J. Josˇ Bonner

July 2007

The Scientific Method

--and why I recommend that we not use it in Elementary Science


As typically presented, the Scientific Method is a formalized logic for scientific analysis.  At its core, it is a procedure for testing hypotheses.  When used appropriately, it is a superb scientific approach.


There are several problems. 


1.  Most disciplines don't use the Scientific Method.  They use different formalized logic patterns.  There are some disciplines that do use the Scientific Method (such as ecology), where it is invaluable.  The particular types of logic and reasoning used in various disciplines depend critically on the types of issues being investigated.


2.  The terminology is commonly used incorrectly.  The formalized logic of the Scientific Method requires the terms hypothesis and prediction.


Hypothesis: this term refers to a researchers current understanding of the system she is investigating.  Essentially, "stating your hypothesis" is saying "I think this is what's going on here..." followed by writing down a complete statement of what you know, what you understand, and how you have interpreted the evidence you have thus far evaluated.


Prediction: this term refers to things that must be true if the hypothesis is correct.  For example:  Having watched dogs interact, we see that they understand one another, as if they are communicating.  We can explain this by the hypothesis that dogs communicate through body language.  That is, it's not all barks and growls.  What things would have to be true if this hypothesis is correct?  For one thing, there should be specific postures that are associated with specific behaviors.  There should be postures that one dog adopts, that are followed by certain behaviors by the other dog.  These are predictions that are useful in testing the hypothesis.  We may not know what dog postures there are, but we predict that such postures should exist.


When we do the investigation, we manipulate variables, or quietly observe, or whatever may be required to figure out whether the things we predicted actually occur.  In the example above, we might put two dogs together in a fenced yard, and watch what they do.  We might see particular postures repeated from time to time, followed by a particular type of interaction between the two dogs.  We might see several such postures.  But this would not prove the hypothesis.  Maybe these two dogs are used to playing with each other.  We'll have to observe other dogs, too.


We may begin to think that certain postures have certain meanings.  This would lead us to more hypotheses (one for each posture), which would necessitate new predictions, etc.


Mis-use of terms: Often, the terms hypothesis and prediction are used interchangeably; students are asked to write their hypothesis, and instead write predictions.  Often, we allow this, because we're a bit uncertain about the Scientific Method ourselves.  It is also common to ask students to predict what will happen in a particular experiment.  In the formalized logic of the Scientific Method, predictions are "the predictions made by the hypothesis," not what an individual person thinks might happen--especially if the individual has not previously developed an hypothesis!


3.  Elementary school students have little scientific background, and therefore are unlikely to be able to develop strong hypotheses about how the world works. 


4.  The Scientific Method is designed to prove hypotheses to be wrong.  This is its strength, but requires deep understanding of the value of eliminating inaccurate explanations.  For young children, it's no fun to be proven wrong all the time.  This is one of the factors that contributes to students not liking science.


What to do instead?


Most scientific disciplines ask questions.  For example, rather than saying "I hypothesize that dogs communicate through body language," scientists could say "how do dogs communicate?"  To answer the question, one could put a pair of dogs in a fenced yard, and watch what they do.  Experimentally, we'd get to the same point as we would using the Scientific Method.  But we don't have to prove ourselves wrong; we don't have to make guesses about what's going to happen; we don't have to know very much about the thing we want to investigate.  Later, as we learn more, and begin wrestling with observations to figure out what's going on in nature, we may develop our own understanding--our own hypotheses--based on the observations.  We may have two alternative explanations--alternative hypotheses--that we can't distinguish based on what we know.  Then we can graduate to the Scientific Method.  When we have an hypothesis we want to test, or even better, two hypotheses and we want to figure out which one is wrong, then the Scientific Method will be just the tool to use.