A randomized experiment is a test that consists in repeating a random phenomenonwith the objective of analyzing it and drawing conclusions about its behavior.
From the definition of random experiment itself, as well as the definition of random phenomenon, we deduce that it is the study of situations dominated by the laws of chance.
Not always if we try to perform a random experiment, we will be able to experiment in a tangible way. For example, let’s imagine that we want to study the behavior of a currency. The currency is tangible, we can see and touch it. Throw it and check the result (face or cross) corresponds to us. Now, let’s assume the example of the weather. We cannot move clouds or change temperatures. At least tangibly.
In line with the above we will have to be aware of the importance of the assumptions of some experiments. For this, the use of the axiomatic method is recommended. See axiomatic method.
Types of experiments
The objective of this article is to develop the concept of random experiment. However, to understand it better, we must understand what types of experiments exist. That is, answer the question: What if the results of the experiment, under the same conditions, are always the same? In that case it would no longer be a random experiment. In this sense we can distinguish two types of experiments:
- Deterministic experiments:Those that can be predicted with accuracy.
- Random experiments:These are those experiments whose outcome is uncertain.
It should be noted that the fact that an experiment is random does not mean that it is unpredictable. In fact, in some cases regularity models appear that allow us to guess a considerable number of times with some probability.
The previous paragraph reflects the importance of differentiating between a deterministic experiment and a randomized experiment. In the first case, it makes no sense to talk about probability. If we can predict, in all cases, the final result, the probability of success is 100% and 0% wrong. However, in random experiments (although there are repetitive patterns that characterize them) we cannot predict them accurately. And therefore it makes sense to talk about probability or possibility. See probability definition
Can it be a random experiment in deterministic reality?
On some occasions, less than we would like, we encounter deterministic phenomena. For example, some matters of physics or chemistry. To illustrate any of them, we know without any margin of error that if a person ingests 1 liter of mercury he will die. In the same way if we throw a stone through the window, we know that in a few seconds it will fall to the ground. We can even calculate the time in a very approximate way.
In other cases, the matter is not so clear. For example, in the case of economics there are currents of thought that indicate that it is deterministic and others that it is random. Or better yet, the case of the stock market. Many operators think it is deterministic, while others think it is totally random.
What should be indicated in this case is the following: the fact that something cannot be predicted (because we are not capable) does not prove that it is random. That is, the absence of proof does not necessarily constitute proof of absence. In other words, that I cannot see it does not mean that it does not exist.
Therefore, in line with the above there are currents of thought from both sides. Going from the most extreme thought that affirms determinism, to the opposite thought that affirms randomness. Among them, there are intermediate positions. For example, we can think that stock prices are deterministic, but as we cannot prove it, we treat them (especially statistically) as if they were random.