Variables (Independent and Dependent Variable)

An experimental investigation is an organized process used to test an idea. In a good experimental investigation, one must eliminate error to ensure that the results are due to the factor or factors being tested.

Let’s look at an experiment with plants. Kaitlin is trying to determine how pH affects the germination rate (time it takes for the seeds to sprout). Click through the following animation to see how the experiment was set up. 

Was Kaitlin’s conclusion correct? Can we make any inferences based on her experiment?

Remember that an experimental investigation is a “fair test” designed to gather evidence to support or refute a causal relationship. Can we tell what caused the results of Kaitlin’s experiment? The answer is no. Kaitlin did too many things differently in the experiment. Each cup was a different size and there was a different amount of soil in each cup. She planted varying amounts of beans in the cups and each cup received a different amount of water.

A variable is something that changes. There are two types of variables in an experiment, the independent variable and the dependent variable.

Let’s compare the two types of variables. Click on each variable to learn more about each type of variable.

Let’s look back at Kaitlin’s experiment.

She covered the beans with soil and added water to the cups. The water had been treated with chemicals to alter the pH. She added the following amounts of the different pH solutions to the cups.

  • 3 ounce cup—3 mL of water solution with a pH of 3
  • 10 ounce cup—10 mL of water solution with a pH 7 (neutral)
  • 16 ounce cup—16 mL of water solution with a pH of 9

She changed too many variables. In an experimental investigation there should only be one independent variable tested at a time!

What should have been Kaitlin’s independent variable? (Remember she was trying to determine how pH affects the germination rate (time it takes for the seeds to sprout).

Constants

In section one, you learned that in an experimental investigation there should only be one independent variable. Kaitlin had too many independent variables in her experiment. What should she have done with all the other variables such as amount of water, size of cup, number of beans, and amount of soil? She should have made them constants. In other words, she should have kept all of these the same throughout the experiment.

Look at the table below. It shows what Kaitlin did in her experiment in the first column and what she should have done to create a “fair test” experiment in the second column.

Kaitlin's Experimental Design "Fair Test" Experimental Design
3 different size cups (3 ounce, 10 ounce, and 16 ounce) Same size cup (10 ounce)
Different amounts of soil. (She filled the different size cups to the top. Since they were different sizes they each held different amounts.) Same amount of soil (225 mL of soil)
Number of beans (1 bean, 8 beans, and 6 beans) Same number of beans (6 beans in each cup)
Amount of water (3 mL, 10 mL, and 16 mL) Same amount of water (16 mL)

A chart like the one below is a helpful tool when trying to organize the parts of an experimental investigation. See if you can complete the chart below to help Kaitlin redesign her experiment.

Groups (Experimental and Control) and Repeated Trials

 In section one, you learned that the independent variable should have different levels or degrees. In Kaitlin’s experiment the different levels of the independent variable were pH of 3, pH of 7, and pH of 9.

Usually one of the levels of the independent variable is the reference point or “normal” value of the variable, and the other levels will be compared with this one in order to draw conclusions from the experiment. This reference level of the independent variable, which other levels will be compared to, is called the control group for the experiment. In Kaitlin’s experiment the group of beans that was watered with the liquid that had a pH value of 7 would be the control group.

The other levels of the independent variable are referred to as the experimental group(s). Each experimental group should only have one factor different from each other, the level of the independent variable, and everything else must remain constant.

Another part of the experiment related to the independent variable is a number of repeated trials for each level of the independent variable. In general, the more times you repeat the experiment, the more valid your results are. The possibility of obtaining misleading or inaccurate results due to experimental errors will be less. Every experiment should be carried out several times and then the results of the individual trials averaged together. The number of repeated trials used will vary depending on the experiment you are conducting. However, as a general rule, the experiment should be repeated at least three times if time allows.

In Kaitlin’s experiment she did not have repeated trials. She planted more than one bean in each cup but she should have planted the beans in at least 9 cups and watered three cups with each level of pH solution. 

Experimental Design Practice

Let’s practice what you have learned about experimental design. Read the following lab scenario. Identify the independent variable, levels of the independent variable, dependent variable, constants, experimental, control groups, and number of repeated trials.

Eduardo placed 450 mL of sand, potting soil, and a mixture of sand and soil into 3 separate 10 oz. Styrofoam cups. In each of the cups he placed a thermometer so that the bulb was 2.5 cm below the surface. He placed the 3 containers under identical heat lamps for an hour. Each cup was 15 cm below the heat lamp. The original temperature of each container was 16°C. After heating the jars in three separate trials the temperatures of the containers were as follows:

Sand: 28°C, 27°C, 26°C

Potting soil: 33°C, 29°C, 31°C

Mixture: 29°C, 29°C, 22.5°C

Directions

Click on the correct location on the graphic organizer below to match the term. If you need a hint, click on the “Hint” link in the graphic organizer.