Perspective

The Power of Plumbing: Infrastructure Supports Electronic Activities

By Robert Tinker

Once every student has a networked computer, what will instructional materials look like? They better be more than simply texts-on-screen. The material should be compelling and thought provoking, and should take full advantage of the flexibility and computational power computers offer.

This issue of @Concord features five classroom-ready lessons from our work that represent a sneak peek at what could replace texts: free computer-based materials that provide interaction, guidance, assessment, and flexible formatting. And, unlike texts, these materials are entirely electronic and can be distributed freely to colleagues, kids, and parents. After using these, textbooks will seem impossibly antique.

Monday’s Lesson, “Motion Two Ways,” illustrates how hands-on experimentation can be incorporated into electronic media. It is one of hundreds of activities using probes and sensors that we and participants in our workshops have produced. This particular activity is a great way to introduce force and motion. A motion sensor is needed, but we give teachers the option of using most commercial motion detectors or one that can be built from an inexpensive DC motor. Constructing the sensor is itself a valuable introduction to electronics, magnetism, and IT careers.

Tuesday’s Lesson, “The Color of Light,” demonstrates how useful it is to embed different models in the same platform. This lesson is a sequence of three short model-based activities that address the atomic basis of color. The first uses a simple model built using NetLogo to get students thinking about the role of absorption and re-emission in determining the color of objects. The second uses our Molecular Workbench to focus on the way atoms can be excited by light and then decay, releasing heat or light. The third uses an open source Java model developed by the Physics Education Technology (PhET) group at the University of Colorado. These three activities were easy to develop using a web-based template we are developing.

Wednesday’s Lesson, “Friction,” demonstrates how electronic media will adapt to individual student differences, the goal of our Universal Design for Learning (UDL) project. Soon the graphs and models will be able to explain their important features. If students have earphones, the text and, eventually, the graphs and models can be vocalized. Colors, layout, and the language used can be modified. When there are questions for students to answer, help will be adjustable for top-down or bottom-up thinkers and the kind of scaffolding will be variable from highly specific to open-ended. Eventually, we will provide different paths through a learning activity and entirely different activities for the same instructional goals, but designed for students with different strengths.

Thursday’s Lesson, “Genetics,” shows how guided inquiry can combine a game-like environment with serious, difficult content learning. Games alone can be unproductive and waste time, but these problems can be overcome by guiding students through explorations of challenges, interesting interactions, and a simplified system modeled after the real thing. This lesson is part of a much larger set of activities that introduce most genetics concepts—including evolution—at multiple levels from molecules to populations.

Friday’s Lesson, “Chemical Reactions,” allows students to make sense of the notation used for chemical equations by getting a feel for chemical reactions at the atomic level. The activity is one of 24 under development as part of the Science of Atoms and Molecules project. The goal is to supplement introductory physics, chemistry, and biology courses with a coherent treatment of atoms and molecules. The activities all feature our incredible Molecular Workbench system that generates highly interactive models of atomic-scale phenomena.

OPTIMISTS have been predicting the demise of textbooks for a long time. Why has it taken so long? Why think it might happen now? One reason has to do with plumbing—of a software kind—and investing in an infrastructure.

Around 1980 we made the first probeware—software that read sensors of various kinds and graphed the results in real time. In those early days, our software was designed to be a pure tool, modeled after a word processor or spreadsheet. We believed that students would master the tool and use it for independent explorations. Naturally, teachers and learners needed instructions, so we produced print materials to complement the tools. As the tools got more sophisticated, the options multiplied and the instructions ballooned. We resisted mixing our “constructivist” tools and models with electronic instructions, because that sounded too much like “instructivist” computer-assisted learning (CAI) and that was considered a bad thing.

Theory finally collided with reality in 1995 with GenScope, a genetics software tool. Paul Horwitz’s careful classroom research found that kids could master GenScope by treating it like a game, but that this did not improve their scores in standardized tests of genetics. The obvious fix was to create a “surround” for GenScope called Pedagogica. This was a programmable environment that provided all the pedagogical support needed: scaffolding, guidance, instructions, and assessments. The result was BioLogica, with the same genetics software now embedded in a series of learning activities that provided an environment for guiding student learning through exploration of various aspects of genetics. A script that Pedagogica runs determines the actual content, whether it is simple dominance rules, or how to interpret a Punnett square. It also determines how open-ended or didactic the activities are.

There are many advantages to this approach. Students do not need to learn the tool before using it. Their attention is not on the tool as such, but on the content, in this case genetics. Teachers do not have to master the tool either, so extensive professional development is not required prior to use. Many different activities can be based on the same tool, but aimed at different content and grades, and employing different instructional strategies. With a good authoring system, developers and teachers can easily change an activity based on classroom feedback or make multiple versions for different students.

A surprising amount of software “plumbing” is needed to make this work. Like plumbing, this software is relatively invisible, but essential. SAIL, the Scalable Architecture for Interactive Learning, the latest version of Pedagogica, grows out of TELS, a collaboration that includes programmers at the University of California, Berkeley, and the University of Toronto. SAIL is the plumbing behind the first four lessons that links together the various parts of each lesson—from text and images to models and data collection, plus open response areas for students to type or draw—and connects all of them to a server.

SAIL is the key to a set of functions that all educational tools and models need:

Editing. Simplifying the creation and modification of highly interactive learning materials by non-technical educators.

Deployment. Delivering materials that students need, when students need them, suitably modified for their level and learning style.

Assessment. Tracking student progress and thinking by monitoring their choices and responses and making this information available to researchers and teachers.

Extensibility. Making it easy to add new functions and link them into the system.

SAIL is modular, free, and open source. Perhaps the advantages SAIL offers will tip the balance in favor of electronic media.

Robert Tinker (bob@concord.org) is President of the Concord Consortium.