Creating A Controlled Experiment

Creating a Good Controlled Variable Experiment
Good science experiments are more than demonstrations of cool things. A good experiment requires that you design a test to perform on something that you are interested in, to answer a question about science you want to discover. This exercise is meant to give you an idea on how to develop a good experiment.
STEP 1- Make a reasonable hypothesis
(What do you really think will happen during your experiment?)
After you have come up with a scientific question you would like to find the answer to, you will need to form a hypothesis, which is a prediction of what you think will happen based on observable facts. It is more than just an educated guess. You prediction should be based on what you expect to observe while conducting the experiment.
 Now we can get to the important stuff: Designing a good experiment!: Let’s take a look at those steps:
STEP 2 – Think About All the Variables
  (What might change in my experiment?)
A variable is a part of your experiment that can change. For example, if you were to do an experiment on plants, some of the things that you could change include the amount of water or light, and the kind of dirt or plant. You need to list all the things that can change in your experiment. Maybe you could let a parent, teacher, or friend check over your list.
STEP 3 – Set-Up a Controlled Experiment
  (How do I keep the variables from changing?)
In your experiment design you can only have one variable that changes. The variable that changes should be the one that you are testing. This is called the independent or controlled variable. For example, if you chose to do an experiment on light and plant growth, you would make sure that the variable that changes is the amount of light. If you have more than one variable that changes, that would make it an uncontrolled experiment. An uncontrolled experiment is an experiment that will not give you the answer to your question or give you the wrong answer. Now use the list of variables that you made and make sure none of them change except the one you are testing!
Let’s Do An Experiment Together!
Here is an example of a simple controlled experiment that you can do at home to help you get familiar with good design:
Which Brand of Battery Works the Best… Duracell, Energizer or Rayovac?
Question:  How would you set up an experiment to test this question?
• Make a hypothesis as to which battery will last the longest.
• Now let’s think about all the variables that could possibly be involved in this experiment.  Remember a variable is anything that could change before or during you experiment. A detailed list is included on the next page, but don’t look at it until you’ve made your own list!!!!!!
• Now that you have made up your list, think about which variable you want to change on purpose. This will become your controlled variable.
• If you have not already thought about it, your controlled variable is the one thing you want to test. This relates directly back to Test Question. So, the controlled variable is the type of battery used in this experiment.
• Now let’s look at a short list of other variables that you want to keep the same:
o The instrument that you put the battery types in (ex. Get some cheap flashlights )
o The brand of instruments (ex. Make sure you get the same brand of flashlights)
o The same size batteries & # of batteries required for each instrument
o Start time for each instrument should be the same.
Some Extras …
• You may want to try this experiment at home and be sure to record the start time and record the time that each battery dies out completely.
• You may use any instrument you like that uses batteries, but be sure that you have the same type/brand for each battery.  Also, to speed the process of battery usage time up, try to make your observations during continuous energy use. In other words, once you turn the instrument on, don’t cut it off until the batteries die.
After doing this simple experiment, think about and use the questions listed in the steps above as you design your own experiment. Remember that a well designed experiment attempts to keep ALL variables the same, EXCEPT the one you are testing!!!!!