Learn how scientists set up fair tests using comparison and variables.
When scientists have questions, they may perform tests or experimental inquiries to try to find the answers. To be confident in the experimental results, scientists need to set up their experiments so that they are fair. This type of an experimental inquiry is often called a controlled experiment or a Fair Test.
There are a number of factors that help to keep tests fair:
A Fair Test involves a comparison.
Fair tests typically involve comparing what happens when you change something or expose something to different conditions. Basically it looks at what the effects are of changing one variable. For example, if you wanted to find out how the incline of a ramp impacts the distance a toy car rolls, you would set up several different ramps (of the same length) at different inclines and let the same car roll down each and measure the distance travelled. The independent variable (cause) is the inclination of the ramp and the dependent variable (effect) is the distance the car travels.
A Fair Test is based on just one difference or change (the Independent variable).
In most experimental inquiries, we want to figure out if the change we make is actually causing the effect we see. In a fair test, it is important to only change one variable at a time. In that way, you are able to tell if the variable you have changed is causing what you observe. For example, if you were testing toy cars on ramps and you changed both the inclination of the ramp and the surface of the ramp, the test would no longer be fair. You could no longer be confident that the ramp incline alone was affecting the distance travelled. The new type of surface may cause the car to roll with less friction, and hence would affect the distance travelled.
A Fair Test tries to keep all other variables constant.
In order to ensure that only the independent variable is affecting the dependent variable (the outcome that will be measured), all the other factors (variables) acting upon the test situation must be kept constant (the controlled variables). It can sometimes be tough to control everything, but it is important to control as many of the factors that could affect the rest of the results as possible. For example, in the toy cars and ramps example, the car should be released from the same place on each ramp, the car should not be pushed down the ramp, etc.
How does Tomatosphere™ model a Fair Test?
The Seed Investigation is a useful way to model for students the design of a Fair Test. The inquiry provides a predetermined testable question (How does the space environment affect the germination of tomato seeds?), independent variable (type of seeds – treated and non-treated) and dependent variable (number of seeds that germinate).
What the Seed Investigation does not specify are the variables that students will need to keep constant (controlled variables).
Have students read the following scenarios and determine whether or not they would consider the tests to be fair.
A student wanted to test whether adding fertilizer to plants made the plants grow taller. The student added the same amount of fertilizer to three pots of plants and put the pots on the windowsill. The student did not put fertilizer into three other pots of plants and put them up on a shelf away from the window, because they ran out of space on the windowsill. Was this a fair test?
A class of students wanted to test the germination rate of two types of tomato seeds. The students planted 30 of one type of seed (“Big Beauties”) and 30 of another type of seed (“Small Wonders”) into the same type of soil and pots. All of the pots were placed along a windowsill where they received the same amount and duration of light. One student was in charge of watering the “Big Beauties” while another student was in charge of watering the “Small Wonders.” The student who was watering the “Small Wonders” went on vacation and didn’t water the plants for two weeks, while the other student continued to water the “Big Beauties” as usual. Was this a fair test?
A student wanted to test the effects of soil type on the germination rate of tomato seeds. The student planted 5 seeds (all the same type, evenly spaced out) in a pot filled with sand, three seeds (all the same type, evenly spaced out) in a pot filled with garden soil, and three seeds (all the same type, evenly spaced out) in a pot filled with compost. The student put the three pots side by side on a windowsill and poured 50 ml of tap water into the pots every three days. The student recorded when the seeds germinated in each of the pots. Was this a fair test?
No, it was not fair because not all of the plants had the same exposure to light which could affect the growth rate of the plants. The independent variable (cause) was the amount of fertilizer. Everything else should have been kept constant, but they were not. The amount of light exposure was different (on the windowsill vs on the shelf), and thus the students would not be able to determine with certainty that the growth rate was caused by the fertilizer alone.
No, it was not fair because not all of the plants received the same amount of water. The independent variable (cause) was the type of plant. Everything else should have been kept constant, but it was not. The amount of water was different, thus the students would not be able to determine with certainty that the germination rate was affected by plant type alone.
Yes. The only factor that varied in the inquiry was the type of soil (independent variable). All of the other conditions (seed type, seed placement, light, water) were kept constant.
Learn More
Variables for Beginners: Doing a Fair Test
This page, from Science Buddies, explains variables using two tests. One to see which toy car is faster, and the other to see if fertilizer makes a plant grow larger.
Fair tests: A do-it-yourself guide
This article from UC Berkley’s Understanding Science uses a delicious set of tests to find out which brand of chocolate chip makes the best tasting cookies.
Collecting data and fair testing
This PDF from New South Wales Department of Education and Training describes a student research project that includes secondary sources, first-hand information, and designing an investigation.
Controlled Experiments (2013)
This video (2:34 min.) from Simple Science by Mark Drollinger, shows how controlled experiments are done using an example with plants.
