The solar-car workshop at Boston’s Museum of Science, three groups of parents and children are trying out model automobiles they have built. Suddenly everyone rushes back to a workbench to change wheels, adjust the tension on a rubber-band pulley that connects an electric motor to the drive axle, and make other alterations. Meanwhile, another team is attempting to get its refined model to run a test race course in 12 seconds. When it comes close-12.8 seconds-the kids cheer. And in another corner, after 35 minutes of nonstop work to correct a model whose wheels at first spun only when the car was held up in the air, an 11-year-old boy looks around and finds someone he knows. “Come look!” he cries.

“I made it work,” and explains what he discovered.

The solar-car workshop represents the cutting edge in exhibits at science centers. Across the country, exhibit developers are creating open-ended experiments designed to stimulate in visitors the kinds of thinking scientists employ. These reformers are discounting the traditional notion that museums can teach little about the scientific process during a typical brief visit. Developers even maintain that their work can serve as a model for enhancing formal science education.

From Push Buttons to Blended Faces

Natural-history enthusiasts began setting up museums hundreds of years ago to display rare and wonderful objects. Such displays even today fulfill an important role, allowing visitors to observe and hence learn about intriguing aspects of unusual objects: rocks, minerals, animals, fossils, and more. And such exhibits still flourish, as the American Museum of Natural History in New York City demonstrated two years ago with the successful opening of its renovated dinosaur halls.

But interactivity-the opportunity for visitors to operate or otherwise manipulate parts in exhibits-was lacking in the first museums. Then, early in the twentieth century, U.S. museum curators brought back word of the approach of the Deutsches Gesundheit Museum in Munich, which was offering classroom-type demonstrations of physics and chemistry, behind glass, in pushbutton-activated exhibits. Three new U.S. science museums-the Franklin Institute in Philadelphia, the Museum of Science and Industry in Chicago, and the Museum of Science in Boston-introduced not only push buttons but live demonstrations by museum staff to stimulate interest in the ordinary behavior of all kinds of objects.

The next wave in science exhibits arrived with San Francisco’s Exploratorium in 1969. This museum’s exhibit developers, who were scientists and artists, took a more hands-on approach still. Instead of simply pressing buttons to activate a demonstration, visitors can more intimately control variables in an exhibit. In one such offering designed to teach people about wave properties, museum-goers can adjust the frequency and volume of sound waves produced by a speaker attached at one end to a horizontal glass tube containing a small amount of liquid. Visitors learn how to play with the controls so that as the air in the tube resonates at particular frequencies, liquid drops fly up to reveal standing waves.

By relying on objects that look as if they have been found in, say, a basement, the Exploratorium’s developers have also tried to create rough-looking exhibits that suggest working prototypes rather than the finished showpieces displayed by many other science museums. Moreover, the museum’s approach is to cluster these items into themes, such as refraction and polarization, with sets of instructions that lead visitors through a series of steps, so that people walk away having gained a logical understanding of some particular phenomenon. For instance, the tube demonstrating standing waves might be located near a machine that mechanically produces a variety of waves, shown by rising and falling ping-pong balls.

The Exploratorium’s approach has been so good that it has influenced science museums everywhere, so that almost everyone working in science centers has come to link the word “interactive” with the phrase “educationally successful.” But although museums have copied the Exploratorium’s approach, they have done so with varying degrees of success. Exact replicas of the San Francisco institution’s exhibits, such as a “half-silvered” mirror that allows two visitors, each on opposite sides, to see their facial features blended in one image, have tended to work very well. (Half the light hitting this kind of mirror reflects back, while the other half transmits through.)

But some exhibit developers have failed to understand that interaction best enhances learning if it engages visitors directly in a phenomenon and have too often incorporated gratuitous and hence useless interactivity in new exhibits. Consider how some 20 years ago, developers at the Museum of Science in Boston proposed teaching bird ecology with the aid of a pinball machine-an object popular with children. The idea was to repaint the game surface so that different areas represented survival factors in the life of a bird. But while a pinball game may convey some lessons on mechanics and trajectories, it does not intrinsically teach anything about birds or how scientists learn about birds. The new labels on the machine would have conveyed all the exhibit had to offer on birds-differing little from museum exhibits in glass cases. And clearly no one was going to read the labels while trying to keep the balls from going down the hole. Fortunately, the developers dropped the idea.

A Twist on Visiting Scientists

The Exploratorium has continued to produce one exhibit after another that uses creative interactions to demonstrate specific scientific phenomena spectacularly well. But a few exhibits the museum has developed over time go beyond this approach to risk open-ended outcomes. Consider, for instance, a table with a central light source embedded in a metal cylinder that is outfitted with slits. Attached to the table by cords are various types of small mirrors and lenses that visitors can experiment with as the light radiates across the table. While such exhibits may not seem as effective in teaching specific content-based lessons, developers elsewhere have recognized that these activities can stimulate visitors to carry out the scientific process on their own, and hence can be quite valuable educationally. Inspired, developers have begun to push the design of such open-ended exhibits further.

The result has been exhibits that resemble working laboratories, where visitors can pursue their own short research projects and find their own answers to the questions they themselves pose. The thinking is that science is not just about, say, the physical, chemical, or biological attributes of the world around us, but a process for learning about the world. No matter what scientific information visitors may take home, the most valuable lesson an exhibit can convey is that process.

The first museum to make a full-fledged stab at the “do-your-own-research” approach was Science North, which opened in 1984 in Sudbury, Ontario. That museum-or, more properly in this case, that science center-includes interactive areas akin to a scientist’s work space. There, visitors can explore various topics with a corps of scientist-teachers. The activity of conducting research is what counts. In Science North’s “Swap Shop,” for instance, people can both bring in and examine others’ small displays of natural objects-such as rocks, animal bones, and shells-using apparatus such as microscopes, tools, maps, and charts.

This approach has sparked widespread excitement among museum developers, and others have started coming up with similar approaches. One of the first to follow up was J. Shipley Newlin, a developer at the Science Museum of Minnesota in St. Paul, whose Experiment Gallery focuses on “experiment benches” where one to three visitors can have exclusive use of apparatus to choose which of many variables to control, devise their own experiments, and hence experience many outcomes.

One of the most successful exhibits is the Electricity Lab, which allows visitors to explore basic electrical components such as resistors, light bulbs, motors, capacitors, diodes, and switches. These click together and can be attached to copper wire to form different kinds of low-voltage circuits. (The circuits are protected so damage does not occur. And to provide more insurance against problems, a staff member stays in the Experiment Gallery.) In a more traditional interactive exhibit, a few components might be bolted to a table, with visitors able to change a couple of variables. But in the Electricity Lab, visitors actually build the circuits and decide what to include.

Evaluations have shown that visitors stay longer at experiment benches than at the museum’s more traditional interactive exhibits-as much as 19 versus 6 minutes. But the open-ended nature of the experiment-bench model also creates some problems. With many options available, some visitors have difficulty figuring out what to do. They have said they need better instructions. The Science Museum of Minnesota has dealt with this concern by adding “Experiment Cards” to the labs that suggest tasks with easy, moderate, and challenging levels of difficulty.

Another center that has taken open-ended exhibits to a new level is Portland’s Oregon Museum of Science and Industry, which in 1993 opened “Engineer It!”-an exhibit that gives visitors firsthand experience in designing, building, and testing an object such as a model boat or airplane. Budding engineers can select from numerous parts in bins and refer to examples of how they fit together. They can then tether their paper airplanes in the wind tunnel to see if they will fly, modify the design of model trucks and test how much drag they encounter while moving down a “highway,” or construct buildings on a table that shakes to see if the design is earthquake-proof.

At this museum, evaluators have found that most visitors pay relatively little attention to the instructions, designing by trial and error instead and learning a great deal from other museum-goers by mimicking their designs and making improvements to objects left behind. In so doing, visitors behave like scientists, who similarly build on one another’s results through personal contacts as well as published papers.

Boston’s “Activity Centers”

The largest-scale effort to develop exhibits that focus on scientific-thinking skills-a series of six “activity centers”-is under way at the Museum of Science in Boston, the institution where I work. This past March we inaugurated the activity center “Investigate! A See-for-Yourself Exhibit.” This center focuses on skills associated with conducting an experiment: asking questions, formulating hypotheses, planning and carrying out a procedure, collecting data, analyzing evidence, and drawing conclusions. At the entrance, visitors encounter a sculpture of a girl standing atop a stack of bedroom furniture, just as she’s about to drop a softball and a golf ball to see which will hit the ground first. This evocation of Galileo’s apocryphal Tower of Pisa experiment symbolizes the exhibit: conducting experiments on one’s own.

Behind the sculpture is a wall of questions and a room devoted to the first step in conducting an experiment-asking a question that can be answered. This room, intended to encourage reflection, contains intriguing objects identified with questions rather than answers: “What is it made of?” “What could you use it for?” “Where did it come from?” “Was it once alive?” Visitors can add their own questions to the wall or their thoughts about the objects by writing these on index cards and posting them for others to see.

From that central room visitors can head in different directions. Those who turn to the right enter a brightly lit room with several experiment stations. Using a wire temperature probe attached by a cable to a computer with a colorful display screen, investigators can measure the temperature of various items as they consider whether, say, Styrofoam keeps a drink hotter than a paper cup does, and how quickly fans blowing on a hot cup cool it. Visitors can also devise experiments involving their skin temperature, such as one to determine whether one person’s hand is warmer than another person’s. “Challenge Cards” offer some initial ideas for research questions, but the exhibit becomes truly successful when patrons start pursuing inquiries developers didn’t consider.

Soon after we opened the exhibit, we noticed one such set of inquiries at the “Drop Stop,” which lets visitors recreate the Galileo experiment. They put all sorts of objects in two metal buckets and push a button to transport them 12 feet into the air. Pushing another button makes the buckets open and drop their contents at the same time. A row of sensors connected to a computer tracks the falling objects and indicates their locations at different points in time.

What we didn’t expect to generate so much interest is a safety-interlock mechanism that prevents each returning bucket from accidentally hurting someone. Visitors who hurry to open the clear plastic door before the bucket is back at the bottom find it stops where it is and resumes its downward movement only when the door is again closed. This discovery has sparked a series of activities involving the interlock: visitors investigate how fast it works, if they can beat it, and where the electrical contact is that makes the gizmo work.

The two most elaborate areas of the activity center-and the ones at which visitors spend the most time-are the solar-car workshop and the “Midden Mystery.” The workshop’s main activity is similar to part of the Oregon museum’s exhibit “Engineer It!” Visitors assemble model solar cars at a long workbench with room for many people to work at once. They can experiment with wheels of three sizes. They can adjust the tension on the rubber-band pulley that connects the solar-powered electric motor to the drive axle. They can move the motor and make a front-wheel-drive or rear-wheel-drive vehicle. And by turning their vehicles upside down on a bench with embedded lights-to activate the solar cells-they can check at any time how the wheels spin. Visitors can also take their cars to a test track where variable light controls and automatic and manual timers allow an entire group to have a role in a test run. For people who are interested, challenge cards suggest activities beyond how to make cars simply work or go as fast as possible. For instance, one idea proffered is to figure out how to run the course in exactly 12 seconds, which usually means slowing down the vehicle.

“Midden Mystery” focuses on drawing conclusions. A midden is an archeological garbage dump. Ours, of course, is simulated; it resembles a large sandbox filled with crushed walnut shells. (They don’t stick to skin and clothes as sand does.) Questions such as “What did the inhabitants of this site do here?” are placed nearby. As visitors brush away the “sand,” they find in it shells, animal bones, arrowheads, and other stone tools. THey also discover some fiber-cast objects, such as of an animal skeleton, embedded in harder layers below.

Near the dig site are workbenches for measuring, recording, bagging, and posting information about the finds. Visitors can speculate on what the objects may have been used for and can take them to tables with reference collections of mollusk shells, small mammal bones, and various stone tools, and they can compare their ideas with experts’ opinions that have been left on answering machines at two “curators’ desks.” Finally, visitors can “publish” their theories using a computer terminal outfitted with a miniature video camera and a microphone, and can learn about other museum-goers’ results and conclusions.

We have purposely built many such reporting mechanisms into “Investigate!” We want visitors to be able to leave behind their questions, speculations, observations, measurements, and conclusions for others to learn from. The idea in that activity center is to supplement the educational voice of the museum to illustrate the idea that scientific truth is determined not by authority but by evidence. We are finding that only a small fraction of visitors record detailed conclusions, but nearly everyone contributes answers to limited questions posed on computer terminals throughout the exhibit. These answers become part of growing databases. We are also learning that strangers frequently talk with one another about matters far more detailed than “What does this do?” Questions such as “How did you get this to happen?” arise.

The activities in “Investigate!” actually constitute the Boston museum’s second activity center. The first center was much simpler, representing our initial foray into the new approach. The activities in “The Observatory,” which tend to be less complex, are designed to encourage museum-goers to become aware of their observational skills. In one area a visitor can operate a remote camera aimed at a terrarium to take a close look at, say, a lizard’s eye or a giant Madagascar cockroach. Another activity is designed for two people to play with emitting and detecting sounds from any of 12 overhead speakers. We are now drawing up plans for the other four activity centers.

Other science museums are moving in a similar direction. When “Investigate!” opened this year, staff from 21 other science centers attended a workshop to examine that exhibit and the way it was developed; many were already making similar plans.

Patrons of the Art

The scientific process is at the center of the latest U.S. effort to reform the teaching of science from kindergarten through grade 12. A pivotal 1989 report published by the Educational Testing Service revealed that U.S. students lagged behind other students not in their knowledge of scientific facts but in applying thinking skills to solving problems. A month later, Science for All Americans, a report based on a three-year study by the American Association for the Advancement of Science, called for massive changes in the way science is taught to emphasize “exploration of questions” and critical thinking over “learning of answers.” Clearly, science museums have a critical role to play in complementing what students learn at school.

Thus the National Science Foundation, which has funded activities throughout the United States aimed at improving science education, has supported the growth of innovative museum exhibits: NSF’s Informal Science Education (ISE) grant program has become an important source of funding for such exhibits. The program not only enables specific institutions to create new offerings but also encourages developers to communicate with other educators. To ensure that others will learn about what succeeds and fails, NSF requires evaluations and dissemination of information about museums’ programs.

Recently cuts have been proposed for ISE, including a 30 percent cut for the federal fiscal year of 1997. Fortunately, recognition among members of Congress of the value of informal education prompted the Senate Appropriations Committee to restore funding during the 1997 budget review process.

In such a climate, however, science centers clearly cannot depend on just this source of funding to advance their new educational goals. The private sector also needs to broadly support innovative exhibit development. Sadly, economic reality has caused corporate philanthropic contributions from U.S. businesses to decline in recent years. Interested in attaching corporate names and sometimes logos to projects, companies’ marketing departments are partly replacing the older form of support. But these branches are less likely to invest in experimental and therefore risky projects.

With just a few museums having created activities that emphasize the process of science, developers have only begun scratching the surface on this powerful and potentially influential concept for effective and enjoyable “informal education.” To help develop such approaches, all those with a stake in science education need to step up and do their part.