Jose Ferreira has quickly become one of the preeminent voices in education data innovation. Ferreira took time to discuss the role data plays in education and what advancements will excite administrators, parents and teachers. He also expands on how administrators can better evaluate adaptive products.
Dr. Berger: There have been turbulent waters in education and none more controversial than the inclusion of data. Help me separate fact from fiction so that school leaders at the district and campus levels have an opportunity to judge applicability of data driven solutions on merit and not propaganda.
Jose Ferreira: Data have enormous potential to help students and teachers better understand how students learn. Publishers today are using new technology to create dynamic products that adapt at every second to each individual’s needs and provide real-time reports for administrators, teachers, and parents showing how best to support each student.
Innovations in data can help improve educational experiences for everyone. Just a few benefits:
Increase transparency into product efficacy.
In a digital world, it becomes possible to measure data to discover exactly how lessons, apps, or games really help drive student understanding. Soon, education companies that create the tools and textbooks students use will have to prove how much they actually improve learning. This means publishers will create ever more effective materials to compete in the marketplace, students will get better outside-of class materials, and both teachers and administrators will have ever more insight into student progress.
Enable students to experience truly personalized learning.
Classrooms are filled with students who have different needs, come from different educational backgrounds, and have different attention spans and interests. Knewton adaptive learning uses data to assemble the exact pieces of content that best improve a student’s progress toward learning goals, at a particular moment in time.
Help teachers target interventions and better serve each student.
Machine learning can provide teachers with analytics that predict student performance, measure engagement, reveal more precisely how well a student understands a specific concept, and more. These metrics help teachers shape their curricula and offer more targeted extra help.
Reduce the need for standardized tests.
Real-time analytics powered by adaptive learning can predict performance and continuously measure student proficiencies at the concept level, as an organic part of the learning process, rather than interrupting learning as standardized tests do. Instead of teaching to high-stakes tests, teachers can focus on each student’s individual needs throughout the school year.
There are a few common fears around the introduction of data into education. In particular, many worry that students’ privacy will be violated. They’ve seen big consumer web companies push the boundaries on consumer privacy; why should edtech be any different? Some companies’ ad-supported models mean they can only make money by selling user data. Edtech companies, who charge by subscription rather than advertising, have no such conflict of interest. In fact, these companies choose subscription models in order to avoid this very conflict. They have a strong interest in protecting student information.
Others worry that data will replace teachers. They won’t. Data are only additive, like x-rays in hospitals and instant replay in sports. Data add concrete information to a teacher’s observations and intuition, but they will never replace experience, personal relationships, and cultural understanding. Think Moneyball. Statistics haven’t replaced talent scouts — they’ve just armed them with more than intuition.
RB: What part of the argument against data should we be wary of and why?
JF: There seems to be a layer of society that fears data, and will use any argument to attack it. Many of those arguments are illogical, like “you’re reducing my child to a data point.” Students do produce huge amounts of data as they study. By analyzing these data, we can help teachers better understand every student, and help students learn faster and better with much improved homework materials. Not taking advantage of this opportunity because it “reduces” a child to data makes as much as sense as doctors not prescribing antibiotics to a child with strep throat because it “reduces” her to chemicals and biological processes.
It’s natural to question new technology, but there’s real danger in letting fear of change automatically prevail. The quality of a country’s education system is closely tied to just about every other sign of national well-being — health, GDP, per-capita income. Ignoring valuable tools to improve learning is doing a disservice not only to students, but to society at large.
RB: Let’s shift the discussion, Jose, to the advancements in data and what they mean for schools and the students they serve. What excites you the most about our ability to differentiate instruction, remedy academic roadblocks and engage students earlier in the process?
JF: Soon parents, teachers, and administrators will have more visibility into how students learn across subject domains and over time. They will be able to see exactly where students have struggled in the past, and exactly what they should work on to address those areas of weakness. This will help eliminate the “Swiss cheese effect,” where students who have persistent holes in their knowledge get passed through to the next level without ever filling in the gaps. Unlike cheese, though, learning gaps get progressively bigger. For example, a student might pass an algebra class, but still struggle with basic equations. They will inevitably then struggle in future math courses, but also in other quantitative courses like chemistry, physics, or economics. Data can make it possible for schools to more easily diagnose these gaps and correct them earlier on.
Data have enormous potential to help improve education for the world. But innovation and technical integrity mean nothing if new tools don’t ultimately work for students and teachers. At Knewton, we understand the importance of engaging teachers in the product development process and implementing effective professional training when implementing new technology in schools. We have a whole team of former educators who consult on pedagogy and help us develop technology that meets teachers’ needs.
RB: I recently interviewed the former CIO of LAUSD, Ron Chandler, who talked quite openly about the failed district-wide iPad implementation project that many have used to support their position that schools struggle to make sound purchasing decisions. Chandler went on to talk about the lack of cohesive messaging and the negative impact on creating buy-in within the district. How can those working with education data prevent similar outcomes and what lessons have been learned from LAUSD’s project and the very public demise of inBloom?
JF: inBloom did two important things incorrectly. First: if you’re going to collect and attach sensitive data to a particular student’s personal profile, you need to have the explicit permission of that student (or, if a minor, her guardian). That way parents can say, “Hey guess what? I don’t want you telling my child’s future teachers about her past behavioral problems, because they might pre-judge her.” Second, if you’re going to receive data about students, the benefits have to be clearly communicated to students and parents. After all, it’s their data.
Education organizations, whether they are for-profit or not-for-profit, should never sell student data. They should only share data when specifically requested by the student (or her guardian). They should avoid holding unnecessary personally identifiable information. They should be transparent about and vigilant with the student data they receive.
RB: What aspect of data-driven instruction should school leadership be aware of, as an upcoming development, that might impact not only their views, but patterns of purchasing solutions in the near future?
JF: As education markets become adaptive, it’s important for administrators to understand how to evaluate adaptive products. Today, many educational products claim to be adaptive. But most of these tools consist of simple rules-based decision trees which use a few checkpoints to guess each student’s path. For example, the decision tree might specify, “Students who get 7 out of 10 questions right on this fraction division exercise can move on; otherwise give them more questions on dividing fractions.” These products are inaccurate and arbitrary. They market the buzzword of “adaptive learning” without actually doing any of the hard technical work of building it properly.
Continuous adaptivity, on the other hand, means the education product responds in real-time to each individual’s performance. Rather than analyzing a student’s knowledge at just a few points during the course, it does so at every moment in time — ensuring that each student gets exactly the instruction she needs, exactly when she needs it. This requires large pools of “normed” content and results in accurate proficiency estimates of what a student knows at any moment. This means a continuously adaptive product can effectively infinite number of possible paths through the material. Instead of funneling students through a pre-determined set of available paths, a truly adaptive learning product constantly updates a student’s path through learning material.
Please connect via LinkedIn and Twitter to suggest interview guests and story ideas. If you are the idea you want to float…by all means connect and pitch me your perspective!
Jose Ferreira is the founder and CEO of Knewton, the world’s leading adaptive learning company. Jose earned his MBA from Harvard. He was formerly an executive at Kaplan, a strategist for John Kerry’s 2004 presidential campaign, and most recently, a partner at New Atlantic Ventures (formerly Draper Atlantic), investing in new media and SaaS companies. He currently serves on the World Economic Forum Global Agenda Council on Education & Skills, as well as the advisory boards for Laureate International Universities, Cambridge University Press, and the ASU/GSV Education Innovation Summit.
Dr. Berger is a global education media personality featured on the Core of Education, AmericanEdTV, in Ed Tech Review India and on RFD TV’s Rural Education Special. Dr. Berger also serves as Vice President of Education for RANDA Solutions an education software and data management firm named three times to the INC 5000. As an industry personality Dr. Berger has interviewed Ministers of Education, leading voices like Sir Ken Robinson, U.S. Secretary Arne Duncan, AFT President Randi Weingarten and other global thought leaders. Dr. Berger is a guest lecturer at Vanderbilt University and resides with his wife and two children in Nashville.