As Of Today, 90 Percent Of Washington State Schools Are Failing

1 min read

Washington State Schools that were making progress a year ago are now failing - not because anything changed in schools, but because the US Department of Education moved the goalposts.

The Office of Superintendent of Public Instruction put out a press release, but Randy Dorn, Washington State Superintendent of Public Instruction, was pretty succinct:

"By losing our waiver, we've had to do some things I think are ridiculous, stupid, ineffective and a waste of resources."

Because the USDOE shifted the definition of "progress" nearly 90% of schools in Washington State are not making enough of it. In Washington "adequate progress" is defined by having every single student in the school "pass" a standardized test - and it's worth noting that "passing" a standardized test is often defined by a cut score.

The maneuvering and posturing in Washington State shows the capricious and political side of school "success." I wonder if the folks who pulled the waiver thought about any effects on student motivation - how do the kids and families returning back to their newly failing school get served and supported by the politics here?

, , , ,

Let's Get Resources For The Ferguson Public Library

2 min read

Over the last few weeks, we have all been following the events in Ferguson.

In the middle of the protests, the Ferguson Municipal Public Library provided resources for community members, and a safe place for kids that met them where they were at. The library worked with Ferguson teachers and volunteers to set up a full ad-hoc schoolhouse until it was safe to open the public schools.

This is what libraries do. They provide a safe place where people can find out more about who they are, what they love, and who they want to become.

The Ferguson Library has been there for the people of Ferguson. I'd like to see what we can do to ensure that the Ferguson library has the resources they need to support their patrons in the weeks, months, and years to come.

On the Ferguson Library home page you will see a "Donate" button.

Use it. Often. While we all still have a lot of work to do to understand and grow from what is happening in Ferguson, a well resourced library in the middle of the town feels like a good step.

In addition to cash donations, you can also buy books for the library via a wishlist set up at Powell's. You can learn more about the creation of the wishlist at this blog post.

,

Personalized Learning and Privacy

4 min read

Personalized learning is frequently discussed, and rarely defined. People can mean many different things when they talk about personalized learning (and arguably, we need to reclaim the term) - but personalized learning can refer to software used for, at least, any of the following:

  • Learning the basics;
  • Remedial help;
  • Reinforcement and review;
  • Taking advanced coursework; or
  • Credit recovery.

Each of these uses carries a range of additional questions: who is being given online or personalized tools? Who supplies and maintains these tools? For students using personalized systems, how was it determined that software was the best approach for that student? To what extent are educational and pedagogical choices driven by "personalized" tools?

The use of these tools are all over the spectrum. There is a significant different between students at a school with a BYOD program supplementing instruction with Khan Academy, compared to large urban districts using online learning to sidestep class size requirements. Both of these scenarios are different than the approach rolled out in Detroit and the EAA, where student learning is mediated by a proprietary software system, and where single kindergarten classes can swell to as many as 100 students split between several teachers.

Which brings us to the point where it is necessary to state the obvious: show me a single school in an affluent district that is using personalized learning tools to increase class sizes. Take your time. I'll wait. If tech companies offering "personalized learning" as a solution want to have any chance at credibility, they need to expand their focus beyond just learning outcomes, and focus on learning process as well. If and when tech can be used to improve the learning process, and can provide feedback to learners that allows them to make informed and empowered choices, we'll be on to something.

This runs us directly into privacy issues. Belated warnings about an already-existing "market in technological learning tools" looking to "utilize and data-mine ... information in name of 'personalized learning'" miss the point entirely. The personalized learning experience of the affluent and the poor are radically different, and existing laws governing educational privacy weren't designed to address the consequences and implications of combining multiple data sets, or the types of data and analytics that can be derived from click-tracking, mouse tracking, pause times, and/or search strings, to name a few.

The timing and tenor of some of the more recent privacy initiatives are oddly ahistorical. Affluent school districts have been sharing sensitive student data with vendors in the cloud for years, via their LMS and SIS. Affluent schools have been using vendors to help write and store IEPs, often storing sensitive IEP data in the cloud. Yet, these long-standing practices used primarily to benefit students in well off schools continue to draw little to no attention, despite the fact that they violate student privacy in identical ways to inBloom. And we will leave the data collection and privacy policies of ETS - for many students, the well paid and self-appointed gatekeeper to higher education - for another post. While there were a lot of issues with inBloom, if the project had succeeded, it could have given traditionally underserved schools some of the basic infrastructure students in affluent schools have been enjoying for a decade or more.

So, when we talk about personalized learning, we need to begin by acknowledging that not all uses of personalized learning look the same, and that the reliance on personalized learning impacts both the learning environment and the privacy rights of students. The experience of a teacher in the suburbs worrying about bandwidth and administrative blowback if he flips his classroom is very different than the experience of a kindergarten teacher in a class with 100 five and six year olds trying to get them to complete their computerized assessments. There are privacy concerns in both scenarios, and equity issues that become obvious as we examine them.

, ,

Privacy Rights in the Education Achievement Authority

1 min read

While doing some research for another blog post, I was looking for information on COPPA, FERPA, or student privacy in Michigan's Education Achievement Authority.

It appears to be in pretty short supply.

EAA Privacy Policy - you get nothing!

, , ,

Give It Away Now: A Review of Digedu Terms of Service and Privacy Policies

9 min read

UPDATE, 10 July 2014: Matt Tullman from digedu commented to announce that digedu has updated their terms. I have not read through the new terms yet, and will update this post when I do. END UPDATE

I've been reading terms of service for a while, and I've seen some bad ones, but the terms of Digedu are just about the worst I have ever seen. If you don't want to wade through the details, please skip straight to the conclusion.

With Digedu, the issue is compounded because they list two separate terms of service and privacy policies. The text of these policies is not the same, which is a separate problem. For this review, I focused on the terms of service listed at https://app.digedu.co/ - this is the login page linked from the home page of the Digedu site. The complete text of these terms of service are included as pdfs attached to this post for reference. The second set of terms are listed of their main site, at http://digedu.com/terms and http://digedu.com/privacy.

Looking first at the Privacy Policy:

They punt - entirely - on COPPA:

the school or district is solely responsible for complying with the Children’s Online Privacy Protection Act (COPPA) as concerns students’ use of the services. If you are below the age of 13, please do not send any personal information about yourself to us other than what we request from your school or what you enter through the app

In fairness, this is not abnormal, and districts should be central on COPPA compliance. But the Digedu request to students to not share information is both hollow, and is undercut to the point of absurdity by stating that sharing personal information through the Digedu app is okay.

When it comes to personal contact information, Digedu requires a first name, last name, unique ID, and appears to also collect emails:

(W)e require a user’s first name, last name, course rosters, and an individual identification number, either designated to each student by his/her school or randomly generated by digedu. This allows the Learning Engine to populate classes with the correct students. We may also require users’ email addresses.

So, Digedu has solid contact information on each individual with an account in the site - this includes both teachers and students.

Moving on to encryption, they make the following curious statement:

We encrypt many of our services using SSL.

The qualifier "many" is odd here. Why not encrypt all their services? What services don't get encrypted? I suspect that this clause is in here because they reuse some content that they don't own, and is served from other sites that aren't encrypted. In any case, this is a curious admission. Given that they don't enforce COPPA and collect real contact info, the lack of end to end encryption is not good.

There are other unsavory aspects of their privacy policy, but in this review, I'm just skimming the detritus that rises to the top.

Moving on to the Terms of Service:

They collect a lot of information on everyone using the service. It's hard to see how this information is essential to running the services they provide. The information they collect seems like an unjustifiably egregious violation of privacy.

We may collect device-specific information (such as your hardware model, operating system version, unique device identifiers, and mobile network information including phone number).

You read that right: they collect phone numbers. Equally relevant: they collect the unique hardware ID of every device, as well as brand information.

When you use a digedu® equipped device, such as the Google Nexus 7, we may collect and process information about your actual location, like GPS signals sent by a mobile device. We may also use various technologies to determine location, such as sensor data from your device that may, for example, provide information on nearby Wi-Fi access points and cell towers.

You read that correctly. They collect location information. Given their stance on COPPA and their partial encryption, this is problematic. Using this data, devices with the Digedu mobile app installed can track the location of their user. With this information, Digedu can get pretty accurate home addresses, and a good sense of where a person goes, during and after school. This is creepy. This applies to all users - both students and teachers. Given that their stance on COPPA allows users under 13 to use their app, this data could potentially be collected on kindergarteners.

We may collect and store information (including personal information) locally on your device using mechanisms such as browser web storage (including HTML 5) and application data caches.

The way this is worded leaves this open to at least two interpretations. The first is that they store data related to Digedu use on your device. If this is the case, they should specify that this data is stored securely (for many reasons, but Apple's location debacle comes to mind). A second read of this - which is far worse - is that the Digedu app scans and stores data from selected other apps - like web browsers - to grab a copy of all sites you visit, searches, etc. If this is the case, this is a horrible overreach. It's difficult to make the case that harvesting this information is of any value to the end user. This feels like Digedu buttressing the size of their data store, in direct opposition to the interests of their users.

Moving on, it keeps getting worse: they can change the terms of service without notice, and you can accept the changes you don't know about just by logging in.

1. We can change this Statement if we provide you notice (by posting the change on digedu).
2. Your continued use of digedu® following changes to our terms constitutes your acceptance of our amended terms.

This is a common poison pill in terms of service, and Digedu's version of it is expecially noxious.

Moving on, we come to how Digedu treats content created ad shared by users. When you share content with Digedu, they own it, forever, even if you leave the site. This is especially troubling for student intellectual property (IP), which gets handed to the company forever. This is equally troubling for teacher IP - teachers can plan lessons using Digedu software, and the company now owns all of that teacher generated work. This potentially could violate contract terms between districts and teachers.

When you upload or otherwise submit content to our Services, you give digedu® (and those we work with) a worldwide license to use, host, store, reproduce, modify, create derivative works (such as those resulting from translations, adaptations or other changes we make so that your content works better with our Services), communicate, publish, publicly perform, publicly display and distribute such content. The rights you grant in this license are for the limited purpose of operating, promoting, and improving our Services, and to develop new ones. This license continues even if you stop using our Services. Make sure you have the necessary rights to grant us this license for any content that you submit to our Services.

Your personal information is Digedu's asset, and it goes with the company:

All of our rights and obligations under this Statement are freely assignable by us in connection with a merger, acquisition, or sale of assets, or by operation of law or otherwise.

Also, it sounds like Digedu can sell user information ("sale of assets") to anyone, at their discretion, with no notice to end users.

And if a point isn't covered in the terms of service? That means it belongs to Digedu.

We reserve all rights not expressly granted to you.

Conclusion

As it currently stands, digedu collects first name, last name, personal locations, and browsing history from all users in the site. They also collect emails and phone numbers from many of these users. They also collect intellectual property (lesson plans from teachers, work from students) from all users.

Digedu claims the right to use all of this information, and the right to change the terms how this information can be used without consulting any users. If the company is sold, user information is an asset in the sale.

To put this into more human terms, a student using Digedu for a year via their mobile app would have their location tracked every day of that year. Digedu would have a record of where that student went for the year, in school and outside school. Digedu would know that student's home address, their phone number. Digedu would have unlimited access and control over the students work. If the student wanted to look at what Digedu has collected they have no recourse. If the student wants to delete their information, they have no recourse. If Digedu gets sold or goes bankrupt, that student information is an asset to be transferred. And, according to the terms of service, data collected by Digedu can be sold

And, at any point during that year, Digedu can change the terms to anything they want, with no notice to end users.

For teachers, it's worse. Teachers have their location and personal information tracked just like students, but if teachers do original work that they distribute via Digedu, that work is then controlled by Digedu, even if the teacher deletes it.

In short, not only are these terms of service and privacy policy incredibly invasive, they leave students and teachers no way of reclaiming or controlling their data or intellectual property. Under these terms, exposing students and teachers to this level of intrusion in the name of learning feels irresponsible and cruel.

Digedu could easily revise these terms and still run their service. One possible explanation for terms of service this bad is an organizational disconnect between their legal team and the education and engineering teams. If this is the case, I hope they address the terms of service and their security practices quickly, and take steps to protect and safeguard user information that has already been entrusted to them.

,

Troubleshooting

4 min read

When working in software development and support, troubleshooting is part of your daily workload. The people who are the best troubleshooters often, when encountering a problem, approach it from the perspective of "what did I do to cause this problem?" Because their first step in addressing the issue involves a close look at how they interacted with the technology, they slow things down, look at their interactions, and gain a clearer understanding of how the software works as a result. Like it or not, computers are consistently, maddeningly logical - they are remarkably good at doing exactly what we tell them.

People who skip over the first part - the part where they examine their behavior in an effort to rule out human error - can still be competent troubleshooters, but for obvious reasons they are prone to committing a PEBKAC.

I was reminded of this over the last few days when, on Twitter, a woman showed up to do her job moderating a twitter chat on LGBTQ issues, and was immediately set on by people who dislike her organization. The conversation degenerated into another session where educational progressives were equal parts tone deaf, insulting, and unaware around issues of race, gender, and privilege (We could also talk about whether it makes sense to attack an organization by using a person as a proxy, but that is a outside what I feel competent to discuss in this piece).

I'm not going to recount details of the issue here, and people are more than welcome to dispute my summary of events. But if your reaction to reading this piece, or to conversations around race/gender/sexual orientation are on the spectrum of "people just need to get a thicker skin" please stop reading my piece and jump to any one of these pieces listed below, or get any of the books on Ashe Dryden's 101-Level Reader.

To re-emphasize: this list is not comprehensive. Reading the items on this list is no substitute for listening to people as they tell their story.

Over the years, I've accepted the uncomfortable reality that the main thing between me and my next mistake is time. It's liberating, in an awkward way. There's an incredible amount of freedom that comes from striving to be competent first, without the burden of always being right. As a straight, white, middle class male working in technology and education, I embody and benefit from enormous swaths of privilege. Try as I might, I have blind spots. There are a range of things that - as a straight, white, middle class male - are outside my lived experience. Fortunately, other people have the expertise and experience I lack, and I can listen to and learn from them - never to the point where my knowledge would supplant their lived experience, but enough so that, hopefully, I will understand more.

When I do make that inevitable next mistake - on an issue of race, on an issue of gender, on a technical issue, or on anything else - I hope that I'll have the grace to listen. I might get that wrong too. So it goes. But when it happens, my single aim is that, at some point in the process, I'll have the grace to listen.

And if/when I mess that up, I hope and trust that my friends will set me straight.

,

Move Out Of Your LMS

1 min read

Over the last few months, we've been working with Lumen Learning to support their mission of increased adoption of open educational resources. As part of this work, we needed a way to import course content stored in a learning management system into more flexible authoring environments. To support this, we built a Common Cartridge importer (Common Cartridge is supported by most systems) that lets us take a Common Cartridge export file and bring it into Candela, the authoring platform used by Lumen.

What you see here is an early version of the functionality; we'll be modifying and revising this as we run more imports. But the import process is pretty simple, so if you like living on the edge and want to free some of your courses from your institution's LMS, there's no time like the present.

At the risk of stating the obvious, all of this work runs on open source code. Feel free to check it out and spin up your own version.

, ,

Use Student Data To Streamline The College Application Process

3 min read

This post is a response to RFI 2014-08649.

Document Citation: 79 FR 21449 Page: 21449 -21453 (5 pages) Agency/Docket Number: Docket ID ED-2014-OUS-0040 Document Number: 2014-08649

Elements of the college application process are repetitive, and require students to pull and collate data already collected elsewhere. For some students, family circumstances can create barriers to getting this information together. Additionally, some students will be navigating the process on their own. To support students looking to transition into college, we should use existing systems to support them along that path.

FAFSA and the Common Application could be streamlined to make use of data already collected, by the Internal Revenue Service and state longitudinal data systems, respectively.

FAFSA already connects with IRS data. Information about this connection is available at http://fafsa.gov/help/irshlp9.htm. However, this connection needs to be updated to support more types of families (or, more types of tax filing, as needed by different family structures). Right now, only a subsection of students can be supported with the FAFSA data transfer tool. While this is a good start, improving and expanding the API used for data transfer from the IRS to the FAFSA should be prioritized. Additionally, the FAFSA data transfer tool appears to only be used for populating the FAFSA form - we will adress other possible supports later in this writeup.

Transitioning to the Common Appication, much of the information required for the Common Application is already stored in state level longitudinal datastores. Over the course of a student's K12 education, the information required by the Common Application has been collected and stored. Given that these datastores implement a version of CEDS, or SIF, or both, we can assume that the data coming from these systems is usable. The Common Application requires information about a student's academic history, demographics, tests, etc, that are all available via data stored in the state-level longitudinal data store. Additionally, some students qualify for fee waivers for the common application. Based on data obtained for the FAFSA, students who are eligible to have their fees waived could be identified proactively.

As students go through college, their loan obligations could be tracked against their income, and their family income. This data could be tracked against other factors (geography, school, economic conditions, family income) to help flag students who might experience difficulties staying current in paying off their student loans. Data could be used to identify students in need of support before they fell behind on their loans (using anything between loan consolidation, deferment, and - in some cases - forgiveness). Programs to support students already exist; data could be used to connect students with support in more timely and efficient ways.

Obviously, there are privacy concerns that need to be addressed as part of the implementation. As a start, connecting different data sources should only happen if a student initiates the process. Students need to own and control where their information goes. Additionally, any implementation needs to allow students to override data that is auto-completed, as machines often make mistakes that require human intervention to fix. But, given that much of the data required to apply to college is already collected and organized, we should not require students to do unnecessary repetition as part of the college application process.

, ,

Data Collection, Political Candidate Edition

3 min read

Right now, there is a very good possibility that at least some of the candidates for the 2056 presidential election are in elementary or middle school.

To meet federal accountability requirements, their schools collect detailed information about their behavior and performance. This information - tied to a unique identifier - gets stored in a state database that records detailed information about learners. All states collect this information for K12 education; 41 track information from preK through college.

Many of these state level datastores are managed, supported, and accessed by external vendors like - for example - Pearson, eScholar, and Infinite Campus.

If this candidate goes to a school that uses Powerschool, then their personal information is shared with Pearson. If this candidate goes to a school that uses Google Apps, then their information gets shared with Google. If this candidate goes to a school that uses Bing for Schools, then their information is shared with Microsoft. If this candidate goes to a school with an iPad program, then their information gets shared with Apple, and all of the vendors of the apps purchased by the school through Apple's closed app ecosystem. If the candidate goes to a school that use Clever to unify platforms, aggregate information from multiple sources flows through the infrastructure of another vendor.

All of these vendors manage information - they have to, in order to deliver the services they offer. But many of these vendors have terms of service that allow them to share information with affiliates and partners, thus expanding the flow of data to even more locations.

As of this writing, few of these data solutions support any streamlined student or parent review. Few, if any, support student input on the data points that people claim make up a "complete" picture of student learning.

It's hard to imagine a scenario where our future presidential candidates won't have their learning patterns stored with at least 10 different vendors.

As we saw with the ConnectEDU bankruptcy, all it takes is a corporate declaration to convert personal details into an asset to be sold. 

What if Bill Clinton, George W. Bush, or Barack Obama went to school in a time where the process of their learning left a long, permanent data trail? What would their data trails be worth? What would be the worth of these data trails to other politicians? Journalists? Business leaders? Other governments? Future datastore hacks have the potential to make NewsCorps' phone hacking scandal look quaint.

Now, in 2014, it's safe to say that we don't fully understand the potential ramifications of our data collection and use. And that's not necessarily bad - new technologies bring us to unpredictable places. That's normal. But, given the amount of data collected about students, it's unthinkable that we continue to exclude students and parents from commenting on and controlling their stories, as recorded through their data.

,

Data Collectors and Vendors: Don't Ask Us To Trust You. Show Us That You Care

5 min read

If you are collecting data on students, students need to be able to see, interact with, review, comment on, and dispute data points to which they have objections. Parents deserve these rights as well. Anything short of full, complete transparency - where students and parents can see the full range of data collected, and interact with it, and learn from it, and interject when the implications suggested by the data are wrong - won't be sufficient.

Whether you see this as a systems design issue with pedagogical implications, or a pedagogical issue with systems design implications, the starting points look the same.

If you are building or working with a data system, it needs to have these two components:

  • Student dashboard - a student should be able to see everything that is collected about them. More importantly, the application should have a mechanism that allows students to comment on, review - and in some cases, remove - data points, or assumptions based on data. More on this later.
  • Parent dashboard - because many students are minors, parents have the legal right to review data collected. Additionally, rights to review some data is guaranteed under FERPA. Really, there shouldn't need to be much - if any - difference between the student and parent dashboard. If there are significant differences between a parent and student view, those differences should be grounded in clearly articulated reasons that are of direct benefit to the learner.

If you are buying a data system for use within education, ask to see these components before you sign a contract. Don't accept answers like, "It's on the roadmap," or "We'll work with our design team to get this in place." It's 2014. This isn't rocket science.

Current data collection efforts treat students as observed objects, not as empowered agents. Additionally, the thrust of much of the data conversation centers around making learning more efficient - and this goal has some merits. But, if we shifted the definition to include and support making all participants in learning - students, teachers, parents, school administrators, policymakers - more informed and empowered, we are on the path of more comprehensive improvements. Efficiency is a short term goal, where education is a lifelong endeavor.

Current data collection efforts - and teaching strategies based on sound data use - don't take into account that we are self-aware. Student insights into how they learn - and, more importantly, creating within students the habit and skills to reflect critically on how they learn - should be designed into systems from the ground up. If our data collection efforts remain focused on building systems to put in front of students, and to report on how students interact with these systems, our thoughts around data, learning, and adaptive systems will remain rooted in the mindset of a student as an object to which learning happens. By shutting student voice out of the system - and no, "student voice" is not the same as taking a survey, or responding to prompts - we send the clear message that student input on their learning is, at best, secondary to the process of learning.

So, data collectors and vendors: don't ask us to trust you. When you collect information without telling us what it is, or why, it's not reassuring. It's creepy. I don't want to hear how your observations and analytics will revolutionize learning. I don't want tech companies, vendors, or really, anybody, laboring under the misconception that they have what it takes to "save" or "disrupt" education. While the intentions might be good, it's patronizing as hell.

Discussions around privacy look very different when we can all see exactly what is being collected and shared. If you are a vendor selling a data-driven service and you think that what you offer is amazing, then have the confidence to let us in so we can see exactly what you collect, and how you use our information. If you want us to agree to data collection, you need to agree to allow us to disagree. If someone doesn't see the benefit in what you offer, you need to have the confidence in the value of what you offer to respect their opinion. In short, my learning - or my kids learning, or the learning of anyone I have ever taught - is for the benefit of the learner first. If you think you can help people learn more with a data driven product, great, but you also need to respect the reality that not everyone will want your help.

If you want us to trust you and your intentions, the onus is on you to design systems in a way that allows for and respects individual choice.

These discussions take place against a backdrop where big data - more information about more things - is becoming increasingly "normal." Outside education, data brokers operate in a largely unregulated gray area. Inside education, vendors with millions of data points are not doing adequate planning to safeguard data. The price of public education should not be privacy. If vendors want to reassure us of their intentions, then please, work with us to make sure sure that learner privacy is respected, and that the learner voice trumps other concerns.

Don't ask us to trust you. Show us that you care.

, ,