Science is a process for asking questions about the natural world and testing the answers. Scientists use the Scientific Method as the orderly process to ask these questions and test their observations. This method begins with the curious scientist observing the natural world, and perhaps making measurements as well. From these observations a question is formed that is answerable using the tools of science. Next, a tentative or provisional answer to the question is formulated, which is called a hypothesis. The curious scientist then makes predictions based on the hypothesis, and tests these predictions through experiments or observations. If the predictions are not proven true, the hypothesis has been falsified. If the predictions are found to be accurate, the hypothesis is accepted as provisionally correct. Hypotheses and their tests must be repeated by others to be considered valid. Hypotheses that have been consistently validated through additional observations or experimentation can eventually be advanced to the status of theory. A theory is a thoroughly substantiated explanation of some aspect of the observable world. Examples of well-documented and widely accepted theories include geology's theory of plate tectonics and biology's theory of evolution.
Theories come as close to objective truth as possible. Like hypotheses, theories are still considered somewhat provisional because the possibility always exists that some day an observation will be made that is inconsistent with the predictions of the theory. Scientists must then construct a new hypothesis that is consistent with all available data. Science is a continually evolving process.
When testing a hypothesis, strict rules apply to what scientists may consider evidence, and how this evidence may be used. For example, correlation (when two things happen at the same time and in the same place or occur in the same way when subjected to the same stimuli) does not necessarily imply cause and effect (when one thing is a direct result of the other). Scientifically, the strongest evidence for proving a cause and effect relationship comes from controlled experiments. Evidence must be verifiable or observable by others, and if experiments are involved, they must yield repeatable results. Scientists are said to be using "critical thinking" when they apply these and other rules of evidence.
Scientific certainty of the kind that is evident in the Periodic Table of Elements is rare; therefore scientists need tools to cope with uncertainty as part of daily life. One tool is to intentionally suspend judgment when they have too little information. They note observations for which there is no present explanation, because they might later be useful in testing hypotheses. The curious scientist resists leaping from observations to cause and effect relationships without employing the Scientific Method to validate any hypotheses that might be posed. For example, a controversial issue that has received public attention is the unexplained strandings of beaked whales (For more information see Strandings). Scientists have been criticized because of their reluctance to give reasons for what happened in the face of uncertainty by those who desired immediate explanations for these strandings.
Scientific theories are often described and summarized using models. Models can be of various types, including mathematical models and computer models. For example, computer models are used to describe how the atmosphere works and to make predictions, such as of tomorrow's weather. Other examples of models are those that have been developed to describe how sound travels through the ocean. These are called acoustic propagation models.
If no data currently exist to answer a specific question, data from a similar situation may be used as evidence in a process called extrapolation. The strength of an extrapolation depends on how similar the two situations are. For example, when assessing the effects of noise on humans, data extrapolated from other mammals are considered more relevant than data extrapolated from reptiles, amphibians, fish or invertebrates because all mammalian ears have the same basic structure. Good extrapolations are fully acceptable in the Scientific Method, such as those used to derive Occupational Safety and Health Administration (OSHA) standards for noise effects on humans, which are largely based on extrapolations from cat and chinchilla data.
Scientists normally attempt to publish their results to contribute to the current body of knowledge in their discipline. "Peer reviewed" journals are the normal publication media. Peer review involves scientific papers being subjected to evaluation by highly qualified experts in the field. The reviews are usually anonymous. Rigorous use of the Scientific Method cultivates scientists that are well prepared to provide unbiased observations. When they review a paper written by another scientist they examine it for flaws that prevent the work from being considered valid. Reviewers (usually 2-3) will then accept or reject the paper or suggest modifications. For that reason, peer reviewed papers tend to be more credible than papers that have not been peer reviewed.
Appearing in print, even in a prestigious journal, does not guarantee that the results are correct, but only that reviewers could find nothing wrong with it at the time. The scientific community considers the published paper within the sum total of what else is known in the field, adopting or rejecting the paper on that basis. Sometimes a paper is originally accepted only to be later rejected as new evidence emerges.
Science is a self-correcting process. The peer review process is used to detect errors and problems. The validation of a hypothesis through repeated experiments or observations insures that errors will eventually be detected as well. The science community is commonly criticized for being too slow and conservative in accepting new results or ideas, rather than too quickly accepting them. Ultimately, it is the rules, skepticism, curiosity, and inbuilt conservatism of the scientific community, as it continually evaluates and self-corrects the theories that it holds, that is the best method for approaching objective truth about the natural world and physical reality.
(adapted from a tutorial by Roger Gentry, NOAA Fisheries)