M-c Shanahan

On Tuesday night, just as I was settling in to read before falling asleep, I took one last look at Twitter to see if anything interesting had been posted. Overseas friends were up for the morning and, not feeling entirely sleepy yet, a nice meaty article or blog post was just the thing I was looking for. One headline from Scientific American caught my eye:

“More Than Child’s Play: Ability to Think Scientifically Declines as Kids Grow Up. Young children think like researchers but lose the feel for the scientific method as they age”

A statement like that would have serious implications for science education and for my teacher education students. It was a must read.

And a slightly frustrating one.

The headline link led to a short research summary by Sharon Begley (Note that the free excerpt actually includes the full piece so you’re not missing anything, even without a subscription). It describes a study published in the September 2011 issue of the journal Cognition by Claire Cook (MIT), Noah Goodman (Stanford), and Laura Schulz (MIT). The study is a clever investigation of preschoolers’ stronger than expected ability to find ways of understanding causation when they are presented with an ambiguous situation. The children were shown play blocks that, when pressed against a toy box, sometimes make the box light up. They were encouraged to play with the blocks and some children found unexpected ways to isolate them so they could test each one separately to see if it made the box light up. The first four paragraphs do a nice job of explaining the research (aside from a misuse of the word variable, but more on that in a second). The fourth paragraph  ends with the sentence “That suggests basic scientific principles help very young children learn about the world.” Cool. And nothing wrong with that statement. It captures the study and the researchers’ conclusions well.

It’s the final paragraph that both inspired the headline and my frustration (underlined for emphasis).

“The growing evidence that children think scientifically presents a conundrum: If even the youngest kids have an intuitive grasp of the scientific method, why does that understanding seem to vanish within a few years? Studies suggest that K–12 students struggle to set up a controlled study and cannot figure out what kind of evidence would support or refute a hypothesis. One reason for our failure to capitalize on this scientific intuition we display as toddlers may be that we are pretty good, as children and adults, at reasoning out puzzles that have something to do with real life but flounder when the puzzle is abstract, Goodman suggests—and it is abstract puzzles that educators tend to use when testing the ability to think scientifically. In addition, as we learn more about the world, our knowledge and beliefs trump our powers of scientific reasoning. The message for educators would seem to be to build on the intuition that children bring to science while doing a better job of making the connection between abstract concepts and real-world puzzles.”

Suggesting that scientific abilities vanish ignores the differences between two types of thinking: finding concrete causal factors (such as which block will make a toy work) and abstract scientific thinking (such as variable manipulation). At first I was thinking that it is like comparing apples and oranges but really it’s like comparing apples and something that on the surface seem kind of like apples but are vastly more complicated (a quantum apple?).

In the conclusion of the study the authors note that in schools and among researchers there has been a tendency to use overly abstract tests of scientific reasoning. These tests underestimate the intuitive skills that young children have for isolating concrete causal factors (which the authors unfortunately call “variables”). The researchers themselves were taken by surprise by one of the strategies that the children used and it helped them notice the novel solutions that the children found. That general conclusion makes sense.

What doesn’t make sense is extending that argument to say that students lose some sort of reasoning or scientific thinking ability as they get older because they struggle with abstract skills such as real variable manipulation. There is no evidence for that. Scientific thinking is abstract by definition because it is about underlying and generalizable knowledge. It is not the same thing as the concrete and situational problem solving reasoning that the children engaged in. It’s like comparing apples to the much more difficult and challenging quantum apples.

The authors of the original study do concede this in their conclusion, writing that “the ability to bring common principles of experimental design to bear on any task, regardless of the number of variables involved and the status of those variables with respect to their prior beliefs, requires an explicit awareness of the principles of experimental design that is, we presume, the exclusive purview of formal science” (p. 348). So while the foundations for this kind of thinking are found among children, there is a second level of complexity that moves this intuitive causal thinking towards becoming scientific thinking. Children do not have scientific thinking and then somehow lose it as adolescents.

Indeed, school children and teenagers continue to understand the basics of experimentation very well. There are several resources for teaching the concept of fair testing in science. They usually begin with intuitive ideas related to general fairness, like using the analogy of a race where everyone must start at the same place and take the same route. Even the idea of a fair test experiment, though, gives a very simplified introduction to scientific investigations. What is much more difficult is, for example, the idea of a variable. And here’s where I disagree not just with Sharon Begley but with the authors of the paper. By trying to isolate which blocks will make the toy work, the children are not isolating variables. There is only one variable – the blocks – and the children have found an innovative way to try to test one block at a time. A variable is an abstract place holder for a quality that is attributable or applicable to objects or systems of a particular type. Learning to use them fluently is hard, really hard. It takes explicit instruction and practice. Even simple variables like length are more challenging than they seem. It is one thing to measure the length of a particular piece of string, quite another to conceive of length as a general property that can be measured or manipulated in any object. This especially true because it is also somewhat arbitrary, requiring the person doing the experiment to choose an operational definition (e.g., by defining length as the measurement of the longest side). There is no concrete thing called length. It is an abstract word that describes a type of measurement. Understanding that is much harder than trying to find a way to measure it in specific objects, which is analogous to what the children are doing in trying to find a way to test each block individually.

This might seem like a subtle distinction but when it comes to taking steps to improve science education, it matters. The implication that students lose some ability to think scientifically because of school experiences or growing up is a misleading one. The headline especially reinforces it. The end of the paragraph gets closer to a real suggestion, which is that teachers need to better recognize the strength of young children’s reasoning and also recognize that learning the abstraction necessary for full scientific thinking is difficult. It requires better efforts to bridge concrete causal reasoning and abstract reasoning about variables and other scientific processes.

But just because that’s hard doesn’t mean that there is anything that students are losing. They just need more support to take the next steps.

Update (October 3): Shortly after I posted this, Matthew Francis added some great insight from his perspective as a scientist on his blog Galileo’s Pendulum.

Edit (September 22): I apologize for not including a link and reference for the original study in Cognition. A link has been added above and the reference information is as follows.
Cook C, Goodman ND, & Schulz LE (2011). Where science starts: spontaneous experiments in preschoolers’ exploratory play. Cognition, 120 (3), 341-9 PMID: 21561605

Further reading:

Bao, L. et al. (2009). Learning and scientific reasoning. Science, 323, 586-587.

Jones, M.G., Gardner, G., Taylor, A.R., Wiebe, E., & Forrester, J. (2011). Conceptualizing magnification and scale: The roles of spatial visualization and logical thinking. Research in Science Education, 41, 357-368.

Markovits, H., & Lortie-Forgues, H. (2011). Conditional reasoning with false premises facilitates the transition between familiar and abstract reasoning. Child Development, 82, 646-660.

Mercer, N., Dawes, L., Wegerif, R.,& Sams, C. (2004). Reasoning as a scientist: Ways of helping children to use language to learn science. British Educational Research Journal, 30, 359-377.

Watson, R., Goldsworthy, A., & Wood-Robinson, V. (2002). What is not fair with investigations? In S. Amos & R. Boohan (Eds.) Aspects of teaching secondary science: Perspectives on practice, London: Routledge Falmer (pp. 60-71).