Scientific+Thinking

Just what //is// the vague idea we call Scientific Thinking? It's important, but like "inquiry," it has a fluid definition. After many years of teaching the subject to graduate students, I have concluded that it's really pretty simple. The core of Scientific Thinking is **Reasoning From Data**. Regardless of the field, whether you use the Scientific Method or shun it, whether you do controlled experiments or complex field studies, you //always// get data. Sometimes you call them "observations," sometimes they are quantitative and can be graphed, and sometimes they are photographic images; whatever form they take, they are the data. Your job is to figure out what they tell you.

You don't spend your days as a scientist memorizing what other people have already done. Nor do you spend it thinking entirely new thoughts and inventing entirely new methods. You //always// start with data, and make sense of it. At the cutting edge of science, you have available a lot of data from others and from your own work. You make sense of it, building an explanatory model. If you use the Scientific Method, you call this model "an hypothesis." If your field shuns the Scientific Method, you probably don't use that term, and call your explanation a "model." Whatever you call it, it is the current state of Scientific Knowledge in that field.

What you do next may be to test your understanding by formally stating what would //have to be true// if your hypothesis is correct (i.e. test your hypothesis), or it may be to //ask questions// about parts of the model that are less clear (making certain not to bias your thinking by predicting the results of an experiment). Either approach leads to the next investigation, and the next set of data.

Now what?

You reason from the data. What do the data tell you?

In many fields (mine, for instance), the doing of experiments can be excrutiatingly boring. You go to the 10-volume set of protocols, get the precise procedure for this experiment, and follow it to the letter. You may spend months following the exact same recipe, changing only the starting materials. At some point, though, you finally get the data. At last, you can actually do science. You can wrestle with the data, and figure out what they tell you.

Marcus Rhodes, one of the inventors of corn genetics, spent his summers in the field, tending his corn. In late summer, he'd go through a flurry of doing crosses. At the end of the season, he'd collect all the new ears and go back to the lab. He spent the entire winter figuring out what the data from the summer's crosses told him. Then, he'd design the next summer's work. The protocol, of course, was to trim the silks of the ear parent, and cover the ear with a bag; put another bag on the tassels of the pollen parent. Then, the next morning, shake the pollen into the bag, and pour some onto the newly-emerged, protected silks from the ear. Put the bag back on, and wait.

Work like this is the origin of the "cookbook lab." There is a procedure that works. There are many that don't. We don't have a lot of time for students to rediscover procedures, so why not let them follow the one that works?

Yes, there is tremendous importance to experimental design, and to the procedures that we use, and to data collection. At present, these are the things that are emphasized in classrooms. But reasoning from the data is not. More typically, we tell the students some stuff, on the theory that they need the background information before they can do anything. Then, we give them the lab. If they get the right answer, then they demonstrate that the teacher was right. Of course, they already knew that, because the teacher already told them the stuff. At best, students learn that "data" serve primarily to confirm the Facts that you already know. They have little opportunity to engage in knowledge-building.

There are several explanations for this.

1. Tradition. Schools used to follow the tradition of Miss Whoever (yeah, Miss. My aunt, who taught in Effingham, IL, had to keep her marriage secret, for fear of being fired) having students recite their lessons. Verbatim. Universities used to (and mostly still do) follow the tradition of telling students the stuff they need to know, and then expecting them to figure it out. "Active learning" is new.

2. Inquiry. From what I hear, Inquiry is more strongly embraced in elementary education than in secondary. Partly, this reflects the view (misguided, in my opinion) that to be Inquiry, students must do all of the investigation, from coming up with the questions to collecting the data. This insistence on Full Inquiry as the gold standard of "teaching science" makes it difficult to move beyond questions that are accessible to students with little background and no sophisticated equipment. Young children can make discoveries with flashlights and shadows; but it's hard to make discoveries that reveal the genetic code or plate tectonics. Often, therefore, "inquiry" is taught in a lab or two (e.g. "which paper towel is best?"), and "real science" is taught in the traditional manner.

3. Concrete Thinking. The long-standing tradition says that students are concrete thinkers. Young children, especially, aren't supposed to be able to handle abstract stuff like reasoning from data. Ha! Anyone have their own kids? Taking Science To School reviews recent data from many fields, and concludes that young children are just as adept at reasoning as adults. The pre-conceptions, or naive conceptions that they hold are often the same as those of adults. The reason that they came up short in Piaget's studies is that he asked them questions in adult terminology. Let the students show how they think, using different types of materials, and we find that they're pretty danged good at scientific thinking. In fact, I conclude that it's geneticaly ingrained in our species. If you can't evaluate the data from your surroundings, make sense of it, and do the appropriate things, you probably get eaten and don't pass on your genes. Much of the material available as Teaching Resources is biult on the old model of students-as-concrete-thinkers. I think we can re-evaluate their abilities, and develop new materials that are actually interesting for them.

4. The Nature of Science. To scientists, it's obvious why Scientific Knowledge is tentative. Such knowledge is, after all, entirely inferences from data. The data are the facts. The rest is inference. It's always possible that new data will be uncovered, and that will cause us to re-evaluate our explanatory models. But in much of the Education Literature, there is a philosophical viewpoint. This viewpoint is clearest in the Elementary Methods textbook, //Constructing Science in Elementary Classrooms//, which refers to students "constructing reality," rather than constructing //knowledge//. The clear message of the example used is that scientists get some data, and then guess at the answer. The earliest guesses are more likely to be wrong than are later guesses. If scientists guess at the answers, then there wouldn't really be much point in having students collect data and then make a bunch of guesses; better to go straight from data to The Answer.

We can solve this, if we have adequate Teaching Resources available -- by which I mean appropriate lesson plans. For example, rather than tell students about plate tectonics, let's give 'em a couple of years' worth of earthquake data, as latitudes and longitudes. Have them plot the locations on a world map. Gosh. Look at that. The danged things line up in interesting ways. What the heck kind of explanatory model can we build for that? Doing it this way, students would learn about plate tectonics, //and// engage in scientific thinking. I bet they wouldn't be as bored as usual, either.

Why does it matter?

The simplest answer is this: Our local economy has gone from industrial manufacturing to life sciences. The life sciences companies need employees -- many of whom do //not// need Bachelor's degrees. The employers report that the major deficiencies in new hires are in three areas, pretty much of equal weight: communication skills, simple math, and scientific thinking. They need people who can get data, process it, and know what to do about it, so that the company doesn't lose 2 million dollars' worth of cell culture in the giant bioreactor because someone didn't know what to do when one of the gauges showed the wrong reading. They're not going to ask their employees to recite the phases of meiosis. And right now, many of these companies are at a loss to find competent entry-level employees. One I've heard of has hired foreign nationals.

Scientific thinking matters. We've got to get it into the curriculum, from day one on.