Can AI Help Your Dyslexic Child Learn to Read? ft. Dr. Coral Hoh

So they had been struggling for a while. And they were able to get from under the 25th percentile. Some of them were in the 10th percentile. Some of the participants in the clinical trial were at the 1st percentile. And to be able to get to the 50th percentile in one school year is, you know, you know that there's something different about this program. They got me diagnosed when I was in third grade. And it was the most relieving day of my life because I finally had, like, kind of a place to hang my hat and say, see, I told you, I'm not lazy. Definitely not stupid. I just learned differently. Like, that's totally okay.

We were dyslexics. We were often, you know, just set up for failure because, you know, we were given books and tasks that really are not appropriate for the level that we were at. So how do you expect anyone to be open to learning when their entire relationship with reading, with learning, is physical pain? I wish I had understood when I was younger that the fact that my brain worked differently than people meant that I would be good at things in ways that other kids were not going to be good at.

I just assumed, I can't be dyslexic because letters don't jump up and down for me. Dyslexia is not a disorder. It is not a disease. Once you sort of own it and embrace it, then you can have the power to deal with it. It's sort of empowering and also enriching to the whole community. To those who don't understand what dyslexia is, that it's not a negative, it's not a problem. It's, like we keep saying, superpower.

Hey, everyone. We are Sonja and Nick, and this is Dyslexia Journey. And today, we are welcoming Dr. Coral P.S. Ho to the show. Coral P.S. Ho, PhD, is a clinical linguist with a PhD in linguistics. For the past 30 years, she has worked with children and adults with dyslexia, learning disabilities, and other language-related disorders. Dr. Ho, welcome to Dyslexia Journey. Thank you for having me. So, we thought we'd just start with what you have developed, which is Dysolve, is the name of it.

And I just wanted to let you, Dysolve AI, like tell us just about that to start with, and then we'll get into, you know, how you came to it. Yes. And I'm sure many people have tried, you know, EdTech programs, so then they're used to game-based programs, including in schools. But Dysolve is different because the games are not there waiting for the child to play them. So, we have this patented, very sophisticated, advanced technology to generate the game specifically for each child for each session in real time.

And Dysolve is an expert system. So, it's the first expert system in this area to do this kind of evaluation and what we call corrective training for language-related difficulties.

So, it sounded like, and full disclosure, we haven't had a chance to try it out ourselves, but it sounds like it does screening as well as practice. Yes, yes. So, it's the first product that integrates screening with training or intervention. Most schools will call that intervention. We integrate the two because it doesn't make sense to evaluate the child, find out that there are some problems, and then just let the data go.

And also, you get much more accurate, precise evaluation if the data are going to be used for some purpose. And in our case, it's actually to resolve the difficulty. So, I want people to understand that for the first time, we can resolve dyslexia. And in 2024, a number of our students who had graduated from the program came forward in the press to talk about the end of the dyslexia. So, it is correctable.

Wow. That's amazing. It's wonderful to have the actual case studies. Yeah. Yeah. And so, how widely is it used currently? It is used in all 50 states. So, we started with having private subscribers, families looking for us and joining in. But now, we have schools. And I also want to mention that these are not just case studies. We have completed a very rigorous clinical trial. And the results were released last October, just so recently.

And this is the first program that has that kind of very rigorous evidence of positive effect, that if you take care of language processing, it's the only program that does that, then reading improves. And the measures we use for reading are the highest bar, which is to use state testing, you know, where you're looking at very global, broad reading achievement. And also, the quarterly assessments that schools run. So, the clinical trial was in nine states.

And, you know, we took, on average, the lowest performing students at the 10th percentile. And they showed that with Dissolve, the reading gains were 33% larger against any other method out there. Wow. In just three weeks of training. Oh, wow. That quickly. That surprised me when you said the half price. And that's what we have seen all along because this is a very rigorously tested program that we didn't actually announce for a long time because we were building it out.

But right from the beginning, with the beta testers, they were able to get, and most of them were in middle school. So, they had been struggling for a while. And they were able to get from under the 25th percentile. Some of them were in the 10th percentile. Some of the participants in the clinical trial were at the first percentile. And to be able to get to the 50th percentile in one school year is, you know, you know that there's something different about this program.

And also, by the way, people talk about evidence-based, you know, methods. The term is actually very loose because it doesn't have to mean positive effect. But for ours, which is run independently by a university team, we're able to get positive effect. Wow. That's great. Let's, can we dig into the pedagogy a little bit of Dissolve AI? Could you explain how it's worked, how it works? Yes. We can't even talk about pedagogy, you see.

That's the thing. You don't start off by taking care of a reading difficulty with reading. You see, the problem is in language processing. And I'm sure, you know, your audience can appreciate that their children are very smart. All right. But they have this issue with not being able to learn to read. So what's the problem? It's a language processing issue. So what the program does, Dissolve does, is actually tackle that at that level of brain processing.

So the first time we are able to do that because of technology. But all other programs and methods can only work on skills. So think about three tiers, right? And so the foundational tier where Dissolve comes in is brain processing. And brain processing supports the acquisition of skills. The reason why the other methods are considered compensatory methods is because they cannot correct that language processing problem. Now you can. When you do that, then very amazing results happen and in a very short time.

So what it also means for the families is that, gee, you know, actually, it's not that big a problem. But, you know, if you don't take care of it, you know, the child will struggle. That individual will struggle for the rest of their lives. So, you know, it's a shame if you don't correct it, especially for children. Yeah, for sure. What, could you give maybe a specific example of what you mean by the brain processing level?

Yes. Like, for example, segmenting words, right? Segmenting words into single sounds, right? And knowing what sounds they are and what sequence, because you need to have that for decoding and spelling, right? A lot of the children who come to us have this kind of problem, right? And so you want to get them to recognize the individual segments, the sound segments, and then retain the order in which they come. Right, right, right.

Right? But, you know, that's already a lot of things going on, right? Yeah. Just to remember what you heard and then, you know, knowing what sequence they come in. So there are already a lot of processes involved in just that one task that people would take to be, oh, it's not a simple task if you can do it. If you can't do it, it is really quite a struggle because it's really already very complex.

So that's what we mean in Dissolve, that when we talk about language processing, it's really to break things down to a very granular level and locate exactly where the problem is for that one child. Yeah. It sounds like a whole task. Something is preventing the child from doing that task, right? So it is possible to correct these areas, and that's why I say that the games are not there, right, because you need millions and millions of such games to hit all the little, little points of difficulty, right?

And that's why the technology now is there to generate it as needed for whatever the problem is. Yeah, so the individualization is really key, it sounds like, yeah. Right, right. And not being able to individualize at this level of detail, that was the problem before. And also not having the expertise because you've got to look at the whole spectrum, right? I'm pretty sure you've heard that, right, from talking to so many families that every child is different.

So you have to consider the broad individual variation across the spectrum. And then, you know, a lot of people don't get diagnosed with dyslexia. So do they have it or they don't? Well, I would say, forget about the title. You just have to take care of whatever the problem is, you know, and most of the time it is language processing. Yes, well, and I'm thinking from the perspective, you know, of the child, like you were saying that, you know, before, if you have just the games there, when do you hit upon it?

I mean, that's been a lot of the issue, too, is that it's not that they're not willing to do some work, but sometimes it's so much to try to ameliorate it. So if it can be so much more targeted, I mean, that's also just so much more focused and so much less time. So let me, you know, actually, that's why some Orton-Gillingham teachers, you know, that's a method that's used a lot, right?

So compensatory. And so the teachers are still working for the, with the students, but have them sign up for Dissolve. And one teacher said it this way, you know, I may hit the problem incidentally, but Dissolve will do it thoroughly, right? Dissolve's not going to miss it. And Dissolve's not going to forget what the child did every single time, right? And retain it and cross-reference it to see, because it's a dynamic program.

So the program is built dynamically for the child. And we call this responsive intelligence, to be responsively intelligent and work with that child one-on-one. And can even young children use Dissolve AI independently, or do they need an adult helping them? Well, you know, the current version really works well with kindergarteners and up. We have adults in two, right? All the way to the 80s. Can you believe it? Yes. Adults don't want to take care of it, right?

So, and we say that only because, you know, young children may have trouble, right, just doing it. But otherwise, really, the families using Dissolve at home, they don't really have to supervise their children. They just have to remind them to do it, because everything is explained, right? Like, if they don't know how to do a certain game, they just hit the how to play button, they see the demo, and they do it.

And when the child logs in, right, with the user ID and a password, you can make it a simple password, then the child sees a game lobby, a game lobby with, you know, different game icons, and only one is activated at any one time. So, there's no decision needed, right, for the child to figure out what to do next. Just do that, click on it, play. So, it's game after game after game.

So, the whole thing is games. And each child will go through thousands of them. Okay. Yeah. And how many hours, when you said earlier that it showed positive effects after only three weeks, how many hours approximately were the children using it during those three weeks? Yes. So, it works out to be about one hour a week. During evaluation, we recommend 15 minutes per session per day. And after that, work up to 30 minutes a day.

Okay. So, it is good to use it pretty frequently, it sounds like. Oh, yes. The day of school, yes. Okay. Yes. Yeah. And, you know, before we jump into your background, I think a lot of our parents listening are going to be wondering at this point how they can get this for their kid. Does it have to be through schools or do you still offer individual subscriptions? All they have to do is go to dissolve.com.

D-Y-S-O-L-V-E dot com and sign up. Okay. Or contact, you know, the support team, right? If they have questions. So, individuals can do it, right? Anyone can sign up. But, you know, if they want this to really benefit more than just themselves, but, you know, other people at the school, then ask the school to reach out to us. After all, schools are required, as you know, right, to give appropriate education, right, for all children with disabilities as well, right?

And so, this one has gone through a very rigorous clinical trial to show that positive effect is there against other methods. Yes. Yes. That's great. I think people will be very excited about this. I know I am already, so that's very, that's wonderful. So, let's, yeah, let's dig into your background a little more because I'm not, I'm not myself particularly familiar with your field, you know, of linguistics. But I'm curious just sort of how you got into it and a little bit about what it entails.

I got into dyslexia really because, it's because of my PhD in linguistics. So, after I got my degree, I wanted to do something practical with it, help people. So, I started working with people with all sorts of language difficulties. And then, soon enough, you know, my friend told me that her son was struggling with reading at school in second grade. And I asked her what she was going to do about it, and she said, I'm going to have him evaluate it.

And I asked her, what does that mean, you know, who's going to do it, you know, how much does it cost? And she said, $2,000. I said, $2,000? You know, how long is he going to do it, right? And what qualification does this person have? And so, I said, why don't you just let me do it, try it out for free first, right? Let's just test, you know, I designed the test.

Let's just see what we can find out for free. And I was able to find out where his processing problems were. But at that point, you know, I really didn't know how to resolve it. So, I started reading the literature, and then I realized, oh, this is what they call dyslexia. And the more I read up on it, I realized that nobody knew how to correct it. And so, that's what I started working on, because then I saw more and more of them, because, you know, the size of the problem, right?

20% of the population at least. But because I'm looking at language processing, it is a much larger size of the population. Especially since now, if you look at the nation's report card in many states, you find that 70% of students are not proficient in reading. So, it's a very big problem. And then I realized that we've got to go to AI. And I was already looking at AI in graduate school, so a long time ago.

And because this is the only way to solve a problem that is that complex. So, my background is important in the sense that then, you know, we can see that the problem is not as simple as what people are making it seem. That, oh, you take care of just this area of language, then it's fine. No, it is a lot more complicated. And that's what we see in the data. And there are lots and lots of surprises in there.

That's what, right? The complexity of the linguistic system in the brain. And then the second one is because, you know, you need technology and computing to take care of the speed of response, right? To capture what the child is doing and then calibrate it very, very finely and also doing it in a game activity that has to work super, super fast. All right? And so, that's the other problem. That's why humans cannot do it.

The third one is, of course, you know, the size of the problem, the capacity that's needed. And also to cross-reference, even for one child, you're talking about billions of data points, really. That's what you end up doing, right? To have a corrective program.

Yes. So, I'm curious about, you said that you were thinking about this problem for quite a while, then it sounds like, because you said that the, you know, you thought of the AI quite a while ago. So, I'm curious, like, how long has it evolved over? Well, you know, I've been in the field for 30 years, you know, doing my practice. And 10 years we took to develop Dissolve. But right from the beginning, even with the prototypes, it worked.

And that's why we have a book out. And the book is titled Dyslexia Dissolved. And it's about the successful cases that we have, 10 cases in there, all with happy endings, because the children, they were mainly in middle school when they came to us, you know, really struggling. And many of them have multiple disabilities. And then, you know, after that, after Dissolve, it didn't really take them that long. One to two years, they were able to get to honors when they got to high school, high honors.

And some of them had graduated. Some of them came back and recorded messages for younger children to, you know, go for it, you know, but stay, stay with it, right? Don't quit, you know, and then you can get out of it just like what happened to them. That's bringing up in my mind, because I know the results came so quickly in the three weeks, but it sounds like, you know, for quite a few years, then someone would continue using Dysolve?

Is that accurate? Actually, you know, it has to do with, you know, how regularly they do it, right? And the reason why we can see results in a large group that quickly is because for some of them, it may not take that much to clear those processing issues, right? Because Dissolve works very fast in identifying them. So the evaluation part is actually two weeks, two weeks of very comprehensive, you know, exploration of where the problems are.

So, but then for others, it will take longer. And I say two years because a lot of them come with co-occurring conditions, all right, and including attentional issues, all right? So they cannot really stay on a session for 15 minutes, 30 minutes, right? Fortunately, each game is short, you know, it could be at most just a few minutes, right? Usually shorter, right? Because these are just, you know, very quick probes, you know, at a very granular level.

That's what I like about that, though, is that it sounds like it still will be helpful, even if someone does have, let's say, an intentional issue. It'll take them longer, you know, in the course of maybe years, but the progress, I mean, it's still possible to progress. So there's still hope for people in that case, too. Let me just mention, right, some of the cases, right, because people like cases, right? Yeah, it's great to hear it.

So, yes, at one of the schools, the teacher noticed changes in six weeks, six weeks. And this was a fifth grader. So the fifth grader had been struggling for a while, got enough attention from the school, you know, all the supports needed. And the administrator said that they used the programs that are available twice with high fidelity, and it still didn't work. And then six weeks into dissolve, then the other teachers, and that's the thing that we don't see.

We only see the game responses, right? But it's the teacher's parents who tell us that the child had changed, right? The other teachers reported that now the child volunteered to read in class, right, which is like a big deal because they try to hide. They don't want to do that, right? And then being able to retain the words in science, right, because these are, you know, technical terms, longer, right, multisyllabic, so harder to retain.

But they're able to do that. So you see the changes happening. And that's what we're studying right now because it is like a hockey stick pattern, right? They will struggle for a while because, you know, I'll be honest, right, because we're trying to push them past the current limits if they have processing issues. So they will struggle for a while, and the scores will fluctuate. But at some point, the brain clicks, and there's a very sharp upward turn for a lot of the things that they do, not just reading alone.

Like, you know, being able to do homework faster without help, volunteering to read books, which they, you know, they stayed away from before, not fighting the parent or the teacher that much, you know, being more accommodating in doing the work because it's a lot easier now. Definitely. That sounds amazing. Are there, are you confident that that Dissolve AI can help everyone with dyslexia, or are there certain variants of dyslexia that maybe aren't appropriate, or what would you say to that?

So far, so good, right? As long as they stay with the program. So the issue is compliance, right? Compliance, right? Willing to spend the time to do it, which is not like a lot of time per day, but just doing it. And when things get hard, because they will, right? So think of a 10-year-old, 10-year-old's brain has been operating in one way, right? And Dissolve is going in to change that. That's going to involve some amount of stretching and struggling, right?

So the child has to persist. So Nick, you asked me, right? What message do I want to give to the children? I say this, right? Now that there is a solution, you do it for yourself, right? This is for your life, right? But you have to persevere. Just like you do in sports or dance, right? Or something that takes some effort, all right? And consistent effort. You do it. You do it for yourself because things will get so much easier.

And that's why our graduates come back and record those messages for the younger children. They want the children to recognize this and understand this. Yes. I'm wondering if you have a message for the general public as well about dyslexia. It is a huge issue, right? Because of the number of people affected. And it also is an economic problem, right? If you use methods that just keep the children in special ed, and because they are so labor-intensive, right?

And there are not enough specialists around, then you lock out so many children. Children who also need help. So you have to figure out the clinical, logistical, and economic problem with this one. And if you recognize it, especially for school superintendents, then you really have to check out Dissolve, right? And also for everyone, Dissolve is also a resource for all this information, new information that we're finding out about language processing. So parents can go to the resources section at Dissolve.com, D-Y-S-O-L-B-E.com.

And we talk about how do you evaluate screeners? How do you evaluate evaluations, right? Before you pay thousands of dollars for it. How do you evaluate the program? So we talk about what it means if somebody says that their program is evidence-based. Is that good enough? No. What about structured literacy? Is it enough? No. You have to consider exceptional populations and not just a typical population. What else do you need, right?

So it's a lot of education, regardless of whether they sign up or not, right? We're trying to educate everyone, right? About the findings that we are getting, right? Now that we have the means to really, in a way, peer into students' brains for those who are struggling and understand why they're struggling. And then also to have expectations as to when will they actually improve.

Yes. Well, so that's great. So it sounds like, of course, to check out Dissolve AI itself, people can go to your website. But also, there's some other resources in terms of how to think about all of this and when you're having your child evaluated. Right. And for the general public, right? Because they do care about school taxes, too, right? Even if they don't have a child with dyslexia, Dissolve is only 10% of what schools are paying per child, right?

And you can use the resources so much more effectively to benefit everyone. Yeah. Well, as we wrap up here, anything you'd like to add that we didn't get to?

I'm just looking at some things I wanted to cover. Yeah, no problem. The biggest impact that we can make is through the public school system because so many children are under that. So I would like people who are in that system to check us out, right? You don't have to believe right away what we're saying, but we would ask that you do the research yourself, right? And that's the thing, right? 40 years of reading research show us that no program had been able to, before Dissolve, to produce, you know, in a very rigorous trial, what we call a randomized controlled trial, similar to medical clinical experiments to show that, you know, you can get the effect, right?

And Dissolve went through that stress test during COVID to take the lowest performing students also beyond third grade, right? Because it's easy to show some evidence before third grade. The challenge is to do that for all the students who have been struggling for a long time, to show that you can still help them. That's important, right? So check us out. Great. Well, this was so exciting to hear about, and I hope everyone does check it out.

Thank you so much for speaking with us today and getting the word out more. We definitely want to get the word out more about this, too. Yeah, and thanks for all your hard work on dyslexia over the years. Thank you. And, you know, yes, there's hope, right? And so it's about spreading that hope. Yes. Yes, absolutely. Yes. Thank you.