Ask Us Anything: teaching the scientific method to middle schoolers

A while back we invited you to ask us anything. Here are our answers to Aaron Mietzner’s question: how would you teach the scientific method to middle school students?

Jeremy’s answer: I know nothing about teaching middle school, so I’m just going to point to my buddy Greg Crowther’s old post for us on teaching the scientific method to high school students and hope that it generalizes to middle school.

Brian’s answer: I guess I would say the question is a little bit leading. A more basic question is “what should we teach middle schoolers about how science works?” I personally disagree a bit with Greg in that: a) I don’t think scientists follow the classic 4 step scientific method, and b) I don’t think it is what makes science powerful. We don’t have a recipe. We have a toolkit and a mindset. What I would teach middle-schoolers about science are:

  1. We have ideas that can be tested by data in the real world. An idea that can never be tested is unscientific.
  2. Scientists challenge and test each others ideas. This usually causes us to be as rigorous as possible with our own ideas before we present them publicly. But when our ideas are proven wrong we accept it and move in new directions.
  3. Science is a contest of ideas. We compare and contrast different ideas, explanations and approaches to each other all the time. And the better ideas accumulate more evidence in support and eventually win. Any idea currently believed by scientists has stood up against tough competition.
  4. We quantify things. This makes it much more likely that we can prove something wrong than if we use vague verbal, qualitative models and test them verbally.
  5. We use experimental and observational methods to pin-point what is going on (i.e. we use controlled experiments and we include covariates like productivity or patient age in observational data).
  6. We learn from past scientists and build on their ideas so we don’t just make things up – we build our knowledge up carefully.
  7. We use formal methods from statistics to quantify the types of variation and measurement errors in our systems.

Personally I would much prefer middle schoolers learn that scientists do those things and practice how they can do these things themselves rather than learn some cook book recipe about how to do science.

And beyond that, data literacy. Let me repeat that: data literacy. I would much rather middle school science teachers produce students who know how to read a scatter plot with a regression line, a trend line over time, or paired bar charts with controls and treatments. I am continually blown away how poor the data literacy is in university freshman.

8 thoughts on “Ask Us Anything: teaching the scientific method to middle schoolers

  1. I think Brian’s list is a great one that cuts to the heart of what science is really about. I continue to question whether that list (or similar lists, like the 8 NGSS Science and Engineering Practices) can be really understood and appreciated by, say, middle-school students with very little prior sense of how science works. Any individual item on Brian’s list is fine, but (as I think he’d agree) it is the sum of all (or most) of those things done together that results in real progress — so how do we teach novices how these components interact? I claim that they must be bundled together in some sort of digestible form, e.g., a 4-step or 6-step process (perhaps with an arrow back to the 1st step, to show the iterative nature of science). The fallacy of science-as-4-step-process does not bother me as a lie that we tell our students because it provides a framework in which naive students can start to appreciate how observations and prediction and data analysis and experimentation fit together.

    • Thanks for the thoughts Greg. I agree with your basic point that the list is longish and some of them are sophisticated. I for sure don’t think it could be effectively communicated to K-5. Having just spent the spring doing climate change outreach with middle schoolers, I’m on the fence whether middle schoolers could get it (some would, some wouldn’t, so I guess it depends on goals there).

      I even get simplification becomes a lie and that is OK.

      But I think our main point of divergence is the 4 step scientific method doesn’t really capture many of the aspects on my list. In fact as I see it implemented in K-12 education it kind of implies that making up hypotheses out of thin air is how we do science. And a single time test is decisive about results. And that is about all they get out of the scientific method that I can see. I’d almost rather they get nothing out than get that out.

      Just my 2 cents.

      • I don’t think there’s anything about the (4,5,6) step method that claims or requires that a single test is decisive. IMO that’s an assumption that many people read into it, but its not there nor is it required. The (lets go with 5) 5 step method is scalable, meaning that it can apply to the process at any scale of resolution. Perhaps that’s an underappreciated aspect of it. Nor does it claim that every answer is yes or no, black and white.

        I like alot of your ideas but IMO a great deal of sculpting and reworking would be necessary to make them useful as a description if what science is. For example I don’t like the way you use the term “data” as though all data were tables of numbers. The numbers in tables are one kind of *observation*. Its *observations* that we use to validate hypotheses. Sometimes those observations take the form if numerical tables and / or charts. Other times they include a wide variety of different kinds of data. For example, there is no tabular dataset that distinguishes a convergent plate boundary from a transform one. The distinction between the two requires substantial knowledge of seismicity, landforms, structural geometry and even sedimentation history. LOTS of great science has been accomplished with relatively few numbers. And I bet there’s no set of numbers that alone distinguishes, for example, the lobo, the red wolf and the grey wolf. And we really don’t use formal statistical measurements to determine errors for this kind of science – but its no less valid for the lack if formality.

        And scientists can and do “make things up” – as in hypotheses – as well they should. The only requirement is that what they “make up” is consistent with much of the existing information. It doesn’t even have to be consistent with all of the existing data or even be internally consistent. Thats why we often have competing hypotheses – because none of the competitors explains all of the observations, right? Even true “theories” don’t explain every detail.

        So you know I see the sentiment behind where your going but there’s *alot* of chiseling still to be done there, and you might be underestimating the five step concept. There much more to it than what’s explicitly spelled out.

      • I understand the discomfort with a cookbook-style approach to how science gets done; I share this discomfort, even though I’ve been doing that very thing in my middle school science classes for 20 years. I continue to do it, though, for good reasons. Many students at this age still think very concretely, and they’re not ready for things to be too free-form. And it’s possible to introduce a “method” without being overly dogmatic about it, or doing it without referencing real practices. You can also connect the writing of hypotheses to prior knowledge or a priori reasoning; I can’t imagine too many middle school science teachers accepting hypotheses pulled from thin air.

  2. I’m not sure at what age the following conversation could be usefully begun, perhaps high school, but at the appropriate stage I’d focus on questions like the following…

    What is our relationship with science? Or put more generally, what is our relationship with knowledge? What has that relationship been historically, how has it changed over time, what is an appropriate relationship with knowledge today, and how might that change over the course of the student’s lifetime?

    The reasoning here is that few of the students will become working scientists, but they’ll all become hopefully thoughtful responsible citizens. They’ll need a basic understanding of how the scientific method works, agreed. But more importantly they’ll need the ability to stand back from the details and develop a big picture perspective that objectively examines the degree to which the scientific process is taking us where we wish to go as a culture.

    The students are the future funders of science, and in the times they will inhabit they will need a more sophisticated analysis than “we should fund as much science as possible”. They will need to make very intelligent and difficult decisions regarding which parts of pandora’s box to open, and which parts to resist. Such decision making will require a much broader and deeper understanding than just how the scientific method works.

    If they arrive unprepared for such sophisticated mature decision making, there’s a good chance they’ll crash civilization, which if true will render all of today’s science basically pointless.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.