Hypothesis

A hypothesis is a suggested explanation for something observed. For a hypothesis to be scientific, it must stem from observations and offer a prediction about reality that can be tested and repeated. This process often starts with an educated guess. When a hypothesis is repeatedly and independently confirmed through experiments, it can mature into a scientific theory. In everyday language, people frequently use "hypothesis" and "theory" as if they mean the same thing, but in science this is inaccurate.

A working hypothesis is a provisional idea accepted temporarily to push research forward. Such hypotheses are often abandoned, and researchers typically know and warn that they are incomplete or even false. Their purpose is to steer investigation in a roughly productive direction, especially when scientists are stuck and brainstorming. In formal logic, a hypothesis is the "if" part of an "if-then" statement. For instance, in the proposition "If P, then Q," P is the hypothesis (or antecedent) of the consequent Q. This P acts as an assumption in a "what if" question, which may be counterfactual. The adjective "hypothetical" describes anything that has the nature of a hypothesis or is assumed to exist as a direct result of one.

Historically, the word "hypothesis" once referred to a summary of a classical drama's plot. It comes from the ancient Greek *ὑπόθεσις* (hypothesis), which literally means "putting or placing under," and later took on many other meanings, including "supposition." In Plato's *Meno*, Socrates uses a method he says mathematicians employ: "investigating from a hypothesis." Here, "hypothesis" means a clever idea or a mathematical shortcut that simplifies complex calculations. Cardinal Robert Bellarmine famously used this sense when warning Galileo in the early 1600s that he must treat the Earth's motion not as reality, but only as a hypothesis.

In common 21st-century usage, a hypothesis is a provisional idea whose merit needs evaluation. To evaluate it properly, the person framing it must define specifics in operational terms. A hypothesis demands further work from the researcher to confirm or disprove it. Over time, a confirmed hypothesis may become part of a theory, or occasionally grow into a theory itself. Scientific hypotheses usually take the form of a mathematical model. Sometimes, but not always, they can be expressed as existential statements (claiming some instance of a phenomenon has a certain characteristic) or as causal explanations in the form of universal statements (claiming every instance has that characteristic). In an entrepreneurial setting, a hypothesis is used to form provisional ideas about product attributes or business models. These are then tested through experiments focused on verifiability or falsifiability.

Any useful hypothesis allows predictions through reasoning, including deductive reasoning. It might predict the outcome of a lab experiment or a natural observation. Predictions can also involve statistics and talk only in probabilities. Karl Popper, following others, argued that a hypothesis must be falsifiable—that a proposition or theory cannot be considered scientific if it does not allow the possibility of being shown false. Other philosophers of science have rejected or supplemented this criterion with others, like verifiability (verificationism) or coherence (confirmation holism). The scientific method uses experimentation to test a hypothesis's ability to answer a question. Unfettered observation is less likely to raise unexplained issues than a crucial experiment designed to test a hypothesis. A thought experiment might also be used.

When framing a hypothesis, the investigator must not already know the outcome of the test, or must ensure the outcome remains under active investigation. Only then can the experiment potentially increase the probability of showing the hypothesis is true. If the researcher already knows the outcome, it counts as a "consequence" that should have been considered during formulation. If predictions cannot be assessed by observation or experience, the hypothesis must be tested by others providing observations. For example, a new technology or theory might make necessary experiments feasible.

A trial solution to a problem is often called a hypothesis—or an "educated guess"—because it suggests an outcome based on evidence. However, some scientists reject the term "educated guess" as incorrect. Experimenters may test and reject several hypotheses before solving a problem. According to Schick and Vaughn, researchers weighing alternative hypotheses may consider: testability (or falsifiability), parsimony (Occam's razor, discouraging too many entities), scope (how widely the hypothesis applies to known phenomena), fruitfulness (the chance it will explain future phenomena), and conservatism (how well it fits existing knowledge systems).