When children are taught how to do science inquiry, the process usually begins with an inquiry question followed by having the student make a prediction about what is going to happen.  This approach often continues through high school lab science courses, even though in these inquiry situations students often do not have very much prior knowledge to base their prediction upon.

Having students make a prediction before doing the experiment certainly increases student engagement because they want to find out if they were “right,” but there are some drawbacks to this approach. One is that students who make a prediction that turns out to be “wrong” can feel a sense of failure. If a student repeatedly makes wrong predictions, she may develop a belief that she is not smart enough to do science or that she is not good at science.  Students often do not have any real background knowledge of the phenomena being investigated, and so their sense of success or failure hinges on how lucky they were at correctly guessing the outcome of the experiment.  A second drawback to having students make a prediction before carrying out an experiment is that it can skew their observations; what they expect to see can affect what they actually do see.

Investigative Science Learning Environments (ISLE) uses a different process to avoid these problems.  The ISLE process starts with an observational experiment in which students make observations or collect data without a prior expectation about what will happen. From these observations and data, students look for patterns and come up one or more possible explanations, or hypotheses, that answer the questions “why” or “how”. These explanations need to be experimentally testable. The next step in the process is designing an experiment to test each explanation.  Only at this point do the students make a prediction, which is based upon the explanation.  Because the prediction is based directly on the explanation, it is the explanation that will be found to be right or wrong, not the student. In this case,  the correctness of the prediction now reflects upon the correctness of the explanation, not the on the student personally.  The experiment becomes a judge of how well the explanation fits the real world, not of how smart the student is.  A prediction that turns out to be wrong can actually be exciting because it means that the there is something more to discover.

I really liked this approach, but thought there still was some value in having students give some initial thought to what would happen in an experiment, even when they don’t have very much to base it upon.

This month, at the Oregon AAPT meeting, Bradford Hill gave a presentation entited “Engaging Students through a Patterns approach to Physics.”  His approach is to call the initial prediction a “wild guess.”  The wild guess prediction is followed by inquiry to collect data, find a pattern, and then making a “Data-informed prediction” for a situation.  This draws the students’ attention to the fact that our confidence about a prediction can vary, and that the the quantity and quality of data used to determine the pattern impacts the confidence that can be placed in the prediction that is based on the data.   Calling the first prediction a “wild guess” frees the student from feeling that its correctness is a reflection of their intelligence or ability to do science.