Prediction means to forecast or foretell some future event based either on an educated guess or informed knowledge, the latter being akin to prognosis. According to the deductive-nomological model, explanation and prediction have the same logical structure. In practice, however, these two scientific activities are by no means synonymous. An obvious difference is that what is predicted has yet to happen, while what is explained has already happened. Thus, whether a hypothesis is to be considered an explanation or a prediction depends on whether its use occurs at the time earlier than the event described by the explanandum (prediction), or subsequent to the time of the explanandum (explanation). Less well appreciated is that a scientific explanation attempts to establish whether its hypotheses are more credible than their negations, and a scientific prediction that its hypotheses are more credible than their comparable alternatives. Furthermore, prediction does not always involve explanation and some explanations have little or no predictive power. For example, Thales of Miletus (634-546 BP) allegedly predicted the eclipse of 585 BP by consulting records, but he was incapable of explaining it. And the theory of natural selection can provide a plausible explanation for the variety of species, but it cannot seemingly predict new ones. Another misconception about the distinction between explanation and prediction is that the aim of any science is the former rather than the latter, that prediction is somehow less scientific than explanation. In truth, relative importance varies from one branch of science to another. For example, developmental psychology’s chief concern is to provide explanations of ontogenetic change. In contrast, the main task of developmental medicine is prediction, a task engendered by its primary concern for gaining control over diseases. In fact, developmental medicine does not necessarily require scientific explanations as, through an accumulated body of clinical knowledge, its practitioners can foretell the course of a particular disease with considerable accuracy from the signs and symptoms. On the basis of this knowledge, they are in a position to make rational forecasts that are predictive statements lacking the explanans of a fully-fledged scientific prediction. Such statements involve the reconstruction of the non-normative conditions under which previous recurrences of the disease have taken place. However, scientific predictions are necessary when the credibility of one predictive statement is challenged by another one derived from clinical practice. In this way, improvements in the control of a particular disease are attained. This aim provides, for example, the rationale for carrying out perinatal follow-up studies.
See Chaos theory, Cross-sectional design, Deductive-nomological (D-N) model, Description, Explanans and explanandum, Explanation, Precision, Prodiction and retrodiction, Prognosis, Theory