Just imagine a world where complete certainty exists. In this case, all the future would change and events would be known in advance and predicted precisely. Therefore, there would be no surprises and no errors. Hence, we could know all our future actions, as well as their exact results. In a world like this one, nothing can be learned, and nothing would be worth knowing (Lindley, 2006). As a result, the accurate foresight would remove the very need of knowledge. This need can arise only if, like in our world, foresight is imperfect, and as long as knowledge is a road map towards bringing preferences. It never follows that everything is uncertain (Cressie & Wikle, 2011). On the other hand, in a world where results of all actions are known with certainty, the idea of everything being certain would not even come into existence (Lindley, 2006).
Therefore, the idea of certain knowledge needs, as its logical partner, the idea of uncertainty. Hence, uncertainty always surrounds us, and we cannot expect certainty. However, certainty is often quantifiable (Cressie & Wikle, 2011). We quantify it when we speak about the degree of uncertainty or certainty. This is the statistical idea of probability: higher probability means higher degree of certainty that something will happen. Statistical techniques are, therefore, designed to help us understand those areas with uncertainty where it is quantifiable. Majority of statistical techniques are based upon probability (Lindley, 2006). It is worth noting that although statistical methods cannot remove uncertainty, they can help us to gain some knowledge regarding its existence. For instance, they can aid us in seeing the patterns through it. We are, therefore, able to quantify the certainty/uncertainty in order to show that the portrayed patterns are true and not just artifacts of our perception or data.
Mathematics scholar Alan Schoenfeld has nicely indicated those areas where statistics might be applied (Cressie & Wikle, 2011). He gives an example of evolution theory, about which scientists generally agree as regards its basic correctness. However, the evidence, which has been used in support of evolution, is accurate in a mathematical sense. The points raised in its support consist of patterns of plausible reasoning along with consideration of other alternative hypotheses (Cressie & Wikle, 2011). In fact, biologists are asserting that there is enormous evidence that is consistent with the theory and broadly construed; there is no other clear evidence that could falsify the proposed theory, and no other rival theory that would meet the same criteria. Therefore, in some areas, it is impossible to expect certainty or anything close to it after just one study. However, when there is a lot of evidence on the basis of high quality research, we can get a high degree of certainty. Cases of reduced uncertainty are referred to as risk. Thus, statistics may be helpful for players in several sectors, such as insurance cover provision among others. Through the use of statistics we are able to change a situation from uncertain to risk (Cressie & Wikle, 2011).
Uncertainty, on the other hand, means that we do not know what will happen next, and neither do we know how its distribution looks like (Lindley, 2006). Finally, the philosophical problems associated with uncertainty are usually tied to a number of conditions. These are limits of what can be known about the past and the future, fallibility of human beings, spontaneity, and contingency of the human decisions and the actions which affect social phenomena.