Students with Disabilities Learning Online: Vulnerable Students in a Rapidly Evolving and Unstable Environment

In this webinar, Students with Disabilities Learning Online: Vulnerable Students in a Rapidly Evolving and Unstable Environment, David Rose, Ph.D., co-founder of CAST, explains the 10 major issues in online learning for students with disabilities, including online accessibility issues.
 

 

Resources developed by the CTD SEA and LEA Technical Assistance Team, led by American Institutes for Research, in collaboration with David Rose, Ph.D., at CAST.

 

Transcript: 

- [Voiceover] Our mandate is to work with the field and all of the great individuals, stakeholders, parents, teachers, policymakers, researchers, and others who are working with students with disabilities. Our goal is to provide up-to-date resources and materials and technical assistance to help inform and guide the work of the field. So again, it gives me great pleasure to introduce Dr. Rose. He is familiar to many of you if not all of you on the phone as the co-founder of CAST, which was started in 1984 under David's direction and vision to create a non-profit center that was focused on both the research and to better understand the application of technology to work with a whole variety of learners and their needs. As a result of the work of CAST, the field of universal design for learning, or UDL, has really become a cornerstone of education, both educational policy and also education and teaching and learning, so that teachers around the country are working to integrate the important principles of UDL in their teaching. David for the last almost thirty years, who's counting, has also been a professor at the Harvard School of Education where he has touched the lives of thousands of students over the last three decades. And I must say, of all of David's enormous accomplishments in terms of his publications, his books, and other articles, one of the things that I think exemplifies his outstanding work is that he was selected as one of the Daring Dozen by Edutopia magazine in terms of highlighting his important contributions to the field of education, psychology, and teaching and learning. So on that note, it gives me great pleasure to hand the slides over to David who's going to walk us through the critical issues facing students with special needs as they venture into this whole world of online learning. David, take it away.

- [Voiceover] Hi, thanks very much, Tracy, and I look forward to spending a little time talking about this topic that has occupied a number of us here at the Center. So let me just show you who's involved in the Center on Online Learning and Students with Disabilities. I think most of you know that Don Deshler has just recently retired, but he was the principal PI at the beginning of this Center. CAST is involved and there are the names of some of the people here, and NASDSE, National Association of State Directors of Special Education, is the third partner and we've been working together now for three-ish years, and this, as the first slide said, a rapidly changing world. In fact, I wanted to start there. I think some of you may be familiar. Oh, by the way, there'll be... Elise is gonna monitor the chat so if people have questions, things that I'm not clear about or I should say more or less about, feel free to type something in. I'm not gonna monitor it visually but Elise will and she'll interrupt me if there's something that we should all talk about and I'd be happy to be interrupted. It's hard as you know, all of you know I'm sure, to speak into this faceless medium so a little interruption would be a good thing. So some of you are familiar with Stephen Jay Gould probably, a very prominent paleontologist that taught evolutionary biology at Harvard and many important books and so on and I pilated two of them here 'cause they're so different. One is his textbooky tome on the structure of evolutionary theory, he could do that, and another one of his more popular books, The Triumphant Tragedy in Mudville, was about baseball. He was not only a legendary paleontologist, but a legendary baseball fan, and he wrote articles about how baseball evolved, that are among the most, what should I say? Certainly the most evolutionary biological look at baseball. Anyway, all of his work emphasizes one aspect of the evolution of everything which is that at the beginning of things, things are highly variable, lot of chaotic variability in anything, and you have much more extremes. As anything grows and develops, it stabilizes, and has smaller deviations, smaller variability. The means make... It's easier to describe things with their means. But so baseball, the one article he's famous for is talking about why there'll never be another 400 hitter in baseball and it's because baseball has evolved such that now, what used to be highly variable in terms of people's skills meant that there were people who were batting 100 and people who were batting 400 in the same league. And now that almost never occurs and certainly not at the professional level. Everybody pretty much bats at about 275 now and a few outliers but outliers only go into the low 300s even, because the pitchers are all good now and the hitters are all good and it stabilized around a much smaller range so his argument is we'll never see another 400 hitter. So the field of online learning is definitely a new field, and it is marked by great variability, enormous variability in every aspect, and when we first began to look at onl... Way was, "Oh my god, it is so highly variable "in what anybody even means by 'online learning' "what people do in online learning, "what kind of students are in online learning." Everything is variable. So that's the main thing that I want to say today and I'm gonna say it in a number of ways, a number of topics, ten specifically. Ten general findings that we have, all of which are notable for their huge variance now. Over time, this field will stabilize and things will become much more alike. It's like now, pretty much every car looks alike. I think you all have that experience. It used to be that cars were really different and now they all look just about the same, and I don't know how Cadillacs even differentiate themselves anymore. Okay, so ten topics, and the first is that in general, there's this large variance of everything and small central tendencies around what really happens. And other than the inclusion of some kind of online content delivery, there's really no consistently uniform procedures that constitute the practice of online learning so the closer you look at it, the less it looks like there's a thing there called online learning. There's a whole bunch of different things, much more like furniture, where there's all kinds of different things in it, than stoves, where there's considerable sameness about them. So here's some of the sources of variance. When you look at any online vendor, this would include blended or completely virtual, you'll see that the populations of students with disabilities is a huge variance, and we'll talk about it a little bit later in how many kids with disabilities will be in any instantiation. The roles that are played by teachers and parents will differ widely among the different varieties of online learning. The quality of the instructional design. Does it have anything that looks like evidence-based practices or not? Huge variation. From our standpoint, basic accessibility and universal design features, huge variation there as well. Some have near zero accessibility, which is a bad start, and some are really quite good. Just a few more to get this range. Whether students can have peer-to-peer relationships within the online environment, or do they interact much with teachers? It's highly variable. How teachers and administrators are trained, and what kind of resources they're given to get them ready. Extremely variable. The ways in which monitoring progress is accomplished or not accomplished, highly variable. How teachers, students, parents, anybody, gets feedback is incredibly different among different instantiations. And the requirements for what you need offline to support kids, particularly with disabilities, is another area where there's great great variation. I'm gonna elaborate on a few of these. I have a bunch of topics that I've got, these little semicolons. I think Tracy knows, I don't really know how to use semicolons, so it might be a different thing would be better there. Anyway, data rich, information poor. One of the things that's notable is that there's enormous amounts of data. Just striking compared to the traditional kinds of data in face-to-face classrooms, and the reality, however, is that the way the environment is now that the lack of communication and interoperability across various systems, we'll talk a little bit about it, currently makes the tracking that one would want to be able to do with that data, nearly impossible. So we have huge amounts of data, but very little information that's usable by a parent or teacher or student to help them either design or actualize a path through materials, so lots of data but not much information. This is just a... We worked with one provider to really try to get a feeling of how the information flow operates within the provider so that blue box is the actual provider, the person that, the vendor that you would recognize if I were telling you and it has all of these elements in it some of which they made, some of which they bought, some of which they licensed, all of them, blah blah, and everything from how did that course get in there to how does assessment happen to what kinds of third-party environments of all kinds happen and so on. But these things all don't necessarily talk well to each other, so there's not even good data flows within a lot of vendors. But they have all these moving pieces, and sometimes there are firewalls between these pieces and sometimes there's just interoperability problems between the pieces, so sometimes they really don't have any idea that these pieces don't talk to each other in the ways that would help students, and it's a problem not all vendors, but a lot of vendors have because they've been cobbled together. They're putting a lot of things together to make this work fast. It's a problem of, you know, the field being early. And here's just some summaries. The vendors are hobbled by this. They have lots of big data that they could potentially could use, but too little information from the schools about the individual students with disabilities, their strengths and weaknesses, their IEP goals, any of those things. Vendors typically don't get much of that information and you'd think they would, but as we'll see, privacy concerns actually erect barriers here so the vendors would like to do a better job and they could do a better job if they knew more about the students, but they actually don't get that information in a timely fashion often, so they're often feel like they're blinded, excuse the expression, by data that is sitting there close by but is not providing them with the information they need. They, I think most vendors would like to do a better job. They'd like to know more about students. Schools, the same thing. They could use this fabulous amount of data that the vendors are picking up. The vendors are often getting every click a kid does. It's kind of extraordinary. So they could know everything, not just whether the kid passes tests or stuff, but they can know what kind of help did they have or what are they communicating about. What'd they do after they made a mistake? What'd they do after they took a vacation? All of this stuff is available in the vendors' data. But it's not routinely shared with schools, and schools often can't get it, for some of the same reasons. Lots of concerns about, well who owns this data? Can a vendor really share it with schools and are the schools really ready to gather in this data and use it? Often they're not, 'cause it comes in a form that schools aren't ready to look at and it uses identifiers and things that make it complicated for schools. So schools are faced with this, all this rich data, that is minute-by-minute what their students are doing but they actually often don't know it, don't get that information. They just get either data that they don't know what to do with, or they don't get any data. And lastly parents and students themselves. For them to make good choices, the same problems That is, even though the data's there, they don't get the information often that they would like or need, often because they don't know what that information is, but often because it's not shared among these three parties. So this is the sad factor of being early in a field. All the parts are not working well together.

- [Voiceover] And David, let me just jump in here for one second, just to note that Andy is, based on his experience, corroborating what you're saying about this whole issue of the data that supposedly is available for teachers in particular but it's often very siloed and that from Andy's experience, he only had information on his own students, but also issues of IEPs and privacy also have an impact on who gets the data and how it can be used. So thanks, Andy.

- [Voiceover] Great, yeah, thanks Andy. So you've been there. I can see that. And so what's hard is there's great opportunity in here and we're not realizing it yet. And true, not like there's blame to be assigned. Privacy's an important issue, but right now it's a barrier to actually doing the kind of individualizing instruction that we'd really like to be able to do, and we'll talk about this later when we get to policies. So a third topic. Old methods in new packaging. The majority of online instructional offerings, as we look at them, are mostly traditional offline materials delivered online, and you've heard this before, largely electronic workbooks, page turning and answering questions. And the online world has some benefits. There's things they can do that make that better but they don't change it in a substantial way. It's just a new delivery mechanism for old methods. And less common are, I have here curricular offerings, but it's probably instructional processes that would take advantage of the unique affordances of online learning, so, we see less than we would like or you would like of really taking advantage... Getting like crazy on all kinds of devices, but a lot of the online environments really kind of are divorced from that. The kids actually don't communicate with each other. And one that we'll talk about a little bit more later actually is the new, one of the new powers here is to really support students in learning to self-regulate their own learning, 'cause there's more opportunities for kids to take autonomy to build independence, but that would take kind of a design that really fosters that and supports it and stuff, and lots of the online providers don't really have that so the kids are kind of left on their own to self-regulate often and that doesn't work very well. One of the great of course powers of these new technologies is that you can really differentiate instruction. Enough variability there to allow kids to take very different pathways and there's not enough advantage taken of that yet. The new technologies would allow kids to demonstrate what they've learned in lots of different ways, but most of them are pretty much multiple-choice answering questions and so on. There's lots of new capabilities that the online world can provide that are being taken advantage of only what should I say, only beginning to be taken advantage of. When you look across the whole spectrum of online learning, it looks pretty traditional. There are great innovators, so I should be careful. Remember that behind every one of these, great variance. We'll see really interesting things happening and lots of really boring things, too. A subpoint of this is that the majority of online providers are still focused on content and kids with disabilities are especially needy of process development, learning such things as self-regulation strategies, how to manage their time, how to build executive function, how to socialize better, how to know when to get help, all of those things. And mostly what you see when you look at online learning is content mastery. Just, we showed you a bunch of things to read. Did you gather that in in such a way that you can answer questions? So again a kind of conservative early stage in online development. Another aspect that we'll come to again is that we have this enormous data capability with automatized analytics to really do rich personalization and really monitor students, not just right and wrong answers but how much are they doing? How much chances are they taking? How much help are they wanting? How much, all sorts of things. What's their academic health? We have ways to measure that that are just completely unusual compared to traditional brick-and-mortar schools and there's great variance. Some companies are doing pretty good at it. Some companies, there's almost none of that. Fourth topic, design. Design matters a lot. And there's enormous variation both between and within vendors just meaning basic accessibility. Now, one thing to highlight here 'cause it makes it really hard for us to do our work but also for a parent or a teacher to make good decisions 'cause a single vendor may have courses, for example, that were made by very different people using very different technologies, and so you may have one course within the same vendor, let's call them, Superman's Learning Objects or something. One part of it may be very accessible to kids with disabilities, and then another part, which looks the same, same vendor, will be completely inaccessible, so even within the same vendor, same company, there's huge variation, which makes it. As a Center, we thought we'd be able to highlight, "Well here's some vendors that do a really good job. "Here's some vendors that really do a bad job," and be able to provide to consumers good information about that but actually it's much more complicated than that. Within a single vendor, we have to go sometimes down to single pages. They're going along pretty well and all of a sudden they'll have a page that's completely inaccessible, but of course that's terminally important to a kid with disability, to get somewhere and all of a sudden you're at somewhere you can't do anything. So the variance within is almost as big as the variance between vendors, a huge problem. And beyond sheer access, UDL has been adopted in various ways by a number of vendors. Some very strongly and some very meekly. Some I'm sure have never heard about UDL. And so variation exists in whether they go beyond access to design in the kind of learning supports that a wide variety of kids with disabilities would need. We've already mentioned for example building in supports so kids can, for kids with executive function disabilities so that they can in fact be able to be somewhat independent but have the supports they would need to do that well. When you look at a vendor, this is sort of at the low end, this is what you can see out of a part of the Center's work we developed a UDL scan tool and it's just looking at, does this vendor provide multiple means of representation? And there was 36 would be a top score and you can see that in fact, they do some things but it's really kind of very meager in this case. Here we are just entirely around perception. Sometimes the whole thing is completely imperceptible to people with who are blind, for example. So we'll have big variation, some do a, we've actually also seen several do a very good job. Really stuck to the web accessibility guidelines and have an institutional policy of trying to make their stuff accessible. Okay, new topic. Context really matters and I think more than we expect it. I think it first came to us with a very, what you'd have to call rigid, online provider that really kind of locked into what you should do and yet when we were able to get some data, what you find out is that it works really well in some states or districts, and really not well in another and yet it's exactly the same. The technology's exactly the same, what the vendor's providing is exactly the same. And you see this huge variance between different locations, and when you look at that data, you're really struck by, how could that be? Like, actually I was gonna say it's like medicine. And it is like medicine actually. If you give me certain drugs that are really good for Sally, they're poisonous to me as it turns out 'cause I'm allergic to them, and it turns out that it's that kind of variance and you'll see, this really works in Idaho. It doesn't work at all in Austin. And it sometimes it'll, if you get close to the data, you'll find out it works in some classrooms and not in others in the same school. And of course that has to do with, what are the teachers doing? What is the community doing? How has this been set up? What is the technology that's available? What is the orientation and motivation about schooling in general here, and so on. All of these have huge effects, which is reassuring as an educator, I think. There's not a magic bullet here. It's not like this stuff really always works, doesn't work. It really depends on the soil in which it's gonna be planted. And this just slide, it's kind of a reminder that, you'll hear about blended learning. And what's blended learning? I don't think we think there's. There's very few things that are not blended now that, it's only a few completely virtual instantiation of online learning. Most of them are blended in some way, but the blends are really quite different. Some, I just want to highlight that there are things that are natively offline, that are often done as part of an online delivery. Kids read a book or watch a video, but that might be offline completely. There are things that are natively offline that are delivered now online, which is again, like I said, not really changing anything. Just taking the textbook and put it online and you click through it. Watching a video online. Is that online or not? Is it really any different than watching a video in your classroom? And then there are some things that are natively online, things that you couldn't actually do except online. Having a group project that's measuring the salinity of all of the rivers and streams all across the country, and all the students gathering the data and analyzing the data, doing experiments. You can't do that offline, it's natively online, and we're finding that there's an increasing but still beginning use of the things that are really natively online done. This is getting at the power of what an online network environment can really do for kids that is different than what could be done in a single classroom. All of that variation that is between how much is offline, how much is online, how much use is real online, there's big variance. But then you have the variance of what happens locally, and this needs a real number of student ratios. How many kids are there that are under the supervision or mentoring of a single teacher and it's like everything else I've been saying, support has huge variation where sometimes a single teacher will have 200 kids that are their students, and sometimes it'll be very small. And by the way this can sometimes happen, we've seen instantiations where a single teacher can have 25 to 50 kids with special needs, all as their students. And I don't think anybody out there thinks, "Well that wouldn't matter what the ratio is." The ratios are gigantically different. In some cases, the variance of what you do with kids with disabilities, if the vendor actually knows, the school has shared who has student disabilities, is just like in brick-and-mortar schools, highly varied. Sometimes special needs kids get special things. Sometimes they're just in with sort of everything that everybody else does, but that's a source of variance. The number of contacts per week or month, highly variable. Some kids in virtual environments have personal contact which would be by phone or online once a week or once a month total and some have it almost every day. Huge variance. Who's responsible for something like an IEP? There's great variation in whether the school retains that responsibility or the vendor takes up some or all of that responsibility and some of the worst-case scenarios where the school has done a full IEP and can't share it with the vendor, so the vendor has to redo essentially an IEP. They can't get the kids going on their learning because they have to do a whole new assessment 'cause they can't get it from the school. And I will say, by the way, I hope I'm not sounding, putting down either of those partners, 'cause they both have problems. We've seen cases where the schools wouldn't share the IEP with the vendor not because of policy but because they were ticked off that there was online learning in their school and they just want to make the job of the vendor harder. So there's... solicit parents directly, and say, "Well your kid doesn't need an IEP in our system because "all of the learning is individualized" but for reasons I already showed you, sometimes the material is completely inaccessible, there's no teachers that's involved anywhere near enough that individualized, there's no executive supports and so on so that's where sometimes this will only become Some vendors essentially are circumventing all of the protections that are built up for kids with disabilities and saying, "Don't worry about an IEP." And so we have that variance. That's what the field is like. Some vendors are desperate to get the IEP so they can do a good job, and some vendors are saying, "Eh, don't worry about an IEP. "It's all automatic," which is not the case. Even who's the primary teacher, as I think I'll show in another slide. In some cases the parent becomes much more the primary teacher than we ever expected. So summary of this, the context is so varied that it has strong effects on outcomes and one can't study, for example, is online vendor X. It's almost impossible to say, "It works." You almost, like other things in our field, you have to say, it works for whom and where and when. It's almost impossible to think about them working in the abstract where vendor A is always good. Rarely happens. Sixth point. Just checking in. Oh we had some...

- [Voiceover] We had some technical difficulty, David, but it looks like everybody's back, can hear. So you're doing just fine.

- [Voiceover] Okay, did I say anything interesting during that time they were off, Tracy?

- [Voiceover] You did but we will be able to post the slides so that people will get to see what they... They could see, they just couldn't hear, David.

- [Voiceover] Okay, the seeing's the best part. So, sixth issue. It's, sorry, the continuation of the sixth, which is about teacher preparation. We looked at, what's the difference if you're teaching online or offline? And all of the stakeholders reported that teaching online is really different. Here's a list of all the things. I'm gonna skip this for now but it's really quite different. Not the same at all. You don't get to see the student, you can't tell who's looking depressed, you don't know that their parents divorced last night, or any of that. It's really, you're in a completely different environment and materials are different and so on. But few teachers are being prepared professionally for this difference. Very few teacher preparation programs. Think of all the teacher preparation programs you went to. They probably still don't teach, they don't give internships or practice teaching or anything in the online world. And yet a lot of people are now teaching in this online world, but they're not being very well prepared. Vendors often provide in-service teaching for teachers but often it's hard to get time to do that well enough and so on. And schools provide it, but it again is a source of huge variance and some teachers are well trained and a lot of teachers are not at all well trained in this whole new medium. National survey, which I know, oh there's the authors. Less than two percent of teachers preparation programs offer field experience in online learning, so people are not being prepared for this, and so on. There's all sorts of early-stage problems here in teacher development. Parent preparation. I mentioned it briefly but we certainly have seen instantiations of ones that are more virtual where parents are playing really the critical role as teachers. Kids are at home and under the parents' supervision and there's only occasional interaction with live teachers again via phone or email or something. And we see the opposite too where parents have no knowledge of how their kids are doing online or know anything about it. Don't see it. So we have the, like the other topics I've had, there's variance between online learning, in one case where the parent is essentially really a provider, and the major teacher, and other instantiations where the parent really can't even get information, doesn't even know what's happening. So the variance is hard to the parents and parents, like teachers, report, that they're not really prepared for this. They would need more. They often don't know what they should do, what they shouldn't do. We've seen, we did a little case study work where we could see, go into the home and see parents working with their kids to see what really happens, and huge variance comes. Parents truthfully are filling in the blanks for their kids. You know, if they want their kids to get a good score on the end-of-unit test, the parents are kind of doing it. I think Tracy would've done that actually, but anyway. And some parents, we found a case where one parent was supposedly helping her kid and supporting them and so on, was actually running a family daycare in another room and almost never there with their kid. So, huge variance again. The biggest, and I'm getting near the end of this, just so you know. On the context issues, the biggest one from our work so far has been policy differences. Every stakeholder in the ecosystem from parents to school defenders complained about the problems that very inconsistent and anachronistic policies get in the way of anybody doing a good job. That means defender fields like, we could do a better job if the policies would allow us to do that. Those policies may be state policies. They may be district policies. They might even be a single teacher's policy that she. You know, a teacher might have a policy, "I never let the kids have text-to-speech" or something as a policy. And yet, that kid actually needs that and the vendor provides it, but the parent won't let the kid use it or whatever. And all kinds of, we talked about privacy being a big issue where everybody would like more data to transfer between these parts, and policies really get in the way. So policies are if anything the big challenge here and I'm not sure we thought that when we got into this work and at the end of the talk, I want to talk about we want to actually go after the policy arena 'cause it's such a dominating effect. Here's some suggestions. I think I'll skip these for the moment just to that there are increasing attention to laws that would protect students. I think you know that. It protects their privacy, which is very important. California, here's the very new policy which really stipulates strongly protections so that student information can't be sold and licensed or given away to other people, especially corporations wishing to sell them, whatever. And those are really good and necessary things. What the challenge is, how to. You can't do a good job as an online provider unless you do have information about the kid. And you can't as a school system evaluate whether the online provider's doing a good job unless you have information tagged to your specific students. Is this vendor's program working for my autistic kid? It's a central question. Is it working for my kids with dyslexia? It's a central question. And to know answers to those questions, we have to share data and that's one of the challenges that stick out there. At the Center we've been running a state policies table. You can look at different states and see what their policies are. And they're changing rapidly as you'll see. Here's the kinds of policies that are there. Policies effect almost everything. What you've got to be prepared as a teacher, what kind of accommodations are necessary, what kind of, how can funding be used, can you give a bonus for kids with disabilities to a vendor, etc. All kinds of policies. And we looked at the policies as they're changing a little bit between 2013, 2014. I think, I don't want to go through these all. You can look at these slides later. But here's, to give a sense of how the policy people in a state, how do they feel? What do they know? And in 2013, 2014, weren't really very different at state levels in terms of, do they even know how many students with disabilities are in online learning? And you can see that most of them feel that they're not very knowledgeable about that, which is, you know, pretty basic thing. Do you have a lot of kids with disabilities taking online courses or only a few? And actually, state heads, these would be state Directors of Special Education, they don't even know that. Can't get it 'cause the data just doesn't work right now. So they don't know whether The largest majority say they're not knowledgeable. I don't know. Which is kind of amazing, 'cause the data's there. And if you ask, how comfortable are you with your policies? Most people are uncomfortable with their own policies. They could see that this is not working. So the majority of people say, "we're not comfortable." And a large number say, "we're not at all comfortable." Okay, that's. That slide got in there by mistake. And we asked them, again state directors, how confident are you that your teachers and administrators know how to deliver in these online worlds to kids with disabilities? 80% say, "I'm not comfortable," meaning, "I don't think teachers are prepared to do this." So again, here's a, this is worrisome that the people who direct special education don't feel that they're ready to do it yet. Now again, huge variance. You'll find a pocket, a school, a district where it's great and you'll find others where it's not great at all. People are doing it that don't have really much training and don't know how to adapt to kids with disabilities and so on. A question that you love to think that it's easy for us to get the answer to is just the simple question, how many kids with autism are doing this? How many kids with physical disabilities are doing that? And as I said, that's very actually hard to get because of this discontinuities between the data sources. There's a few things that are highlighted here. You can see this difference when we looked at different vendor-school partnerships, we could find some partnerships were only 3% of the kids enrolled were kids with disabilities. Another partnership had 35% of the kids enrolled were kids with disabilities. We've heard that it's as high as 60% in some places where there's vendors who are preferentially going after kids with disabilities. So a difference between 3% and 35% is pretty huge. Pretty variable. You see the stats on why the families choose an online-environment school and a lot of them, I have to say, the more vibrant ones that come out of it, aren't so much as big numbers, but partic. Oh, this is where you can see how problematic it is. For some kinds of disabilities, the child being bullied is the majority reason they had the kid go online. For some it's a minor thing, but for some types of kids, kids with social disabilities, it's very prominent for parents to say, "I want him out of traditional school "because he's being bullied." And I have someone here, Janet's pointing out that Andy said something.

- [Voiceover] Andy was corroborating. He was talking about his work with autistic students who actually thrived in online environments due to the low social interaction and the highly structured nature of the learning environment and then Alice chimed in that she worked at an AT center where they used several online packages for reading and literacy, but she knows that they were coupled with individual instruction so the students have access to that traditional face-to-face individual instruction.

- [Voiceover] And that's a key part of the variance. Some schools provide that and some don't so whether the online's gonna work depends, I think Alice is saying, on you getting that individualized instruction offline. And yes, but I want to go back to Andy's point 'cause it's critical. Here's the issue. There are that we don't know good enough answers to. There are kids that are being pulled out of traditional brick-and-mortar schools and put online because of their behavior problems. They have social problems. Some of them are kids with autism. And both school and parents will report things are better. 'Cause they had fewer behavior problems in the school and the kid is not bullied or whatever. But we as educators have to look at this closely because what we're essentially doing is going backwards. We're actually putting the kids back in the basement. We're isolating them again. So we have to always be looking at, what are the goals here? And if the goal is mastery of content, then probably pulling the kids out, putting them in the basement would work. But the goals of educating kids with disabilities are much broader than that, and how are they gonna learn to exist in the culture they're gonna grow up into if they're essentially in isolated conditions when they're learning in school? So these are complicated issues and we're wrestling with making sure that we measure the right things and that's what I was talking about, that sort of, like executive function and self-regulation. These are things kids really need to learn and you're forced to learn them in face-to-face social environments with other kids and with a highly structured program, you actually aren't forced to learn those things as strongly so you may never learn them. So I hope that's good, but thank you for bringing that up, Andy. Definitely autistic kids are often thought of as being the kids who most do well in the online environment by vendors and schools but I would again say, "Well, at what? "Are we doing the right thing?" Okay, and this is just, do kids persist? They complete? What are their outcomes? And here's actually that same issue. I would say that the vendors believe that if they really knew, they think that kids with autism for example really persist and do well in their program, as long as, again, issues are content mastery. But again, it's very hard to tell given the sequestration of the data so they can't answer the questions we'd really like to ask. The primary barrier to research on these key variables is that we'd just been talking about. Are they, what are the outcomes? Are they really getting the full outcomes that schooling's meant to do? Are they persisting and completing these courses? We need to know by what kind of disability for example. What is the difference for autistic kids and kids with dyslexia? And right now it's very hard for us to study that even with supports from the Department of Education. So I have just a few minutes. I want to just complete with where we're going with our research 'cause we've learned that a whole lot of the ways we thought we do research are not gonna work, and that in many ways we see the online system as having a disconnection syndrome, which is here's the three main players and there's the offline context I've talked about here that matters hugely. What's happening in the local environment really determines a lot of whether this is gonna be successful or not. The online provider, what they make, how they provide it, how they supervise it. All of those things are huge factors. And then there's the individual. Does the person have autism? Does the person have social phobias? What is the individual thing and what's the parent's capabilities here? These three sort of domains in the ecosystem are the determiners all together in whether we could say it's working. And they should be working together. That would be the best thing, that parents really know what happens when their kid's online, and that parents know what happens in their school at the same time, what's the school doing, and the school knows what happens when the kids are online. "Oh I got data back tonight of what happened to Billy and "oh Billy kinda fell off the channel here. "What happened to Billy? "Why didn't he make any progress today?" And schools should know that. And the online providers, they need to know what's going on in Billy's home context. What's going on in his school? Like Alice's question, is he getting some instructional help on reading or not? It'd be good if we knew that. So these need to be working together and disconnection syndromes. I'm a little bit too late to go into it, but the nervous system has disconnection syndromes where the parts are fine but they don't work together. And that's what's happening in our online world. The parts may work together. The online provide may be doing a good job, but it's disconnected from two key sources, knowing about the individual and their parents, knowing about the, what's happening in the school, what extra services are being provided, what else is going on. Is that kid playing with kids? That right now, these are disconnected parts. And the center is connecting its research moving forward largely by working with all three of these entities at the same time, in what we're calling Research to Practice Partnerships and they have advantages, which are here. Some of them I'm gonna skip, but by not being connected, there's two kinds of disabilities, internally, the system actually can't monitor its own progress. It's hard to know, just like I said, whether kids are doing well, or for what reason are they doing poorly even. It's hard to tell. Maybe it's just 'cause they're not getting something in the offline world that they need. Maybe it's because the online provider's providing something that's completely inaccessible. Maybe it's because the parent has disappeared when they should be there, whatever. But internally we can't tell because the information's, there's data but no information. And externally, researchers like us can't really understand the parts without looking at it in this holistic way. So I'm gonna give an example, and then I'll end, of how it feels disconnected. Audio-supported reading I think a lot of people are familiar with that, having, a lot of these environments have a lot of text in them. Guaranteed, that's gonna be problematic for a lot of kids with disabilities, so most kids with disabilities face kind of this text barrier, and we really don't know very much about when and how audio supported, that is letting kids use text-to-speech engine, the fairly sophisticated ones, you know, we need to know when should they use it, when should they not use it, all those sorts of things. And so we want to study that but let me just show you how it works. The provider in this case that we're gonna work with actually knows every time a kid clicks on a word and gets it read out loud, every word, they know how fast they said it, they know when they did it, they know if they did it at home or school. They have great data. They can record every time a kid uses it, which would be great, a great way to start our research. But they don't have access to the data on who the kid is. They know it's kid number 1473. They don't know who they are, what their disability is, what's happening in their home, or anything like that, whether they're poor, rich, anything like that. So that limits how much they can know or do, and they don't have data about the school. It would say, what is this. In some cases, teachers say, "No no, don't use that," and we watch that. Teachers say, "No, don't turn on that text-to-speech" but the provider doesn't know if it's because the teacher told the kid not to or if the kid doesn't want to. You know, what's going on? So you have this disconnect. That's what disconnection looks like. So what we're doing is working with all three of these things together to do our research so that we can really start to understand is it something in the context that made the big difference? Is it something the provider's doing that made the big difference? But mostly it's gonna be their interaction, just like everything else in life. So we're reimagining our research going in that direction, which is reducing the barriers between these functional parts and getting them to be able to communicate well and using that as a test bed for the last point I'll make, which is, what are the policies and procedures that would allow these parts to talk together. Right now they can't. Sometimes it's technical. A lot of times it's policy. So we want to actually work in local implementations like this, to change policies and technology practices so that the parts talk together and so that we can do our research and that the school can get better, the online provider can get better, and individual kids can get better, so our research is gonna look like that over the future. Always with at least three components in a partnership and I think we'll be able to get more progress and learn a lot about how kids with disabilities can do in an optimally connected environment. Okay, I finished all I needed to say right on time, Tracy.

- [Voiceover] Yes it is, David, and you've done really such an outstanding job of identifying not only the challenges that are facing all of us working with students with special needs, but also the great opportunities that are out there. We did have quite a robust discussion but we're now at 4:00 so what we will do is make sure these comments are posted online as part of this discussion and hopefully David we can entice you to come back and to continue this very rich and interesting discussion.

- [Voiceover] Well it was a pleasure. Thank you for setting it up, Tracy, and I think I probably had a slide that says we have some nice... oh good you got some next steps here Tracy, but if anybody knows a good partnership site where you know there's a school district that is really interested in doing better with kids with disabilities and they have a vendor, which is, a promising vendor, and that we could work with to make a research partnership, contact us. We'd like to get probably one or two more so we can do these kinds of studies.

- [Voiceover] That's great. Well thank you, David. Can't say enough about how informative this was and how important the work is that you're doing, particularly as we find ourselves seeing online learning being rolled out with really very little information about for whom is this working, under what environment, and what are the supports that students need to be successful. So take care, everyone. Hope to see you again online, and thank you all. Take care. Thanks, David.

- [Voiceover] Byebye.

- [Voiceover] Byebye.