I can design and conduct a valid experiment.
Did a math teacher ever tell you to check your work? If you’re adding numbers, that means repeating the calculation to see if you get the same answer the second time. If you get the same result twice, then the answer is probably correct. But if you get a different answer the second time, at least one of the results must be incorrect. Then you have to add the numbers a third time and hope that the third answer will agree with one of the other two.
Scientists also have to "check their work." By conducting the experiment multiple times and getting similar results, scientists are able to build confidence in the accuracy of their data. It means they are more likely to be able to accept or reject their hypothesis. In addition, conducting multiple trials allows scientists to take the average (mean) of their many data points. To calculate the average (mean), scientists calculate the sum of the measurements and dividing that sum by the total number of measurements. This helps them to better represent the differences amongst their data points.
Scientists also have to check one another's work. The results of an investigation are not likely to be well accepted unless the investigation is repeated—usually many times—and the same result is always obtained. Getting the same result when an experiment is repeated is called replication. If research results can be replicated, it means they are more likely to be correct. Repeated replication of investigations may turn a hypothesis into a theory. On the other hand, if results cannot be replicated they are likely to be incorrect.
Because replication of scientific experiments is necessary, it is of utmost importance that scientists document their procedures thoroughly and specifically. If one scientist is going to repeat the experiment of another, she must have the exact steps the original scientist took to complete his experiment. When writing a procedure for an experiment, a scientist should write the steps in a clear way that is easy for others to follow. The steps should be numbered and include an illustration of the set-up when it is difficult to interpret or visualize from the words. Scientists should also ask for others to read their procedure to determine if there are steps that need more clarity or detail added to ensure the experiment could be repeated exactly as intended.
Why Consistent Results are Important in Science: An Example
The following example shows why replication is important in science. In 1998, a British researcher published an article in a medical journal reporting that he had found a link between a common childhood vaccine and autism, a disorder affecting children that results in trouble communicating and forming relationships (see Figure below). According to the article, children in his study developed autism soon after receiving the vaccine. Following publication of the article, many parents refused to have their children vaccinated. Several widespread outbreaks of disease occurred as a result, and some children died of the diseases.
This child is receiving a vaccine.
Soon after the original study was published, other scientists tried to replicate the research. However, it could not be replicated. No other studies could find a link between the vaccine and autism. As a result, scientists became convinced that the original results were incorrect. Eventually, investigators determined that the original study was a fraud. They learned that its author had received a large amount of money to find evidence that the vaccine causes autism, so he faked his research results. If other scientists had not tried to replicate the research, the truth might never have come out. You can read more about this case at the URL below.
Teacher Note: The controversy around vaccination presents a great opportunity for a Philosophical Chairs activity. Students could research and present both sides of the debate over whether or not to vaccinate.
- replication: getting the same results when an investigation is repeated.
Scientists often repeat their own experiments to see if they can replicate the result. These repeated experiments are called trials. At the following URL, read about the repeated trials in an investigation involving crows. Be sure to watch the movies showing the crows during different trials. After reading the article and seeing the movies, answer the questions below.
- Why were repeated trials done in this experiment?
- Why might other researchers in other labs try to replicate the results?
- What is replication?
- Why is replication important in science?
- Scientists do not try to replicate every investigation, but some are repeated many times. For example, many researchers tried to replicate the vaccine-autism study described above. Why do you think so much effort was made to replicate this particular study? What was the outcome of the additional research?
- 7.1.A.b: Identify and describe the importance of the independent variable, dependent variables, control of constants, and multiple trials to the design of a valid experiment
- 7.1.A.c: Design and conduct a valid experiment
- 7.1.C.c: Determine the possible effects of errors in observations, measurements, and calculations on the formulation of explanations (conclusions)
- 7.1.C.d: Evaluate the reasonableness of an explanation (conclusion)
- 8.2.B.a: Describe the difficulty science innovators experience as they attempt to break through accepted ideas (hypotheses, laws, theories) of their time to reach conclusions that may lead to changes in those ideas and serve to advance scientific understanding (e.g., Darwin, Copernicus, Newton)
- 8.2.B.b: Describe explanations have changed over time as a result of new evidence
- 8.3.B.a: Describe ways in which science and society influence one another (e.g., scientific knowledge and the procedures used by scientists influence the way many individuals in society think about themselves, others, and the environment; societal challenges often inspire questions for scientific research; social priorities often influence research priorities through the availability of funding for research)