The Scientific Method (2)

Procedure

The procedure is the practical part of the scientific method, it's the steps you take to test your hypothesis.

The purpose of science is to discover things about the world with accuracy, truth, and objectivity. Scientists

An important part of the scientist's process of discovery is the procedure followed. A procedure is like a recipe—it's a list of steps. The steps you plan to take to test your hypothesis must be clearly written out so that anyone could repeat what you have done. Your procedure

Experiment and field work

Scientists test their hypotheses either through experiment or field work. Experiment

Experimental observations are made in a controlled environment that you create. How? You make a simpler, smaller-scale version of the part of the real world you want to study. You focus your attention on just a few things, instead of on everything that can happen.

In an experiment, a scientist tries to look at how just one thing affects a subject. The difficult part is creating an environment in which only that one thing changes. That is why you often see scientists using test tubes, petri dishes, and other small, enclosed settings for their experiment. It is much easier to control things in such environments.

Field work

In field work, a scientist goes into an uncontrolled environment, a specific place in the world, and records exactly what is observed there at the time. Because you are studying a unique situation every time, field work is almost always new and original.

The difficult part with field work is that while you are recording your observations, you must make sure that you yourself are not interfering with your subject simply by being there. For instance, you cannot count birds in a tree if you scare any away while you try to count them.

Variables, Controls, Groups, and Trials

Scientists are like detectives—they try to solve mysteries. Experiments are part of a scientist's detective kit. When you want to prove a theory true or false, create an experiment that will test one thing you can observe.

You have an idea—that if you set up a controlled situation and purposely change only one thing, this change will cause something else to happen. The thing you purposely change is called the changing variable. If your change causes something else to happen, this "something else" is called the responding variable, because it is responding to the change.

You must plan your procedure carefully to be sure that you change only one thing in your experimental group.

Suppose you want to know what would happen if you played music for an experimental group of plants. You will play music, your changing variable, and watch for any signs of a responding variable, which you expect to be bigger or faster growth.

But how will you know if any growth is a change? How will you know what is bigger and faster growth? You need a way to compare the rate of growth. You need to have something to compare your experimental plants to—something to show what normal growth is. So you need a control group. You need to raise some other plants in exactly the same way as you raise your experimental group, except that they will not experience the changing variable. You will treat them exactly the same as you do the experimental group, but you will not play music for them.

Your experimental group—Give these plants x amount of food, y amount of water, and play music for them.

Your control group—Use the same kind and age and size of plants, give them the same amounts of food and water, but do not expose them to any music.

You can measure the growth of the plants that you expose to music against the growth of the plants that you don't.

Still, you need to consider some other things. Can you think of anything that could affect the plants? How about diseases or pests? Could some of the plants have been healthier than others before you even started the experiment? That is possible, even though you looked them over carefully before you began. To ensure that any recorded change is from your changing variable, and only from your changing variable, you should test in groups of at least 25 subjects.

For example, if you tested only one or two plants and they both died, you could not be sure that their death resulted from your experiment, or if they were weak before being part of the experiment and were about to die anyway. But if you tested a group of 25 subjects and only two died, you could more confidently conclude that those two plants had been weak or ill before the experiment began.

To be reasonably sure that nothing happens randomly (by chance), you also should run at least three trials—do your experiment three times. For example, if you ran your experiment once and correctly used a group of at least 25 subjects, but they all died, you could not be sure their deaths were the result of your experiment. Perhaps they had all been weak or ill before they were affected by your experiment. If you have at least three trials, and the results are similar each time, you can feel more confident that your results are accurate. If one of the trials gives results that are inconsistent with the others, you can suspect a problem with the odd trial.