Designing an Experiment

Related Resources
Great Science Fair Projects
Develop Your Hypothesis
Physics 101 - Basic Information

Elsewhere on the Web
Hundreds of Science Fair Projects for Students
Science Fair Central offers ideas for science fair projects and experiments

Your experiment needs to be designed so that it tests your hypothesis. The key point to designing an experiment is to control your variables.

Variables are anything that can affect the outcome of your experiment. There are three important classes of variables to consider:

• Independent variables are the few (or even better, one) quantities that you deliberately vary during your experiment, in order to asses their affects.

• Dependent variables are the ones that change in response to your deliberate variations of the independent variable. At least, your hypothesis says they will.

• The remaining variables are usually fixed at the same value from trial to trial, and are called controlled variables. Also variables that do change, but not in an important way between trials. For example, in many experiments the time of day, which changes as you conduct trial after trial, will not be important. However, bear in mind these variables that are incompletely controlled – if your hypothesis fails, these are often the first place you should look for an explanation.

You also need to bear in mind confounding variables. These are variables that change in an important way during your experiment, but you don't control, or at least account for. For example, if you were testing an hypothesis relating the amount of light in a room to the position of the dimmer switch, you would need to account for the differences in light coming in through the windows if some of your trials were conducted in the morning and some were late at night.

Be very careful if you do have more than one independent variable in your experiment. You need to know what variations are due to which changes. Only vary one independent variable at a time – this way you will be able to see the independent affects, then you can try varying them together to see if there are any interesting relationships.

You should draw a diagram of your experiment before you get started. Be sure to list everything you need, so you can have it on hand and just do the experiment. Nothing kills momentum like having to stop and find some more items you forgot you needed.