Facebook, College Summit & the College Knowledge Challenge

If you ask eighth-graders if they are going to college, 95% of them say yes. But gaps between high- and low-income students develop throughout high school. By age 24, 82% of high-income students earn a bachelor’s degree. The completion rate for low-income young people plummets to 8%. Here’s a look at the numbers at key points in time:

High Income Students Low Income Students
Say going to college – 8th grade 95% 95%
Say going to college – 10th grade 80% 60%
Graduate from high school 94% 70%
Begin college after high school 84% 41%
Finish college by age 24 80%   8%

Last year our team worked with a partner to create a picture of how these numbers map with dynamics that support or inhibit students reaching the college aspirations they all hold in middle school. You can explore a prezi of it here.

We support grantees and partners working to change these numbers in many ways. One project we’re part of aims to create new ways to help young people better navigate the college-going process. We partnered with College Summit and Facebook to launch the College Knowledge Challenge (CKC). My teammate Emily Dalton Smith designed and leads this project (and many others).

CKC is a $2.5m competitive grant fund for developers of Facebook applications aimed at making the college-going process more transparent, collaborative, and easy to navigate for low-income and first generation students. The goal is to spur the development of creative apps that make the most of students’ participation in the world’s largest social network. The CKC was open to for-profit and non-profit developers with ideas for great apps to help students in one of three ways:

  • Helping students build, test, and implement personal academic pathways that grow out of college-career aspirations and are supported by informed decision making.
  • Helping students build social capital and a college-going peer group.
  • Rectifying information asymmetries in college admissions, financial aid, and college selection processes that disadvantage low-income and first generation students.

Last September, CKC kicked off with a day-long hackathon at Facebook headquarters with more than 100 developers, engineers, high school & college students and College Summit mentors:

facebook hack

The hackathon generated a ton of attention for CKC, and College Summit followed up with webinars and other resources to help developers understand key barriers to college for low-income young people. Experts were available throughout the fall to talk with app development teams if they had questions.

In the end, more than 100 teams submitted ideas to CKC. Last week a judging panel of students, experts in the college-going process, technologists, investors, and philanthropists selected 21 winners based on demonstrations of prototypes. Each winning team received $100k to further develop the apps. Next fall, all will be fully functional and widely available.

The project is an example of how our team is trying to catalyze more innovators to work on the same challenges in different ways, rather than always looking for the needle in the haystack. If you get a chance to poke around on the CKC website for more info about the contest parameters and the winners, let me know what you think!

Personalized learning and accelerated student outcomes

A couple of weeks ago I blogged about our support for schools that are redesigning their instructional models in order to personalize learning for students. I listed design principles shared by these schools:

  • Student Centered: designed to meet the diverse learning needs of each student every day
  • High Expectations: committed to ensuring that every student will meet clearly defined, rigorous standards that will prepare them for success in college and career
  • Self-Pacing & Mastery-Based Credit: enables students to move at their own optimal pace, and receive credit when they can demonstrate mastery of the material
  • Blended Instruction: optimizes teacher- and technology-delivered instruction in group and individual work
  • Student Ownership: empowers students with skills, information and tools they need to manage their own learning
  • Financial Sustainability: sustainable on public per-pupil revenue within four years
  • Scalable: designed to serve many more students if it demonstrates impact

In the earlier post I wrote a couple paragraphs about “self-pacing & mastery-based credit” and “student ownership.” I received a few emails asking about specific expectations for performance at the school level, since “high expectations” is listed as a principle. We’re thinking a lot about this with our partners.

Most of the schools we support serve high percentages of students who enter with academic knowledge and skills well below the benchmarks for their grade. To help students catch up, the school teams are designing instructional models and using technologies they hope will accelerate learning gains. The new schools we are funding have a goal of achieving an average of 1.5 years of growth for their students each year.

When I said this in a keynote at a conference in October, some folks in the tweet stream said it was illogical — one year is one year, and whatever a kid learns that year is a year’s worth of learning. That’s true in one sense, but we’re saying something different. Independent assessment instruments often have an objective notion of how many concepts students typically master in a school year and at specific grade levels. When students master fewer or more than these, they are said to achieve less or more than a year’s growth.

A more practical way to think about it is this: Imagine a 9th grader arrives at high school with 7th grade reading and math knowledge and skills.  Unfortunately this is a pretty common phenomenon in low income communities. The only way for this student to graduate in four years meeting 12th grade academic expectations is to learn 1.5 years of material every year.

To evaluate whether their approaches are working, the new personalized school models we fund agree to participate in a common, multi-year research and evaluation study conducted by RAND. The study focuses on learning outcomes, implementation issues, and financial sustainability.

For learning outcomes, “years of growth” will be measured at the student and school level in math, language arts, and problem-solving using the Measures of Academic Progress (MAP) adaptive assessment, given at the beginning of the study to establish a baseline and then at the end of each academic year to evaluate growth. RAND will also collect and analyze state test scores, attendance, behavior, and graduation rates and compare them to matched comparison groups, along with non-cognitive measures of grit and academic mindset.

But back to the main idea of the post — regardless of the specific measurement instrument, what do you think about this 1.5 years of growth concept?

What teams of educators and developers can create in 30 hours

A couple of days ago I wrote about my team’s efforts to better connect teachers and developers in order to catalyze better learning applications. In the post, I mentioned last weekend’s SLC Camp app challenge and promised to post something about the winners.

Last year the SLC ran focus groups with 800 teachers from several states and distilled those conversations into 10 scenarios that teachers say are important to them. (You can find them here by clicking “key scenarios.”) SLC camp has run in a number of cities over the last several months, and each time the teams of educators and developers are asked to focus on 2 of the 10 scenarios. At the Silicon Valley camp last weekend, teams were encouraged to create prototype apps that address:

1. Interventions, flagging action, and measurement

2. Communication, collaboration, and sharing

Multiple teams formed around each  scenario. After a 30 hour codeathon and winnowing process, 7 finalists told their stories and showed their prototypes. Here’s a summary of what the three winners came up with in the 30 hour window:

1st place: 3R radar helps teachers communicate with parents about a student’s area of weakness, suggest strategies and activities for assisting, and report back on results. Prize = $5,000 cash plus $1,000 in Amazon Web Services and 4 hours with an AWS Solution Architect.

2nd place: NOTE-e-FI (from CaseNex) helps teachers track and communicate about student performance, attendance and notes. It is designed to facilitate both parent contact and collaboration among staff. Prize = $3,000 cash plus $500 in Amazon Web Services and 1 hour with a Solution Architect.

3rd place: Rapid RtL uses cohorts and smart goals to help teachers design, deliver, monitor, and assess group interventions for students with common needs. Prize = 1,000 cash plus $500 in Amazon Web Services.

You can read descriptions of all seven great ideas here.

Connecting teachers and developers

I often say that when it comes to K-12 learning technologies, we need lots of innovators working in different ways on the same opportunities. My team and I believe this can help speed up the innovation cycle for lots of great, affordable solutions aimed at tough instructional challenges.

Grant-makers, including my team at Gates, often strive to fund the “best” intervention with hopes it will produce results and be able to “go to scale.” But this needle in a haystack approach can be ineffective when it comes to early stage learning technologies. New applications and tools are rarely “right” from the beginning. They need time for testing with users and room for shifts in functionality and business model. Philanthropy can have the unintended consequence of buffering entrepreneurs from the natural give and take between product developers and customers that makes solutions better.

That’s why we’re trying to spend more of our time and capital on increasing the number of needles in the haystack so they are easier to find. I’m not sure this haystack analogy works, though I promised myself I’m not going to edit these posts much. But you get the picture. Rather than always looking for THE solution, we’re trying to figure out ways to strengthen incentives for lots of players to tackle tough problems, trusting that in the long-run customers will pick the winners based on what works. I talked with edsurge about this a few months ago and will share more of what we’re thinking and doing on this front on this blog over the next few months.

One of the projects we’ve been working on with partners for a couple of years is the Shared Learning Collaborative.  We’re funders and I chair the board.  The project is complicated, but one of my favorite descriptions shows how the SLC tech services help Mr. Thompson, a fake middle school math teacher, personalize learning for his students.

Though good ol’ Mr Thompson is fake, the scenario he represents is very real. By working with partners to solve key pain points at the infrastructure level, the SLC aims to create more incentives for innovation in applications that help personalize and accelerate learning for students. At this point we’ve talked with around 1000 educators about what they need from instructional technologies. One thing we heard loud and clear is that many applications don’t take into account the reality of the classroom — which problems are most important, how kids and teachers interact, what resources they have available.

So, we are working to link teachers more closely with developers & entrepreneurs who want to put their creativity and talent to work on solutions that matter most for kids and teachers. In that spirit, over the weekend I was in Silicon Valley at  SLC Camp with 300 educators and developers who formed teams to generate ideas and translate them into prototype apps over about 30 hours.  Here’s a team of teachers and developers at work on their idea:

SLC camp

At the end of the weekend, seven teams of educators and developers pitched their ideas and demo-ed their prototypes. Three won cash prizes. I’ll do a short post later in the week with brief descriptions of the winners. In the meantime, let me know what you think about our intention to spend more time on strengthening incentives for lots of innovators and creating better connections between teachers and product developers.

Design principles for personalized learning in schools

Over at Impatient Optimists, the official Gates Foundation site, my teammate Scott Benson and I blogged about our investments in personalized learning in schools. It’s a solid high-level overview of our approach, but the IO editorial process resulted in a bit of a formal tone.

So, I wanted to use this blog to spark a more direct dialogue about the set of design principles we look for in schools we’re investing in:

  • Student Centered: designed to meet the diverse learning needs of each student every day
  • High Expectations: committed to ensuring that every student will meet clearly defined, rigorous standards that will prepare them for success in college and career
  • Self-Pacing & Mastery-Based Credit: enables students to move at their own optimal pace, and receive credit when they can demonstrate mastery of the material
  • Blended Instruction: optimizes teacher- and technology-delivered instruction in group and individual work
  • Student Ownership: empowers students with skills, information and tools they need to manage their own learning
  • Financial Sustainability: sustainable on public per-pupil revenue within four years
  • Scalable: designed to serve many more students if it demonstrates impact

Obviously not every school we’re investing in has every principle figured out. The two that are wide-open for innovation right now are self-pacing & mastery-based credit and student ownership. We think both are important and we’re following their evolution closely. Here’s a little more about what we mean by each:

Self-pacing & mastery-based credit

As soon as students learn something and can demonstrate it, they should get credit for it and move on. Move on to the next set of concepts and skills on their learning path, and move on to deeper engagement in topics they’re interested in. No more sitting through hours of classroom instruction that’s too basic or too advanced. No more tyranny of the clock and the calendar. This is such a compelling idea, and really hard to implement in a school, a network, or a district. It requires a rethink of time, classroom organization, teacher roles, student support, and curriculum & assessment. Some school leaders and policy-types are doing and saying important things on this topic. Diane Tavenner, CEO at Summit Public Schools, recently blogged about their efforts on this front over at Blend My Learning . The folks over at the Nellie Mae Education Foundation are supporting several schools that are experimenting with what they call competency-based curriculum and credit – you can read more about it here.

Student ownership

The NYT recently chronicled the college-going experiences of three low-income girls from my neck of the woods (or rather, coastal plain) in Galveston, Texas. Even though they had done well academically in high school, the girls struggled with all sorts of challenges at college. A growing body of research and practice is focused on the kinds of factors that contribute to stories like theirs. Paul Tough wrote about it in How Children  Succeed, my favorite education book last year. Some critics of high performing charter networks say the “no-excuses” model of intensive teacher support employed to help students score well on state exams and graduate fails to develop the habits and initiative required to successfully navigate learning opportunities after high school. Many of the same school leaders are dismayed by the college completion rates of their graduates. While their completion rates are often higher than others like them, these students still graduate at a rate far below those of higher-income students with similar academic records.  A number of these schools are tapping into a body of research that might help set students up for longer-term success.

The research team at the Chicago Consortium for School Research produced a great literature review  on “non-cognitive” factors that lead to better academic performance, and Carol Dweck and colleagues reviewed the research on what she calls academic tenacity. Both have extensive bibliographies if you’re interested in reading more. For a look at what a couple of schools are doing to implement the ideas, check out the KIPP NY initiative and Diane Tavenner’s description of what they are doing at Summit.

We’re at the beginning of our learning curve on these aspects of personalized learning and privileged to be working with so many innovators. What do you think about our list of design principles? Have you seen interesting implementations of self-pacing & mastery-based credit or student ownership we should know more about?

If writing is thinking…

…blogging is thinking out loud. That’s not a new idea, but it’s appropriate here. I like to think about what I’m going to say before I say it. I’m much more comfortable sharing something I’ve written after I’ve edited it 87 times. The paradox, though, is the sooner I kick my ideas around with others (or have them kicked around by smart people), the sooner they get sharpened. And the faster I learn.

That’s the goal of Thinking Out Loud. My team and I are interacting with so many innovators through our next generation learning work at the Bill & Melinda Gates Foundation. This is a place where I’ll try to share ideas and reflections about what we are seeing and learning, before we’ve had a chance to fully make sense of it all. Hopefully this will spur dialogue with lots of folks who will help shape the ideas and strengthen the insights.