Alexa, What’s on the High School Menu Today?

When the Montour School District launched America’s first Artificial Intelligence Middle School program in the fall of 2018, many questions arose. Why middle school? Why teach Artificial Intelligence? How? (Just to name a few). But, as a student-centered and future-focused district, the thought process was not if we should teach AI, but what if we don’t teach AI? Also, why isn’t everyone teaching AI?

To better answer these questions Dr. Justin Aglio, Montour’s Director of Academic Achievement, and I met with two eighth grade students who were part of a special project that tapped into AI. The students provided great information about the “why” and “what” for learning about Artificial Intelligence and Justin added some key elements explaining how the program will grow in the upcoming years.

Tema and Aidan, two of the four eighth grade students, really played up the fact that it’s not about teaching AI in school today, but why hasn’t anyone started sooner. The “Fourth Industrial Revolution” report from the World Economic Forum points out that we need to understand our changing environment, challenge our assumptions, and continuously innovate. Schools and all institutions will need to begin to think about the impact of AI and Robotics. It’s not only jobs that will be affected. It’s our moral code; it’s our training for all learners to become informed citizens for the 21st century.

For K-12 schools that means we need to rethink how we define and evaluate learning. Montour is one of school districts who are rethinking what graduates in the 21st century need to know, understand and do. Montour believes that all students need to become data fluent. They need to know how to analyze, interpret and create data to solve problems. They need to be able to design frameworks to solve real world problems using data. Learners also need to understand the underlying processes and ethical issues behind modern technology. In Montour that’s exactly what students discover. Through a Media Arts course developed by the Massachusetts Institute for Technology (MIT) all students investigate moral issues related to Artificial Intelligence. Topics for the Montour course include: algorithmic bias called gender shades, the trolley problem, or ethical matrix design.

However, as Tema and Aidan pointed out that doesn’t explain how AI works. You need to get behind the modern 8 ball and discover how to use AI to solve a problem. We don’t want students to become just better consumers of AI; we want them to become better creative producers using AI. So what’s a problem that middle school students have on their minds? The Montour middle school students wanted to know what to expect when they head to high school. It’s one thing to read about it, but it’s something different to develop a “skill” for Alexa to explain what to anticipate at the high school level. Amazon provides a developer’s kit that the Montour students used. The Montour team first conducted a survey of their peers and then used the data to develop the questions that would be part of a pilot project with Witlingo. Each of the four students in the development team not only conducted the research, they also recorded their answers. So, today when you ask for Montour “Hey Google, talk to Montour High School,” or “Alexa, Launch Montour High School,” you now can learn about the high school program through the research and voices of Montour students.

While the middle school program has had great success, it’s not enough to just drop AI in the middle of a student’s life. What will prepare students for the world of machine learning? According to Justin Aglio you need to arouse student curiosity at the elementary level. So next year Montour will include in its elementary program Experiments with Google AI to introduce students to AI concepts and traveling AI robots around the school that interact with students.

Once you have the students asking questions and conducting research, you want to have them go further. At Montour high school students will soon have mentorships with companies, like Google or Argo. Students will have opportunities during their Personalized Learning Time (PLT) to take additional courses, such as AI4ALL’s Open Learning program.

Driving Innovation: Tech Enablers

In 2018 the Consortium for Schools Networked (CoSN) transformed the K-12 Horizon Report into The CoSN K-12 Driving Innovation Series with three reports. The reports are based on the work of over 100 educators around the globe who look at emerging technologies through three lenses: Hurdles, Accelerators, and Tech Enablers. As the co-chair of the CoSN Emerging Technologies Committee, I was selected to be part of the process. The Advisory Group engaged in several months of discourse about the major themes driving, hindering, and enabling teaching and learning innovation at schools. After each phase, final thoughts from advisory board members were distilled in surveys discerning the top five topics to feature in each publication.

Currently I’m working with the CoSN Emerging Technologies Committee to expand the work of the Advisory Group around Tech Enhancers, focusing on Analytics and Adaptive Technologies.

According to the Driving Innovation report:

Tech Enablers are tools that support smooth leaps over the hurdles and expansive changes in K-12 education. The top five enablers , which were ranked in order of closest proximity to mainstream adoption are:

To understand how Analytics and Adaptive Technologies have evolved I interviewed two key experts: Steve Ritter, the Chief Scientist and one of the founders of Carnegie Learning, and John Pane, a Senior Scientist and one of the leading educational researchers with the RAND Corporation. Both Steve and John have years of experience and have witnessed the changes in how Analytics and Adaptive Technologies have created new opportunities for a better understanding about learning and how to personalize that learning.

According to Steve Ritter, the role of analytics is changing in K-12 education with availability of Big Data. For Carnegie Learning data plays several key roles:

Evidence of student learning based on existing assessments;
Improve existing products to better identify learning issues;
Provide teachers with real-time information about student learning.

Carnegie Learning has partnered with the Miami-Dade School District in Florida and other schools to develop “LiveLab,” a real-time analytics dashboard that provides data to educators based on student progress within MATHia™ software. The dashboard identifies which students most need help so that teacher can make best use of their time. It also helps teachers better understand why students need help. According to Carnegie Learning’s website, “LiveLab highlights each student’s progress through math concepts in real-time so teachers can guide, intervene and coach effectively. Indicators and alerts help teachers assist struggling students and celebrate students for hitting key milestones in Carnegie Learning’s MATHia software.” Teachers from Hopewell School District, outside of Pittsburgh, PA, have been testing out LiveLab. Ray Smith, one of the teachers, describes the experience as game-changing. According to Ray, “Using the analytic tool provides all the student information at your finger tips.”

John Pane has examined a variety of adaptive software tools. The results are not always positive. For instance one study of the Cognitive Tutor Algebra from Carnegie Learning in 150 schools in over 50 school districts showed a significant improvement, but mainly in the second year. In a separate study of Cognitive Tutor Geometry in a single school district there was negative effect on learning. According to John you need to look deeper. The Carnegie Learning programs are not intended as 100% computer based. The software is intended to supplement small and whole group instruction. These studies measured the entire package of software, classroom implementation, training and the ability for the teacher to change their practice to work in concert with the adaptive software. According to John there is a conflict between policy and practice. The software pinpoints the need for many learners to work on prerequisite skills, but the teachers are often under pressure to “cover” the content and are uncomfortable letting the students move at their own pace. As a result they may have students use the software less, or override the software to push them into more advanced content before they have mastered the basics.

John also discerned a potential risk with the use of Adaptive Technology. If educators are not careful, they can end up tracking students and have lower expectations for student performance. He also highlighted some other challenges around privacy and security. However, even with these concerns, John remained optimistic and believed that adaptive technologies and analytical tools when used appropriately and with teacher training and buy in can provide greater opportunities for student mastery of knowledge, skills, and dispositions.

As we increase the use of enhanced technologies that provide analytics and adaptions, educators need to be cognizant of both the opportunities and the possible negative consequences. We need to be mindful of privacy and security issues, for instance. In addition, it’s very difficult to gauge the success of some technologies when we’re using measures that only look at content growth. In order to truly “Drive Innovation” we need to not only understand the Hurdles, discern the Accelerators for growth, and identify the enabling tools, we must also think what are we really trying to achieve. Is it enough for students to demonstrate proficiency on standards-based assessments? Or do we want to provide our learners with the tools for life-long success so they can continuously learn, relearn, and even unlearn in order to become creative problem-solvers, communicators, and collaborators in a global society?

Driving Innovation: Accelerators

Currently I’m working with the Emerging Technologies Committee to expand the work of the Advisory Group around Accelerators, in particular Data-Driven Practices. The CoSN Emerging Technologies Committee felt that all five themes (see graphic) were important, but for the CoSN audience, Data-Driven Practices had the greatest relevance. Plus it has been over three years since CoSN had worked on examples in this area.

According to the Driving Innovation report, data-driven activities can be defined according to this statement: With more engagement, performance, and other kinds of data being collected, schools are leveraging that data to make decisions about curriculum, hiring, technology investments, and more.

CoSN’s previous discussions on Data-Driven Practices focused on administrative issues relating to privacy, security and uses of data to inform instruction, with a major focus on compliance issues relating to No Child Left Behind. Now with the move towards student-centered learning, there’s a growing interest in looking at other ways of using data in the educational arena. The Data Fluency project at Carnegie Mellon University’s CREATE Lab is a great example of how data is now viewed as a tool for empowering both educators and student learners.

According to the Mission/Vision statement from the Fluency Project:

Fluency is a process of deep inquiry, case-making, and advocacy. Guided by shared values, we explore how technology and data can serve as tools to enhance the voices of teachers and students. Co-powering teachers and students to be “Fluent” means they can gather information, reconcile it with their personal experience, and influence public discourse. Within this framework, the focus is on creating an individual path for exploration based on self-knowledge, in the context of the world around you. While students will have access to new tools for understanding data and creating compelling media, we believe it is the Fluency process that will lift up their voices, and mold them into critical thinkers and active citizens.

In order to understand how this looks in a K-12 world I interviewed school leaders and teachers from two school districts in the Pittsburgh, PA region who are taking a lead in using data to enhance student agency – Carlynton and Allegheny Valley. The principal of Carlynton Jr/HS, Michael Loughren, introduced me to two of his English teachers who have taken the lead on the Data Fluency project – Kristen Fischer and Wendy Steiner. We don’t usually think about data projects in English. Kristen and Wendy have discovered a new approach give students a voice in their writing, oral and digital communications.

For the study of Shakespeare’s Macbeth, students now analyze the character in relation to episodes of PTSD. They have to find details (data) in the form of repeated words and phrases that supports an argument that Macbeth suffered from PTSD. For another project students had choices of expression for an an oral history project on a self-selected element of family history. The students used a different tool to express themselves – a podcast format. According to Kristen and Wendy there have been a number of benefits. Student work is now always original with no elements of plagiarism. All students are engaged and see a purpose in their writing. According to Michael Loughren the Data Fluency project has deepened and strengthened relationships between teachers and students. In addition, he has witnessed a decrease in the number of discipline problems.

At Allegheny Valley, Brett Slezak, the technology director, has seen similar benefits using the Data Fluency approach. Student voice has been amplified by allowing each student to make their case, which in turn has led to more student engagement. Brent emphasized the importance of using an inquiry-based processed. Students need to start by asking essential questions. At Allegheny Valley the essential question for one high school project was: What is air quality? Why is it important? Students used a SPEC sensor from the CREATE lab to monitor the air quality in multiple classrooms. The students then had to analyze the data and make their case. The problem required the students to “scrub” the data and visually represent what was happening. The students discovered patterns that led to conversations with teachers. The students had to develop a narrative so the data created a story. The students then had to advocate for changes within the classrooms. The students discovered how data revealed solutions for real world problems.

There were more projects that Carlynton and Allegheny Valley teachers created. In every case students voice became amplified. Data provided a way to gain insights into real-world problems. Students discovered that data can be more than numbers. Students took their ideas to new levels by becoming agents of change advocating for solutions to solve real-world challenges.