Investigate It: A Grade 3 Fair Test
Students ask a testable question, make a prediction, conduct a fair test, record observations, and use evidence to explain their results.
Illustrations are auto-generated and may be placeholders. They can be refreshed to match the narration.
Ask a Testable Question
A testable question can be answered by collecting observations or measurements. It names what you will change and what you will measure. For example, you might ask, “How does the height of a ramp affect the distance a toy car travels?” You can change the ramp height and measure the car’s travel distance in centimeters. A question such as “Which toy car is best?” is not testable until “best” is clearly defined. Before testing, gather helpful information from reliable sources, such as a science book, a trusted educational website, or an adult expert. These sources can explain how ramps and motion work and help you plan safely. Record the source title and one useful fact from each source. Then check that your question can be investigated with the time, tools, and materials you have.
Make a Prediction
A prediction states what you think will happen and gives a reason based on what you already know or learned from a source. It is not a random guess. For the ramp investigation, you might predict, “If the ramp is higher, then the toy car will travel farther because it will speed up more as it rolls down.” Write your prediction before you begin testing. A useful prediction names the changed condition, the expected result, and the scientific reason. You can use an if-then-because sentence to include all three parts. Your prediction does not have to be correct. Scientists learn from results that support a prediction and from results that do not support it. After the investigation, you will compare the measurements with your prediction rather than changing the prediction to match the data.
Plan a Fair Test
A fair test changes only one variable at a time. In this investigation, the changed variable is ramp height. Test heights of 10, 20, and 30 centimeters. The measured variable is the distance the car travels after leaving the ramp. Keep the same toy car, ramp board, starting line, floor surface, and measuring tool for every trial. Release the car without pushing it. These controlled variables help make sure ramp height is the main reason for any difference in distance. Test each height three times because repeated trials make the evidence more dependable. Place a meterstick from the end of the ramp and measure to the front of the stopped car. Plan the steps in order, gather materials, and check the testing area for safety before beginning.
Observe and Record Data
During each trial, watch carefully and record what happens right away. Measurements are quantitative observations because they use numbers. Descriptions, such as “the car rolled straight,” are qualitative observations because they use words. Suppose the car travels 55, 60, and 65 centimeters from the 10-centimeter ramp; 85, 90, and 95 centimeters from the 20-centimeter ramp; and 115, 120, and 125 centimeters from the 30-centimeter ramp. Record every trial in a data table. For this example, the average distances are 60, 90, and 120 centimeters. Display the averages on a scaled bar graph. Label the horizontal axis Ramp Height and the vertical axis Average Travel Distance. Use equal intervals of 20 centimeters from 0 to 120. A scale with equal intervals lets readers compare the bars accurately.
Explain Results with Evidence
A results explanation answers the investigation question by using evidence from the data. First, state the pattern you found. Then include measurements that support the pattern. Finally, explain whether the evidence supports your prediction. For example: “The toy car traveled farther as the ramp became higher. Its average distance was 60 centimeters from the 10-centimeter ramp, 90 centimeters from the 20-centimeter ramp, and 120 centimeters from the 30-centimeter ramp. These results support my prediction that a higher ramp would make the car travel farther.” Do not claim more than the test showed. The investigation tested one car on one surface, so it does not prove that every car will behave exactly the same way. You can also describe possible errors, such as measuring from different points, and suggest a new test using another surface.