Scientific studies are often made via controlled experiments, which are essential to confirming the validity of the experimental results. In this lesson, you’ll learn more about controlled experiments.
Controlled Experiment Defined
Imagine conducting an experiment to evaluate the effectiveness of sleeping pills. You gather a group of test subjects and measure the average time it takes them to fall asleep after taking the pills.
Afterwards, you measure the average time it takes these same test subjects to fall asleep without taking the pills. By comparing the results, can you make a valid scientific claim about effectiveness of the sleeping medication?The answer is no, you can’t. Why? Well, let’s examine some arguments against the validity of such an experiment.Each test subject has different life circumstances that could affect his or her tendency to fall asleep. For example, before taking the sleeping pills, a significant percentage of the test subjects might have been very tired from working extra hours at their jobs, making it easier for them to fall asleep. Or, for the second part of the experiment where they didn’t take the pills, these test subjects might have been back to working regular hours.
There also might have been a psychological effect on subjects’ sleeping patterns because they believed the sleeping pills would help them fall asleep.By conducting the experiment as previously described, we can’t tell whether these and/or other factors influenced the results. To fix this issue, we need to make a controlled experiment in which the test subjects are divided among two or more groups, where at least one of these groups is a control group while the remaining are experimental groups.
The control group is composed of test subjects who remain in their normal state for the duration of the experiment. For each experimental group, one variable is changed from the normal state. Let’s take a closer look at what this means.
A Closer Look
In our example, the test subjects in the control group would continue living their normal lives, taking placebos (or ‘fake’ pills) to offset any type of psychological effect. That is, they would think they were on sleeping medication just like the test subjects in the experimental group, when in reality, they wouldn’t be affected by the pills.The remaining test subjects would be placed into one or more experimental groups. If we were interested in the effectiveness of one particular type of sleeping pill while all else remained constant, we’d only need one experimental group.
But if there were other variables that we wanted to test, such as the efficacy of other sleeping medications, we would need more experimental groups. Now, let’s look at setting up a controlled experiment.
Setting Up a Controlled Experiment
There are some important considerations in designing a controlled experiment. First, you need to determine what exactly is being tested and use this knowledge to create your experimental groups. In addition, you need to account for all the factors that could undesirably affect the results of your experiment.
For example, if you’re testing the boiling point of water with and without salt being added to it, the elevation and ambient temperature at which the experiment is carried out can have an effect on the boiling temperature. Also, the water being tested should not have significant amounts of impurities in it.
A controlled experiment involves having a control group and one or more experimental groups. In the control group, the test subjects are kept in their normal state, while one variable under study is changed in each experimental group.
Before setting up your controlled experiment, make sure to understand all the factors that could influence your results.