CDT Takes On EDTECH –– Talking Tech w/ Elizabeth Laird, Hannah Quay-de la Vallee, Adarsh Mahesh
The Center for Democracy and Technology is ramping up its student privacy work – and here's your introduction to our new team: <br><br>Elizabeth Laird, Student Privacy Senior Fellow<br>Hannah Quay-de la Vallee, Senior Technologist<br>Adarsh Mahesh, Research & Communications Associate<br><br>More on our student privacy work: http://bit.ly/cdtstupriv<br><br>More on Elizabeth: https://cdt.org/about/staff/elizabeth-laird/<br>More on Hannah: https://cdt.org/about/staff/hannah-quay-de-la-vallee/<br>More on Adarsh: https://cdt.org/about/staff/adarsh-mahesh/<br>More on our host, Brian: bit.ly/cdtbrian<br><br>Attribution: sounds used from Psykophobia, Taira Komori, BenKoning, Zabuhailo, bloomypetal, guitarguy1985, bmusic92, and offthesky of freesound.org.
Top Keywords
- privacy 0.023
- data 0.018
- student 0.017
- student privacy 0.017
- students 0.016
- hannah 0.012
- schools 0.008
- school 0.008
- districts 0.007
- adarsh 0.007
- data integration 0.007
- education 0.007
Transcript
Welcome to Tech Talk. Bye. CT.
Tea.
Welcome to CDT's Tech Talk, where we dish on tech and Internet policy while also explaining what these policies mean to our daily lives. I'm Brian Wasilowski, and it's time to talk tech. Tech Talk is back after a short break, a break that was completely my fault. I went and got married, yay, and then did a honeymoon. But now I'm back and so is Tech Talk, and we have a great episode for you today. CDT is seriously ramping up our work on student privacy, and we have a shiny new team that has plans to address topics such as student data portability, almost lost that one there, system interoperability, and how digital decisions in schools can be used in positive and not so positive ways. That team joins us today. Elizabeth Laird, Hannah Quaid Alavalay, the best name in our office right now, and Adarsh Mahesh. Did I say that right, Adarsh? There we go. Oh, I should know this. I with you. So if I don't, you can correct me. Welcome to all of you student privacy team. Say hello. Thank you. Hi, Brian. There we go. Three of them. This is our largest podcast not ever, but, it ties it for the largest podcast. So be brilliant. You gotta cover all this and we're gonna give you a lot of time. So we should be. Brain trust, you know. So let's kick it off. Why are you all so interested in student privacy? Like, why did you get into this topic and why do you work on it? Sure. So, I'll start. This is Elizabeth Laird and I'm a senior fellow for student privacy at CDT.
I did not grow up as as a young girl thinking, you know what I wanna do one day when I grow up? I'm gonna work on student privacy. No? No. I got interested in student privacy when I began to work in education policy and quickly realized that there is not good information or research available on what is improving outcomes for students. And so that means as a policymaker, you can't do less of what doesn't work and do more of what does. So initially is what got me interested in data, and then learned, the hard way, by working in states for five years that you can't effectively use Dan technology to improve student outcomes if you're also not equally obsessed with protecting their privacy. So that is really the lens that I bring to this work is how do we use Dan technology to improve outcomes for lens that I bring to this work is how do we use the end technology to improve outcomes for students, while being, as equally, protective and obsessed with making sure that their information stays secure. Cool. What states did you work in? I worked in Louisiana,
and DC. So two very different different places. And DC should be a state, but isn't. Right? Yeah. Oh, wait. That's a separate official CDT
discussion point. Yeah. It's a separate podcast.
Hannah, Darce, what why did you get into student privacy? This
is, Hannah I'm the one of the senior technologists here at, CDT. So I'm coming from a tech background. So my interest in student privacy is basically throughout my background in tech, I moved further from, like, very extreme whiteboard math, like, very disconnected from doing day to day tech work, more and more into user facing work. And I came to realize that, at least in my opinion, a lot of the issues around privacy and technology are going to have to be solved from a advocacy and policy standpoint. And I think that's exceptionally true in education where students and parents and even sometimes schools don't have a lot of say in what tech they're using and how they manage their tech. So it becomes all the more important for the policies to be in place to protect the students and their data.
That's great. And Adarsh, who actually joined us from, just moved down. You were working upstairs at a different organization and moved down seven floors or something like that. Three floors. Yeah. And my commit did not change at all. You're welcome.
Which is very easy on me. This is, Adarsh Mahesh here. I'm the research and communications associate on the student privacy project. This is my very first podcast. So I mean, if I if my voice sounds a little shaky, that's just me freaking out. I mean, Y'all are doing great. I'm not a recording issue. But, yeah. I mean, on the city side, what brings me to, the student privacy space as, having seen firsthand, sort of education being the great equalizer, going to serve a public school in India to coming here and, working here in Washington, DC, the capital city of United States Of America, working on some of the most timely and important issues of our times alongside some of, you know, the most brilliant minds, in their respective fields. And, having sort of always wanted to work in the education sector, and sort of increasingly, we've seen, technology being used more and more in classrooms, as as we call EdTech. So you can't have a conversation, around education without also, you know, speaking to, the transformative, effects that data, brings and also at the same time, the privacy implications, of of students and sort of, as I'm sure it's but it'll explain later, sort of talking about the benefits of technology and making sure that it doesn't come at the expense of the privacy of students. So that's what interests me most about this space. Great. I sense a lot of passion and energy for this topic. This is great.
So what are y'all actually gonna do? I mean, you obviously, team of three, what are the big issues you're gonna be taking
on? What are the the core projects that are gonna be part of your work? Yep. Sure. So I this is Elizabeth again. I will kick us off, but, my, more technical friend and my more research and and communications oriented friends will certainly jump in. So I think from, us, you can expect three things. When we are thinking about student privacy, we are approaching it, from through three different perspectives. I think the first thing that you can look for from our student privacy project is balanced advocacy. And by a balanced, I mean, issues of data and technology and privacy can be very, divisive and you can quickly fall into two camps. One camp that believes that data and technology are going to cure all of society's problems and the other camp that believes it is going to exacerbate them, and and do harms to individuals. I think our approach to this work is is we believe, in the power of data technology to improve outcomes for students. We also believe deeply that that cannot come families. So when it comes to advocacy, that is, that will be the approach that we take is how how do we do both? We can and have to do both. I selfishly care a lot about the second thing that you'll see from us which is to provide policy solutions. I care deeply about this because, I in part took this job to create things I wish I had. So leading student privacy policy for a couple of states was oftentimes, an isolating experience that there weren't a lot of resources out there that I could easily apply to my job. So I think we wanna help those who are doing the work and actually provide them with policy solutions. And and lastly, looking at Hannah, although you can't see that because this is a podcast, but Hannah She did look at her. I can confirm. We wanna translate those policy solutions into technical recommendations. And again, this is to help the people who are on the front lines doing doing the work and need things that can easily be adapted to what they're doing. So I will pause there and see if Adarsh or Hannah have anything to add, in terms of that approach.
I think that about sums it up. Yeah. Mhmm. That's great. I I can't imagine doing anything better than that.
So what this kind of as I listen to this, obviously, I've been at CDT longer than any of you. It sounds very much like a CDT approach to policy making. Is Is that kind of the unique lens that CDT is bringing to student privacy? We've been in privacy for a long time, have done some work in student privacy, but you all represent a major, as I said in the intro, ramp up of our focus on it. Is that, you know, what you just described the unique part that or approach that CDT brings to this? Yeah. I think so. I think,
I think, again, we did a lot of stakeholder older engagement when we first got here, both talking to folks who are working at the national level as well as folks working in states and districts to get feedback about what would be the most valuable, what would be the most valuable resources that CDT could provide, and that's what we've heard. So as much as that sounds like a well honed three point approach, we started really from scratch. And and that's very much informed by folks who are doing the work, and getting feedback from them about what would be most helpful. Cool. Well, let's dig into some more specifics about what you're gonna be do doing because I've seen some of your plans.
Predictive analytics. This is obviously a big part of education and used to predict a lot of different things about students, about, interventions. What are some of the ways that they're used beyond those in education systems, and why do you wanna take a look at the use of predictive analytics?
So this is Hannah. Predictive analytics is basically any sort of system that takes in data and outputs some sort of recommendation or answer to a particular question that you're asking. Predictive analytics gets used in education in a number of ways. Early warning systems is, is a good example. Those are systems that take in data about their students, either social data or academic data, and try to identify students who are at high risk of dropping out or not finishing. Another example, is school choice systems that help that take in a number of factors to assign students to particular schools. Another area that, we're seeing a ramp up in is school safety. So similarly to early warning systems taking in both social data, academic data, even social media data about students to try to identify students who are at risk of committing a violent act. So these systems can provide a lot of benefit. They can help notice patterns and trends that might be difficult for human teachers or administrators to spot. But they can also be quite dangerous in the sense that algorithms, the sort of engine behind predictive analytics, can can exhibit bias. And certainly because they take in data, they're only as good as the data they use. And because we can we live in a world that has a lot of bias, a lot of our data is biased. So for instance, if you think about a school safety analytics system, something that's trying to find students at risk of committing a violent act, if one of the things that takes into account is, past interaction with law enforcement. Well, we know historically that students of color are much more likely to have had an interaction with law enforcement, not necessarily because they are more likely to have done something to cause that, but because of the nature of policing in America. And so if your intervention the intervention that you take based on the recommendation of this analytic system has any harmful effect on the student, that can, further amplify the bias that we already see in our society. So you have to be really careful with how these systems get used and really conscientious of the potential
damage they can do. Yeah. And we've talked about, you know, we've turned to a digital decisions at CDT before. We've talked about kind of the the issue with bias and algorithms. I would imagine that educational data is some of the most sensitive data out there. And certainly, the ones when you think about kids and parents wanting to protect their kids, it's one of those user cases that definitely jumps top of mind for a lot of people. Yeah. So great that you're working on that. There's actually a bill out there, isn't there in Florida related to privacy and school safety?
Yeah? Yep. So Florida, in response to the Stoneman Douglas shooting passed a bill which I recently learned was written and passed within a week. Wow. So it was a piece of legislation that was drafted and passed very, very quickly. Those data sources are from other state agencies. They can include juvenile justice, health and human services, and also collecting social media data as as Hannah described. Integrating all of that, which has never been done by a by a state education agency before and then partnering with law enforcement to then share that information. Wow. So it is really unprecedented in terms of, the the scale of of the data integration. And because it hasn't been done before and because we really do lack good data, and research on school safety, I think we have some really serious concerns about how do you play put in place the right safeguards, including really strong data governance, really strong limitations on access, limitations on how that information can and can't be used to make sure that that information
isn't used to harm the various students it's intended to protect. Yeah. And sounds like there's not a lot of information on whether it would actually be an effective intervention
yet. Yeah. I think, it's funny you say that. Recently, there was, an article by NPR, and the article followed up on a report that was released by the US Department of Education that provided information on, the number of of, schools at which a a a, a gun had been discharged. And so this was coming from the civil rights data collection, and they added this new question. And what they found is that the what was self reported by school districts was off by about a factor of 10 Wow. Of over reporting. So in the article Interesting. The the or the the report that was released by US Department of Education had school shootings, I believe, in, like, the 250. Every town, has a number of school students at shootings at 26. So we don't have good information about even when these shootings are occurring, much less what is contributing to them. So it's a real it's a there's real tension between the urgency to act which is totally understandable and a real lack of data and research. And so policymakers are trying to reconcile these two things to act in the best interest of students.
But they have to do so without good data and research. That's tough. So let's pivot a bit. You mentioned the word data integration. I guess that's two words. And then also you've talked about data portability, interoperability. And I know this is also going to be a core part of what you're all looking at in terms of educational systems. So maybe tell us what those terms actually mean, for you all and what you're planning to do in your project around
them. Alright. So this is Hannah again. So, one of the things mentioned there is portability. And that, for our purposes, essentially means students, maybe parents, even school districts being able to move student data across systems. And that can be systems within a single school. It can be systems across districts, across states. And there's, there's a few reasons why that might be important. A big one is basically that it it helps avoid what we call lock in, which is essentially where you've put so much data into a system that the inertia is so high that getting out of that system is basically impossible or prohibitively difficult. And so that both prevents, schools and students from taking advantage of the best the advances in technology, new systems, new capabilities, and also places a lot of power in the hands of whatever company or envy is behind the system that they're currently using. So we see data portability
experience. And if I could just add on in terms of portability, another another reason this is a really important issue in education is, student mobility. So students are are moving all the time. I know I moved three or four times when I was growing up, and we all have our own personal experiences around what that was like. But we see that that the neediest students are the ones who are moving the most and are really, negatively affected if that information doesn't follow them. So for example, a student with a disability, if he or she were to move across, schools or districts and that information wasn't to follow, they don't have they're not receiving the services that they're entitled to and that they need to receive. So I think in terms of of portability, we also care, because of, you know, students are moving and the most vulnerable students are the ones who are moving the most and they're the ones who most need that information to follow them so that the educators that are receiving them can can best prepare to meet their needs.
Makes sense.
And then in terms of data integration, I think that's a great question and, you know, it can mean a lot of different things to a lot of different people, so I'll sort of speak to what it means here. Data integration in education is not not a new issue. I think it's long been acknowledged that students do not go to school in a vacuum. They are being touched and served by lots of different agencies and different stakeholders. And so the issue of data integration is one that that school systems have long struggled with. How do we, integrate data? What I'd say is horizontally. So at a moment in time, the the adults and the agencies that are serving a student, how can they integrate information to better meet their needs, but doing so appropriately and in a way that protects their privacy? Another form of data integration is also to follow students longitudinally or over time. You know, we know that what what I what initially got me interested in data and research was actually being able to follow students over time and understand how their experiences earlier in in school or even in early childhood before they get to school ultimately influence, what happens with them once they graduate high school or go to college or get a job. And that's really the the lack of information that policymakers, have been limited by and therefore have made, you know, the best decisions they could, but they're uninformed decisions because they're not good information about Yeah. The kind of experiences that students are having and then how that ultimately determines some of the outcomes in their lives.
Great. So that's a lot of data that we're creating that needs to be interoperable and integrate and all these sort of things. I know you're all looking into the notion of chief privacy officer for schools. Is that something all schools need? I mean, a lot of major companies have them even non, you know, tech companies have chief privacy officers as they deal with more and more data of all different types. Is that something that you think is needed at, you know, the school level, the district level, or does it vary? Yep. I think it's a great question.
And I I will, speaking of bias, show mine immediately, which is that I was a former
former privacy officer. So I, of course, think it's a very important job. So we need to clone you and put you in all these different places. This would be great.
Yes. In addition to that no. So, yeah, I think I think building capacity around privacy is really, really critical. I think there there is, maybe a misconception that privacy student privacy is a new issue. It's not. The Family Educational Rights and Privacy Act or FERPA, which is the main student privacy bill, has been around for a long time. Do you care to guess what year it was what year it was passed? I should know this. This is not a fair test. Does your team know?
Hannah. Hannah has her hand up.
Go ahead, Hannah. I was trying to I was trying to give Brian signals. 1974?
Yeah. Oh. I'm really glad I got it. Wow. That's even before I was born. That doesn't happen much anymore. That's fantastic.
So, so even even just, like, the basic federal protections around student privacy, schools should have been doing this for over forty years. So privacy is not new. Yeah. What is new is are the demands that are being placed on schools and districts and states around data and technology. But they the the systems and the bureaucracies haven't responded. What also is not new is that these organizations are still incredibly siloed. Privacy is split among multiple divisions. No no single, person or division owns it. And what happens when everybody owns something is no one owns something. And so you see this rash of data breaches, all of which are many of which, I shouldn't say all, many of which are just the result of human error. So so I think think we are are very much looking at ways to build privacy capacity, and that could look like hiring a chief privacy officer. I think especially for a large state education agency or for an urban district that are well resourced and they're serving hundreds of thousands of students, should you have a chief privacy officer? Yes. Absolutely. With that said, you know, there are small smaller districts. There are some single site charter schools that are never going to have Sure. A chief privacy officer. But it doesn't mean that they are still not responsible for protecting student data. And so in those cases, I think we would look to more creative solutions including, you know, if you have a state chief privacy officer, part of their job probably should be to make sure that all the districts and schools in their state have at least a baseline understanding of privacy. And I know, when I was in Louisiana, I passed, a restrictive student privacy bill. I traveled the state training, districts on on privacy including how to comply with the law but also about privacy more generally. And the law didn't require me to do that, but I did it because it was the right thing to do to make sure they could protect student data. So yeah. So privacy capacity is a is a huge issue. The structures have not really changed even though the demands have changed, but the time to do that is is now.
So I'm gonna give you one last question. Advice to parents because I feel that a lot of times parents are the front lines of navigating the, you know, data driven educational system for their kids. So any advice to parents for how they could help their kids in this kind of data driven world or educational system they live in? Yeah.
I'll start, but I would love to hear from Hannah and Adarsh. I think, my advice to parents right now is to know your rights. I think because there has been a lack of training and a lack of awareness around, about what what some of the privacy requirements are, I think that sometimes awareness on the part of schools and districts. So I think that is something I care a lot about, is making sure that right now, there are things that parents have rights to make decisions about, and they should know what those what those things are.
I I would just add on to that, for parents to just be generally aware of, what new technology is being used in classrooms, how the technology is being used, what it is being used for. I just generally I mean, it's just I don't know that. I mean, I know, knowing your rights and knowing of, what what I do have under federal laws and state laws is important, but at the same time, the technology that's being used in, you know, and why it's being used. It's it's actually very important questions to ask, teachers or schools. Yeah.
Yeah. I don't I don't have too much to add to that. I think those are sort of the big ones. Just and just don't don't be afraid to ask questions about the way what exactly what data they're collecting, how it's being used. Yeah.
Great. Great advice. Well, you are quite the talented trio and charming trio. I told them they had to be charming to be on the podcast, and they did not let me down. Thank you so much for joining Tech Talk. Look forward to seeing all that you accomplish. Thank you, Brian. Thank you, Brian. For having us on this. And that is it for this episode of Tech Talk. For the very latest on what CDT is doing on student privacy and data issues, follow us on Twitter, like us on Facebook, or visit our website cdt.org. I'm Brian Waslowski. Thanks for listening.
