Scientists use an orderly approach to learn new information and to solve problems. The methods scientists may use include observing to form a hypothesis, designing an experiment to test a hypothesis, separating and controlling variables, and interpreting data.

Forming Operational Definitions
Operational definitions define an object by showing how it functions, works, or behaves. Such definitions are written in terms of how an object works or how it can be used; that is, what is its job or purpose?

Example
Some operational definitions explain how an object can be used.

• A ruler is a tool that measures the size of an object.
• An automobile can move things from one place to another.

• A ruler contains a series of marks that can be used as a standard when measuring.
• An automobile is a vehicle that can move from place to place.

Forming a Hypothesis
Observations  You observe all the time. Scientists try to observe as much as possible about the things and events they study so they know that what they say about their observations is reliable.

Some observations describe something using only words. These observations are called qualitative observations. Other observations describe how much of something there is. These are quantitative observations and use numbers, as well as words, in the description. Tools or equipment are used to measure the characteristic being described.

Example
If you were making qualitative observations of the dog in Figure 16, you might use words such as furry, red-gold, and long-haired. Quantitative observations of this dog might include a mass of 34 kg, a height of 90 cm, ear length of 10 cm, and an age of 1277 days.

Hypotheses  Hypotheses are tested to help explain observations that have been made. They are often stated as if and then statements.

Examples
Suppose you want to make a perfect score on a spelling test. Begin by thinking of several ways to accomplish this. Base these possibilities on past observations. If you put each of these possibilities into sentence form, using the words if and then, you can form a hypothesis. All of the following are hypotheses you might consider to explain how you could score 100 percent on your test:

If the test is easy, then I will get a perfect score.
If I am intelligent, then I will get a perfect score.
If I study hard, then I will get a perfect score.

Designing an Experiment to Test a Hypothesis
In order to test a hypothesis, it's best to write out a procedure. A procedure is the plan that you follow in your experiment. A procedure tells you what materials to use and how to use them. After following the procedure, data are generated. From this generated data, you can then draw a conclusion and make a statement about your results.

If the conclusion you draw from the data supports your hypothesis, then you can say that your hypothesis is reliable. Reliable means that you can trust your conclusion. If it did not support your hypothesis, then you would have to make new observations and state a new hypothesis-just make sure that it is one that you can test.

Example
Super premium gasoline costs more than regular gasoline. Does super premium gasoline increase the efficiency or fuel mileage of your family car? Let's figure out how to conduct an experiment to test the hypothesis, "if premium gas is more efficient, then it should increase the fuel mileage of our family car." Then a procedure similar to Figure 17 must be written to generate data presented in Figure 18.

These data show that premium gasoline is less efficient than regular gasoline. It took more gasoline to travel one mile (0.064) using premium gasoline than it does to travel one mile using regular gasoline (0.059). This conclusion does not support the original hypothesis made.  

Separating and Controlling Variables
In any experiment, it is important to keep everything the same except for the item you are testing. The one factor that you change is called the independent variable. The factor that changes as a result of the independent variable is called the dependent variable. Always make sure that there is only one independent variable. If you allow more than one, you will not know what causes the changes you observe in the independent variable. Many experiments have controls-a treatment or an experiment that you can compare with the results of your test groups.

Example
In the experiment with the gasoline, you made everything the same except the type of gasoline being used. The driver, the type of automobile, and the weather conditions should remain the same throughout. The gasoline should also be purchased from the same service station. By doing so, you made sure that at the end of the experiment, any differences were the result of the type of fuel being used-regular or premium. The type of gasoline was the independent factor and the gas mileage achieved was the dependent factor. The use of regular gasoline was the control.

Interpreting Data
The word interpret means "to explain the meaning of something." Look at the problem originally being explored in the gasoline experiment and find out what the data show. Identify the control group and the test group so you can see whether or not the variable has had an effect. Then you need to check differences between the control and test groups.

These differences may be qualitative or quantitative. A qualitative difference would be a difference that you could observe and describe, while a quantitative difference would be a difference you can measure using numbers. If there are differences, the variable being tested may have had an effect. If there is no difference between the control and the test groups, the variable being tested apparently has had no effect.

Example
Perhaps you are looking at a table from an experiment designed to test the hypothesis: If premium gas is more efficient, then it should increase the fuel mileage of our family car. Look back at Figure 19 showing the results of this experiment. In this example, the use of regular gasoline in the family car was the control, while the car being fueled by premium gasoline was the test group.

Data showed a quantitative difference in efficiency for gasoline consumption. It took 0.059 gallons of regular gasoline to travel one mile, while it took 0.064 gallons of the premium gasoline to travel the same distance. The regular gasoline was more efficient; it increased the fuel mileage of the family car.

What are data?  In the experiment described on these pages, measurements were taken so that at the end of the experiment, you had something concrete to interpret. You had numbers to work with. Not every experiment that you do will give you data in the form of numbers. Sometimes, data will be in the form of a description. At the end of a chemistry experiment, you might have noted that one solution turned yellow when treated with a particular chemical, and another remained clear, like water, when treated with the same chemical. Data, therefore, are stated in different forms for different types of scientific experiments.

Are all experiments alike?  Keep in mind as you perform experiments in science that not every experiment makes use of all of the parts that have been described on these pages. For some, it may be difficult to design an experiment that will always have a control. Other experiments are complex enough that it may be hard to have only one dependent variable. Real scientists encounter many variations in the methods that they use when they perform experiments. The skills in this handbook are here for you to use and practice. In real situations, their uses will vary.