USDA Agricultural Research Service

What picture comes to your mind when you think of agriculture? Do you see some guy in overalls plowing the dirt with an old tractor, or women picking corn in a field? Maybe what you should picture is a scientist using a satellite-linked computer program to examine global climate changes, or a researcher developing a new product that can absorb 1,600 times its own weight in water.

Agriculture today is high-tech. It has to be because keeping U.S. agriculture productive requires research that is every bit as complex as the space program. As a matter of fact, many agricultural research scientists work with NASA on a regular basis. They use satellites to examine fields for crop damage, to map soil conditions, and to look for changes in the environment that could affect or be caused by farming.

While agriculture itself isn't one of the official science fair project categories, keeping agriculture working so we always have food to eat and renewable resources for everything from clothing to biofuel requires knowledge and research that fits into all of the categories.

So don't let an old picture of agriculture like overalls and dirt make you think that agriculture isn't a science.

When your teacher asks you to come up with a science project for class or as part of a science fair, think agriculture!

The idea behind a science project is to see what happens if . . . What happens to one thing if you change something else while you keep all of the other conditions the same?

That's the heart of all research, and a science project is just another name for research.

The idea behind a science project is to learn something new—through an experiment. You might guess the result beforehand, but you won't know for sure what will happen until you try out the experiment.

While your science project may be simpler than a scientist's, it should still follow the same basic steps that make up the scientific process.

Exactly what do you hope to figure out? What is the "what if " question? You should be able to write the research question in a simple sentence.

In fact, keep the whole project simple. This is important to the scientific process: the simpler the experiment, the easier it is to keep "all other conditions" the same and change only one thing.

That's how you can be sure that the thing you are changing is actually causing any difference you measure.

Return to Top

The Hypothesis ...

"Hypothesis" means "what do you expect to happen in your experiment?" Suppose your research question is, "what happens to seeds if I change the temperatures at which they are kept before they are planted?" The hypothesis might be "the higher the temperature at which seeds are kept, the quicker I expect them to sprout."

It's important to phrase your hypothesis correctly. For example, don't say "higher temperatures are better for seeds." "Better" cannot be measured. Decide on a hypothesis that can be proved in a measurable way. For example, "higher temperatures will make the seeds sprout faster."

It is perfectly fine for your experiment to disprove your hypothesis. If something unexpected happens during your experiment, the project doesn't need to be trashed. You just discovered something new and showed that what we expect is not always what we get.

Do some studying before you decide on your hypothesis. Sources of information include school and public libraries and the Internet. Also, once you have some background, you might consider writing, or e-mailing a scientist who works in the field you've chosen for your project.

Return to Top

The Procedure ...

The procedure is how you plan to do things: how you are going to conduct your experiment.

An experiment can only have one variable. That means you can only change one condition in each experiment.

For example, with the seed-sprouting experiment, if you vary the temperature at which the seeds are stored before you plant them, keep each group of seeds at that temperature for the same amount of time. And make sure all of the seeds get the same amount of light and water after you plant them.

If there's more than one variable, the experiment becomes flawed. It can be hard to figure out what other conditions must stay the same. But it may help to think it through before you start your experiment.

Also think about how long your experiment will take before you decide on your procedure. If you only have a few weeks to do your experiment, don't decide on a procedure that will take months to carry out.

Think about your "sample size." How many seeds will you test at each temperature? Allow a big enough sample so you can have a few duds in each group.

Once you decide on a procedure, write it down step by step. That way, you can prove what you did and can follow the same procedure if you need to repeat the experiment.

Return to Top

The Results ...

This is where you collect the information or data. Your data should be in numbers, not just what you see. For example, say that some of your plants grew 1 centimeter the third day. Don't say that the plants "look bigger today than they did yesterday." Words like "bigger" mean different things to different people, so reporting your results using only words can lead to confusion. You want to tell people exactly how much your plants grew.

Keep all your results in one notebook.

Return to Top

The Conclusion ...

It can be hard to understand the difference between results and conclusion, but the two are very different.

Results are the specific data collected during the experiment. The conclusion is what you learned from doing the experiment, and what the results mean. You might also think of the conclusion as a summary. In just a few sentences, you need to explain what happened in your experiment and whether it agreed with your hypothesis.

Did your data (the measurements you took) support your hypothesis? If not, that's a result, too. It doesn't mean that the experiment didn't work. Also, consider other possible explanations for your results. Did your treatment kill your plants or was it that you left them outside and some insects ate some of the leaves? You're not out to "prove" your hypothesis but to test it. Think more along the lines of "here's what I thought was going to happen and here's what actually happened." Then go on to explain why you think things happened the way they did.