What does it mean to make a rational decision in a complex and dynamic environment?
A rational decision is one that is not just reasoned, but is also optimal for achieving a goal or solving a problem. In everyday life, we must deal with probabilities, varying degrees of certainty in a complex world. Rational decisions therefore necessarily involve some form of probabilistic reasoning. This may take the form of an inductive inference. Rational decisions are based on inferences constructed from the available information relevant to a given task. The premises of an inductive inference lend some degree of support for the conclusion. In other words the truth of the premises will indicate with some degree of strength that the conclusion is true. When the truth value of the available evidence is not certain, the probability that an inference is sound will vary accordingly.
Aristotle, the founder of formal logic, held that inductive logic was more appropriate for practical affairs than deductive logic. A deductive inference is 'valid' if the conclusion follows from the premises, and the premises must be true for the inference to be 'sound.' There is no element of probability. But in everyday life, we are not always presented with sufficient information to make a deductive inference. If I am on Earth, the constant of acceleration due to Earth's gravity is 9.8 meters per second per second; I am on Earth; therefore, if I free fall, the force that Earth's gravity exerts on me will be related to that constant regardless of other factors. If I am investigating a crime, however, I must rely on clues from which I may inductively infer the probability that x happened as opposed to y. (Note: the principle of Occam's razor is invaluable when thinking about probable explanations.).
In the process of making a rational decision there are parameters. This involves calculating probabilities based upon empirical evidence, i.e., making inductively valid inferences, and taking the action that maximizes the probability of achieving a desired outcome. It follows that a decision may be deemed irrational if it is predicated on a logically invalid inference, i.e., the conclusion does not logically follow from the premises, and does not maximize the probability of achieving the desired result. The degree to which a given action may be viewed as irrational will depend upon how counterproductive the action is to the objective.
A rational decision is by definition deliberate and based upon available data; however, many situations arise in the everyday world where one is required to make quick decisions for the sake of economy. One may make a rational decision with a limited set of data if the data strongly suggest that the available alternatives are counterproductive. In such a case, further inquiry is not always necessarily warranted. Ongoing inquiry and updating of hypotheses to reflect new evidence is however vital in scientific research.
Note: the term, "Irrational," should be distinguished from "arational." Arational behavior does not involve reasoning; neither rational nor irrational apply. Autonomic processes are neither irrational nor rational. Behavior over which one has (at least by default) no conscious veto power whatsoever, instinctively blinking when an object comes too close to one's eyes, for example, ducking when one hears a loud explosion, or other stress responses to perceived danger are arational.