There’s a popular video on the internet in which Richard Feynman summarises how science works in just 63 seconds.
“I’m going to discuss how we look for a new law. In general we look for a new law by the following process; first we guess it.” The audience of students laughs. “No don’t laugh, that’s really true.” continues Feynman. “Then we compute the consequences of the guess, to see if it’s right, we see what it would imply. Then we compare those computation results to nature or we say compare to experiment or experience. Compare it directly with observation to see if it works.”
There’s a pregnant pause, while Feynman, the consummate story-telling teacher builds anticipation to his punchline, his key point, his lesson objective.
“If it disagrees with experiment, it’s wrong.” Feynman clearly states, pointing at the flow diagram he has drawn on the board.
“In that simple statement is the key to science. It doesn’t matter how beautiful your guess is, it doesn’t make a difference how smart you are, who made the guess or what his name is, if it disagrees with experiment, it’s wrong.”
Feynman gestures a crossing action with his arms as he says this, emphasising and repeating his key message – you can see how he was a masterful teacher as well as scientist.
“That’s all there is to it” he quips and his enraptured audience laugh again.
It’s a favourite video of mine, and a few Elvis Juices to the good I can be persuaded to make it the basis of a sincerely affectionate Feynman impression.
In just one minute Richard Feynman summarises the essence of science. It beautifully illustrates the liberal and egalitarian nature of science – that there is no scientific authority and anyone’s guess can be wrong. Anyone’s guess can be right. The only authority is the careful observation and application of common sense.
This thinking is taken to an extreme conclusion by progressive educators who believe that students can discover the laws of nature for themselves if only they are taught how to conduct experiments that are fair. They believe in the egalitarian scientific world all guesses are equal until they are compared to scientific observation. Therefore the guesses of novice student are as equally valid as a Nobel Prize winners’ guess and everyone is doing “science”.
But Feynman would disagree. However, you’d only know that if you dig around for the full version of the Messenger Lectures. In longer versions of the video Feynman goes on to discuss a hypothetical computer that makes a succession of random guesses and computes the outcomes. All possible guesses can be calculated. Guessing becomes random – a dumb man’s job. But Feynman explains, that’s not how science actually progresses, because existing knowledge means most of those guesses are impossible, or at least very unlikely. The best guesses are made by the well-informed. The knowledgable. The experts.
When I started my Ph.D. in 2003 the first thing I was told to do was read. Read everything on the subject that I was about to study. I was sent an enormous 54-page review, written by my PhD supervisor, to read weeks before I even started, and when I started in his lab I was given selected references from that paper to read in further detail and several more papers that had been published in the year or so that had passed between the publication of his review and my starting. Given that the purpose of a PhD is to push back the boundary of human knowledge (nicely illustrated in the gif below), it’s important to know where the boundary is.
I spent at least 3 weeks reading before I even touched any lab equipment – then two months learning how to use it all the equipment. Then the next 3 months performing control experiments. Finally around 6 months into my PhD I got to do my first real experiments (as planned out by my supervisor) which would yield new data. When I showed those first results to my supervisor I recall his response was “Oh good, that’s what I expected.” You see, he’d predicted the outcome of almost all the experiments that I conducted during that first year. That’s what happens when you’re an expert. The guesses aren’t random. They’re carefully considered from a position of expertise.
It was about a year and a half into my PhD before I started formulating my own hypotheses, and about two years before I finally conducted an experiment that hadn’t already been thought of, and correctly hypothesised by my supervisor – I was finally overtaking him as the expert in my miniscule niche of knowledge, I was on the path to passing my PhD.
A decade or so later when working as a long-in-the-tooth postdoc in a lab a colleague remarked
“the problem with modern PhD students is that they’d rather spend 3 weeks in the lab discovering something they think is new, than spend one hour in the library discovering that someone else already discovered it.”
I wonder now, with a good few years teaching under my belt, if science education in schools has not contributed to this problem – that students believe that they are best placed to discover new things for themselves, rather than read about them?
I’ve been considering for some time how to square this circle: that good science is done from a position of expertise, that practical work is a core part of school science, that school students need to be trained how to undertake good science, and yet they are, by definition, novices. The key, I have realised is in the Hypothesis stage of an experiment.
Question, Hypothesis, List of Variables, Equipment List, Method, Risk Assessment, Results Table, Graph, Conclusions. But the most important of these is the Hypothesis.
Students are in the strongest position to understand how science works when they can take their existing knowledge then make, and explain, a good hypotheses from a position of relative expertise within that domain.
Take the following example from our Year 7 scheme of work:
Question: What is the relationship between the force of friction and the mass of an object?
When forced to hypothesise the answer to this question students are faced with just three options:
- As mass increases friction stays the same
- As mass increases friction decreases
- As mass increases friction increases
Students can easily guess any of these and proceed with their experiment with little thought as to why their guess might be true. If indeed they are asked to make a hypothesis at all. Not only should students have to make a hypothesis they should also have to explain the reasoning for their hypothesis:
- because the roughness of the surface is always the same regardless of the mass
- because the roughness is smoothed out by increased mass pushing down
- because the roughness is more difficult to overcome when increased mass pushes down
Only when students can predict the outcome of the experiment from a knowledgeable position are they ready to carry out the experiment. Having to make an accurate hypothesis and explain why they have made that prediction using their knowledge makes students think, which is surely the name of the game? (Learning is the Residue of Thought as Dan Willingham says). In practice, school students can make their hypothesis freely, or use two-part MCQs such as above if scaffolding is required, but they should have to explain why they have made it.
Rather than concentrating on identifying variables and making tests fair – which seems to be the focus of much school science – should we put more emphasis on the hypothesis and ask students to spend more time on the hypothesis of their experiment? We will likely have to teach students concepts before practicals, but by asking students to make informed hypotheses before they start practical work they will have to carefully apply their existing scientific knowledge to a problem. This has the added advantage that their data analysis becomes much more meaningful and anomalous results or completely incorrect results (due to poor experimental design) become desirable cognitive dissonance that students have to explain, rather than accept as “correct”.
This is also much closer to the reality of How Science actually Works. Science is not the random guessing of the outcomes of experiments, but the careful guessing and comparission from positions of expertise. Which is why Feynman was one of the best.