have to use an average idea of how the brain is located in the cranium and which parts of it respond to signals.
Q: how to understand how the parts of the brain are connected? Pathways or anatomical structures that connect?
There is a special method for this - this is just DTI. In this situation, we are interested in how water reacts to a magnet, since you will have a change in the direction of water molecules. And in these areas we are building the paths of white matter. Here you need to understand that this is the part where a very large amount of programming takes place, because an MRI is not a camera.
I often use this metaphor when I tell students or colleagues about how MRI works: that is, there is no other way for me to know 100% what is happening in the body other than an autopsy. But I certainly want the subjects to be alive.
MRI allows us to predict with high accuracy how our paths look, but these are still models. Including when you go to the doctor, and he is trying to understand from your scan how your different parts of the brain are connected. Different methods have different accuracy, and over the years the quality of the methods increases - this is also part of my tasks. People with different expertise work here, and it will always be an interdisciplinary science, where people with education in physics, engineers, mathematics, and also, including people involved in cognitive psychology and developmental psychology, will build the most plausible model.
Q: but for all people, the vessels in the parts of the brain are equally located in space?
No, there are differences. I don't think I can say for sure now, but then I leave the links about it. In general, we always hope that everything is located in space in a similar way, extremely close, but there are deviations - and disorders, and such that do not in any way affect the quality of a person's life and the perception and processing of information.
Q: how do you use EEG in research?
You have to understand that none of the research in which I participate can be done by one person. This is a huge myth: science has long ceased to exist in the format of a “gray-haired man sitting in a white tower”; large teams always work. The projects in which I work are often carried out by 30-40 people. Always large samples, a lot of work on the search for subjects, data management.
For example, now in our laboratory in St. Petersburg there is a large project on the study of biobehavioral indicators in people with experience of institutionalization. By institutionalization, we mean people who lived in orphanages or children's homes.
Everyone has heard about orphanages, and children's homes are structures of an orphanage for children under 4 years old (after 4 years they are transferred to orphanages, boarding schools). We are interested in how such early experiences affect development - both in childhood and adolescence and adulthood. Our main focus is language development. There is a large hypothesis that institutionalization at an early age reduces the variety and quantity of linguistic information addressed to a particular child.
There are great social changes that are taking place in this area: when people who work with children - educators, nannies - are taught to interact more with children, to diversify their communication experience. But everyone understands that this is still not “their own” child, and no matter how well these people do their job - and I have met wonderful educators and nannies - a certain specificity of living and working conditions is still obtained. And we are interested in how this specificity, the absence of a close adult, affects the child. By a loved one, we mean the one on whom the child always relies, who is almost always available, with whom he forms a secure attachment and with whom he has the closest relationship. Usually, often, this is mom. This adult forms, gives the child linguistic information,helps to form his cognitive system. We consider this influence from the point of view of psychophysiology; I am interested in how children process language information differently (or in the same way), language stimuli in experiments, and whether these differences persist in adolescence and adulthood.
For example, there is such an experiment, presented in the oddball paradigm paradigm - I love it very much. The component that is presented there is called mismatch negativity. So, for example, for children, typically developing, the following is characteristic: up to a year of life (approximately), they can distinguish the sounds of different languages. So, for example, children will distinguish between the sounds "ta", "ha" and the sound that I cannot pronounce (the throat sound between "G" and "D", part of Hindi). With age, by the year of life, language statistics accumulate, and the child ceases to hear the difference between the sound of Hindi and the Russian language, he simply ceases to be special, different. It manifests itself as follows: you present audially to the child a repeating series of sounds "ha" (a lot), this becomes a line of no change. Further, when the child hears the sound "ta"which is different for him, on the brain activity you see a sharp change, a leap. This is the negativity of mismatch. Up to a year of life, regardless of whether you present a "ta" or a throat sound with a rare stimulus, you will see the negativity of the mismatch. If you conduct this experiment on typically developing children after a year of life (Russian-speaking, this is important), then you will not see this negativity of the mismatch to the sound from Hindi, or, in principle, to the sounds from non-native speech.If you conduct this experiment on typically developing children after a year of life (Russian-speaking, this is important), then you will not see this negativity of the mismatch to the sound from Hindi, or, in principle, to the sounds from non-native speech.If you conduct this experiment on typically developing children after a year of life (Russian-speaking, this is important), then you will not see this negativity of the mismatch to the sound from Hindi, or, in principle, to the sounds from non-native speech.
We wondered if the same memorization of sounds occurs in children living in children's homes, or whether the negativity of the mismatch remains at an older age. We ran an experiment. It turned out that phonological awareness - the ability to hear the difference between the sounds of Hindi and the native language - is also erased in children from children's homes. That is, they receive a sufficient amount of linguistic input in order to have information about their native language and speech - that is, they can accurately determine their native language.
At the same time, if we conduct experiments related to higher structures of information processing - for example, when children are asked to name objects, or when they are shown a picture of a boy and said that it is a flower - information is processed longer in children from children's homes. Therefore, we believe that there is some lagging behind in linguistic development. We see this also in behavioral techniques; we never work only with psychophysiology, we believe that it is not very informative. Working at the behavioral level also allows us to compare what is seen at the level of behavior and at the level of brain activity. Is there a completely identical interpretation, or will I see the differences, and I will assume in such a situation that there are sometimes no differences at the level of behavior, but at the level of brain activity there are, and this is not a completely smoothed difference.Although it can be smoothed over the years, depending on when a person got into the family, how long he lived in a child's home or orphanage.
Q: which of the interpretations of the word "consciousness" is relevant?
I am not concerned with the problems of consciousness. I understand that there is a wide variety of interpretations of consciousness, but now I use this concept at a fairly everyday level - as an opportunity for awareness.
Q: Have you seen the content of Victoria Stepanova? Can we consider her a psychologist? She believes that she can determine sexual orientation from a photo.
No, It is Immpossible. I don't know who Stepanova is.
By the way, I can talk about pseudopsychology and why it infuriates us too. Perhaps even stronger than people who do not study psychology. This is a big problem: when I go somewhere, I do not introduce myself as a psychologist. Because otherwise everyone will think that I will now talk about Vedic women, solve other people's problems, talk about how to live rightly and recommend a psychotherapist.
As for psychotherapists, I can recommend communities that I trust or people with whom I studied (or studied with). But all the other points are completely past me, I do not give such advice.
Pseudo-psychology annoys me because it is more difficult for me to deal with it. I find myself in a situation where I see a person with whom I was on good terms 10 years ago, and he suddenly went into pseudopsychology, and I understand - oh, it looks like we will no longer communicate. It's painful for me. People who only look at pseudopsychology from the outside can simply say: well, I just won't get to know him. This is the first thing.
Secondly, pseudopsychology is bad for scientific psychology. She automatically questions her scientific nature; "Is this science" is a normal question that everyone asks. In addition, it is necessary to talk about the research that has taken place in psychology from the very beginning. We have to start with Karl Popper's methodology; a lot more happened after him - there was Mark Poloni, for example, and a large group of British and Hungarian methodologists. And all this has to be told very quickly in order to then tell about just one study.
I think that all this in the same way can affect a large number of people who are responsible for funding various fields of science, especially in Russia, and who do not have the opportunity, time or desire to understand in detail how certain areas of knowledge work.
I wanted to tell you about my other two favorite projects besides the institutionalization project. The second large area that interests me in our laboratory is the project on the prevalence (prevalence) of autism spectrum disorders in the Russian Federation. In terms of how we calculate the prevalence, and how widespread each particular disease, disorder or other developmental feature is, we, as a government, will have to change the funding for each specific area. At the moment, there are no accurate statistics for Russia on how common ASDs are.
The laboratory where I work exists in close collaboration with the Exit Foundation, which helps families with people with autism spectrum disorders. Accordingly, we started a project in St. Petersburg, in the Primorsky District, to assess the prevalence of ASD. It took us three years and we continue to do it. This is a very large project. It is associated with the need, firstly, to develop an understanding of demographic statistics in general in a particular region. The team is large, multidisciplinary, employs people engaged in population research. For example, when you are evaluating a population study, you are building a model in which you try to understand how you can best represent the general population (these are all the people to whom you plan to transfer this result; for example, when we talk about a specific area,these are the children living in the area). You need to estimate the size and demographic diversity of the general population in order to understand where you can get quality information about these children.
Such research takes place in two stages. The first stage is the screening stage; that is, we work in a certain way with polyclinics, we interview as many people as possible using a short questionnaire. In it, they answer the peculiarities of their child's development, answering yes / no questions. There is the concept of "red flags" - features that are typical for children with ASD or for developmental disorders in general. When a person picks up a certain number of "red flags", we contact the family and invite them to the second stage. In addition, we try to find children with a minimum of "flags" - potentially neurotypical, and we also invite them to the laboratory. In the second step, we do a full development assessment including the battery associated with the ASD assessment. Thus, we determine the prevalence of ASD in a particular region,and this will allow us to create a design study that can be fully carried out in Russia.
For this I love working in my laboratory. It not only gives me the opportunity to do different projects, but also makes me feel that I am doing something socially useful.
The third project that I wanted to talk about is a project about language development disorder in an isolated population in the far north of Russia. This project is interesting to me because by “isolated population” we mean a population of only 800 people. This particular population is known for this disorder. We assess how the disorder of language development is transmitted in them, and see how this is represented at different generational levels. We collect stories from grandmothers, mothers, all close and distant relatives in order to assess their linguistic development, we conduct a full range of assessments of linguistic development in children, we look at how connections occur within one family.
Q: where to start to learn how to help your child's development?
I cannot answer this question, it is too individual. In addition, it is absolutely not necessary to do a development boost. I believe that people who visit Habr, read, write on Habr are a very developed audience; keep just playing and talking with your kids.
I am not an individual counselor. I'm not a clinical psychologist, that's important. I have the right to work with clinical populations, but I do not have the right to consult (and never intended to).
Q: what research can be done on false memories?
Check out Elizabeth Loftus, which is awesome.
For everyone: false memories are an event that you think happened to you, you are sure that you remember it, but in fact it did not happen. Elizabeth Loftus is a psychologist, memory researcher, the person who has invested the most in the development of research on false memories; it is very interesting to read about it, even if you are not interested in psychology.
The story with false memories looks like this: when you find yourself, for example, a witness to an accident, the way you are asked a question will influence which answer and how you give. If you ask different people how fast the car RACE and with what speed the car RIDED before the accident, then, statistically, those people who were asked with the word “rushed” will give higher speed ratings than the second group.
Elizabeth Loftus worked a lot (and works, I think) with the system of prisoners and people in pre-trial detention and this influenced the change in interrogation protocols in the United States.
Q: are false memories deja vu?
False memories are when you think that something happened to you some time ago, and you are sure of it. For example, the fact that you were at Disneyland. And your relatives say - we weren't at Disneyland when you were a child, I'm sorry. There may be some more extensive memory. And deja vu is when it seems to you that you knew about an event that just happened in advance. Deja vu and false memories are believed to have common roots, but I cannot say more about deja vu.
Q: is the detail of a dream an example of false memories?
After all, as a rule, we dream of blurry images, and already the clarity and consistency of the events that occurred in the dream we think out, retelling the dream.
I’m afraid to lie, so I’ll say I don’t know. I have my own ideas, but this cannot be considered expert knowledge.
I also wanted to talk about “what to do if I have a physics, mathematics, CS education and I want to work in neuroscience or cognitive science”. Firstly, it's cool, come to us, we have cookies. Second, read the literature before doing this. I am very pleased with people who are entering cognitive neuroscience with knowledge from other areas; I, for example, lack a person in the team with an engineering education, or with knowledge of classical mathematics, graph theory.
I have a project on brain connectivity in children with institutionalization experience, which is a project that I am doing this semester. I'm trying to take different connectivity (connectivity) metrics and write a script that will allow me to process the same set of data using different metrics and get different results, and then write to this each time a separate interpretation and a separate idea of which a conclusion can be made. And do the same with the data of adults with institutionalization experience. The problem is that I write in Arc, but I write poorly in Python and I do not know how to work with Matlab at all, which would be very useful. It would be nice for me to have someone with whom to discuss different metrics. I am starting now with graph theory, using metrics from graph theory, but a wide variety of scoring systems remains,for example - dynamic systems. That would be very helpful.
There is one problem that happens to people coming from areas that are considered more natural science. It's like in that picture from xkcd, where different scientific fields are ranked in terms of their accuracy, and the mathematician says: "You can't see you from here at all." People who come to cognitive science or neuroscience from more precise fields believe that now they will take the method they used before - for example, machine learning - and apply it, and then they will tell me how the whole field of knowledge works. It doesn't work. Because in order for you to interpret ML results (including), or to make a more accurate model, you need to understand what is happening inside. And for this you need to get at least part of the education that people receive for 10 years.
This is not a hit. I really encourage you to come to us and try something if you're interested. I will ask you to leave my contacts under the video. Maybe you can join one of our projects and even get to us on a permanent basis; I hope that in the near future we will have the position of an engineer. But this classic problem remains.
I have a good friend who is writing a dissertation on using ML with MRI data. He is a physicist himself. And about once a month he says to me: some incomprehensible data, what's going on here? And I retell him a piece of cognitive neuroscience. Of course, I understand that in order to have enough knowledge on this topic, it takes a lot of time, and he simply did not have this time. Therefore, if you want to work in neuro- or cognitive science and want to come to a project - write to the laboratories that do this. You will definitely find a project that will suit you, and in which it will be pleasant and interesting to communicate, and you will be able to adequately hear.
By the way, there is a great funny story about "what to do if I have a physics, mathematics, CS education and I want to work in neuroscience." Elon Musk some time ago announced a neuro-related company to build artificial intelligence. And he made an announcement on Twitter: come everybody, you only need prior knowledge in engineering or programming, and you need to have no prior knowledge in the field of cognitive psychology, neuro or something similar. Everything would be fine, but after a while they made a presentation of this company, and in the picture describing the brain they confused the left and right parts (this is easy to do if you have never looked at the brain and have no idea how it works). There is such a joke now. Although I really respect Elon Musk.
Q: does a cat have consciousness?
I will now take advantage of this moment. I have a wonderful experience where I have a great opportunity to discuss the existence of consciousness, behavior and other things in animals. My wonderful friends have a project called "Bobbin" - a scientific and educational project in which they talk about interaction with dogs, about what aspects of life and behavior of dogs exist. They post cool links to research. There you can read about consciousness in dogs - this is a similar question, and the answer will also be similar.
