Information about the natural world comes from many sources. In controlled experiments, the responses of similar groups to a treatment are compared, and differences in the responses suggest that the treatment may have had an effect. Where controlled experiments would be impossible or unethical studies that compare conditions in two or more similar situations that differ in place or time may be appropriate. Statistical analysis allows investigators to evaluate the probability that observed results are due to chance. Historical records, natural records such as fossils, oral traditions, traditional ecological knowledge, and observations from citizen scientists and parataxonomists are also important. Researchers often develop models to predict how a system behaves under specified conditions. This is useful when a system, such as the Earth’s climate, cannot be observed directly. Science provides a framework within which results can be compared to predictions and conclusions can be modified as new evidence becomes available. Arguments and information about the natural world should be evaluated critically for misleading statements and potential bias.