Saturday, January 30, 2010

The Mars Rovers and Science Education

What does the Mars Rover program have to do with science education -- aside from studying the Mars Rover program?

It makes a useful analogy to science labs in classrooms around the world. That may seem a bit far fetched. As you read this analogy, don't assume it's crazy. Read to the end before passing judgment. You may be surprised at how apt the analogy is.

When NASA designed the Mars Rover program, it had a number of goals and restraints. Assume that it could consider just about any approach and then had to adapt to the goals and restraints, a brainstorming session. What were the range of options available?

At one extreme would be no trip to Mars. At the other extreme would be a manned trip to Mars. In between is the idea of a remote robotic explorer.

At one point during your brainstorming session, a software developer jumps up and proclaims that you can have a software program that includes all known information about Mars. This program can then simulate the data that a trip to Mars, manned or unmanned, might produce. The program not only could produce data but even could put together simulated images of the Martian surface. Just look at the benefis.
  • low cost (compared to a Martian trip)
  • complete safety (no astronauts at risk)
  • short time (writing software instead of building equipment and sending it to Mars)
At another point, a rugged test pilot stands up and says that the only way to explore Mars is in person. Simulations are for wusses and robots are for geeks. Lots of people like this idea, but it has some problems.
  • very high cost (compared to a robotic mission)
  • extreme danger (never been done before, may not be able to return, etc.)
  • very long time horizon (years of preparation, very lengthy trip)
When discussing the options, the simulation idea comes in for some criticism. The scientists tell the software developer that simulations won't generate any real science. They may look real, but they certainly will not match what the actual science will be on Mars. How can they publish papers on Mars using investigations of a simulation?

The scientists carefully explain that computer science is not science in the usual sense. It's actually an engineering discipline that produces tools used by scientists and by society.

In the end, of course, the robotic mission wins out as the least expensive real science option for exploring Mars. The scientists have a number of options regarding how to handle the data from the mission. It could be streamed live continually (sort of), or it could be stored on the rovers and sent later. The received data could be stored in a database and available for retrieval at any time in the future, sort of prerecorded for use by many different people at many different times.

While bringing NASA into this discussion does exaggerate the situation, it also shines a very bright light on how best to teach science, especially the use of science labs. In today's discussions of science labs in science courses, you'll find two extremes: those who insist on 100% hands-on labs and those who, with equal vehemence, insist on using simulations instead.

Fortunately, some are finding middle ground. At MIT, they're working on the iLabs project, which allows real-time remote robotic experimentation. Unfortunately, these labs are mostly engineering labs, and the likelihood of covering a reasonable range of science labs with this technology is very remote at this time.

The fact that all Mars Rover data are stored and usable by many scientists in many locations opens up a different approach: prerecorded real experiments. Images, videos, data, and other information can be stored for retrieval by students. The science certainly is as real as hands-on and remote robotics approaches.

The pedagogy depends on the software and the instructors. People who write the software and create the experiment videos cannot also create the instructors. They can only provide software that's easy to use and instructions for correct usage. Better science teachers know how to incorporate science lab experiences into their classes.

Data collection forms a very important aspect of the science lab experience. Data should not be precollected or automatically collected. Just as in a science lab, students should take their own individual data point by point. Each point represents not just the experiment but also student care and judgment, an important factor in understanding the nature of empirical data.

Each video should tell a story and provide means for collecting experimental data. If the video itself doesn't tell enough of the story, then the lab units should be supplemented with text, diagrams, animations, and videos that complete the story: tell the students enough so that they truly understand the details of the experiment.

Finally, sufficient supporting materials should be provided so that both students and teachers are able to succeed. This approach and list form the basis for Smart Science® education, a system of more than 150 lab units for use in science courses from grades 6 through college.

© 2010 by Paracomp, Inc., U.S.A.
Follow this author on ETC Journal

Hands-On Labs are Not the Answer

From the beginning of science labs in education in the mid-1800s, they've been hands-on labs. Until the latter half of the 20th century, the only other sort of lab was the paper-and-pencil lab. Some of you may not have encountered these desktop labs.

In a paper-and-pencil lab, the instructor hands out copies of laboratory "data," which may have been created from equations and not taken from real experiments. Students then fill in provided tables with the data and calculations based on that data. Finally, they answer questions about the results.

What are the benefits of paper-and-pencil labs?
  • low cost
  • minimal time required
  • high safety
  • lab technique does not affect results
What are the problems with paper-and-pencil labs?
  • no experimental design
  • likely to have unreal data
  • no kinesthetic experience
  • no visual experience
  • data not dependent on student technique
  • data not dependent of student judgment
I'm sure that you can add to these lists. You'll note that these features, except for lack of visual experience, match those of computer laboratory simulations being hawked by a wide variety of vendors, instructors, and amateur scientists. With simulations, the visual experience is generally poor, being limited to cartoon-like animations.

With so many defects in these labs, whether pencil-and-paper or simulations, you can see why so many educators have pushed back very hard to the point where they insist that only hands-on labs can be appropriate for science education. It's a natural reaction by those appalled by the large infusion of simulations into the laboratory part of many science curricula.

However, these hands-on purists are throwing out the baby with the bath water. By denying any lab but a hands-on lab, they're making advances in science education difficult and limiting their student experiences severely.

They should be searching for means to make new advances in technology available in science education. The goals must include the following.
  • lower cost of true science lab learning experiences
  • improve safety of science lab experiences
  • expand range of science lab experiences available to students
  • use student class and homework time more efficiently
  • provide exposure to the nature of science and all that it implies
Hands-on labs can be great learning experiences. Those that extend over many periods and involve iterative redesign and exploration can open up new vistas in students' imaginations. Instructors should not give these up entirely. However, recognize that such experiences are time-consuming and expensive. Usually, they require that students work in groups, and some in any group may opt out of the experience, just tagging along for the ride.

On the other hand, many hands-on labs are merely exercises in lab technique. How many students will find pipetting techniques valuable in the future? Other hands-on labs have been structured as "verification" labs, a class of labs that was railed against by F. W. Westaway nearly a century ago and by Carl Sagan much more recently. Students know all of the science and the numerical result expected before entering the lab. They are simply to verify this information.

Technique and verification labs do not teach science. They are a waste of time and money. Worse, they give students the impression that science is dull and uninspiring.

What's to be done? One way to view the options is to look at the Mars Rover program. It's real science, and not science pedagogy. So, you must be careful about drawing too close of an analogy. I'll be posting more on this analogy soon.

© 2010 by Paracomp, Inc., U.S.A.
Follow this author on ETC Journal