# Null hypothesis | Volkig

The **null hypothesis** is an assumption that is used to affirm or deny a certain event that is related to one or more parameters of a population or sample. When a conclusion is reached related to an experiment, the person doing the research should establish both the null hypothesis and the alternative hypothesis.

If it refers to the **null hypothesis**, it is a contrary statement to the one that the researcher or person has reached when he tries to reject it. In case you have enough evidence, you could check that the opposite is totally true. This means that the **alternative hypothesis** would be the conclusion reached by the researcher through his research.

The statement of this **hypothesis** cannot be rejected, unless the data in the sample shows that they are false. That is why mostly a no or an unequal a, in its statement.

# Why is it Called the “Null”?

The word** “null”** can be understood as “unchanged.” Normally, a null hypothesis is the standard assumption that is defined as the prediction that there is no interaction between the variables.

**For example, **the **null hypothesis** indicates that there is no causal relationship between a new treatment and a reduction in disease symptoms. In other words: a new treatment does not offer any improvement with respect to the reference treatment, and an observation or improvement is the result of chance.

Such a claim can be proven by scientific study, such as a clinical trial, and the application of appropriate statistical tests. If a clinical trial discovers that a relationship really exists and the new treatment produces an improvement, the **null hypothesis** is shown to be false and can be rejected. In this case, the alternative or **research hypothesis** can be adopted; in this example, it means that the new treatment is better than the reference treatment.

# What is the null hypothesis?

In statistics, a hypothesis refers to the affirmation of a population parameter and is represented by H0. When it comes to the **null hypothesis**, it means that it is the statement that two or more parameters have no relationship to each other. It is the starting point for a certain investigation that is not rejected unless the data is not correct.

Basically it is an application in statistics of the method of reduction to the absurd, where it is assumed from the beginning, the opposite of what you want to prove, until the evidence or conclusions that are acquired allow the demonstration that the starting point was false, then it would be rejected and the opposite is concluded.

Since this **hypothesis** has the logical formula of a universal statement, in order to confirm that it is true, it is necessary to study the entire population.

Its term comes from the first agricultural and medical applications of statistics, this in order to affirm how effective new fertilizers or new medicines are. In that case, the hypothesis where it begins shows that there was no effectiveness or difference between the samples that were treated and those that were not treated.

If the results of the sample were to fail to support the hypothesis, then it should be rejected and the conclusion accepted and confirming the existence of any link between the samples would become alternative H1 hypotheses.

**The** **null hypothesis (H 0 ) is an essential part of any research design and is always tested, even if indirectly.**

The simplistic definition of the null hypothesis is almost contrary to that of the alternative hypothesis (H 1 ), although the principle is slightly more complex.

The null hypothesis (H 0 ) is a hypothesis that the researcher tries to refute, reject or annul.

Generally, **“null”** refers to the general opinion of something, while the alternative hypothesis is what the researcher really thinks is the cause of a phenomenon.

The conclusion of an experiment always refers to the null, that is, it rejects or accepts H 0 instead of H 1.

Despite this, many researchers neglect the null hypothesis when testing hypotheses , which is bad practice and can have adverse effects.

# Examples of the null hypothesis

## An investigator can postulate a hypothesis:

And a null hypothesis:

It is important to carefully select the null text and make sure it is as specific as possible. For example, the researcher can postulate a null hypothesis:

There is a big flaw with this H 0 . If the plants really grow more slowly in the compost than in the soil, you reach a dead end. H 1 is not supported and neither is H 0 , since there is a difference in growth rates.

If the **null hypothesis** is rejected and there is no other option, the experiment may be invalid. For this reason, science uses a series of deductive and inductive processes to ensure that there are no errors in the hypotheses.

Many scientists neglect the null hypothesis, assuming it is just the opposite of the alternative, but the right thing to do is take time to create a solid hypothesis. It is not possible to change a hypothesis retrospectively, not even H 0.

# Significance tests

If the significance tests generate a 95% or 99% probability that the results do not fit the null hypothesis, then this is rejected in favor of the alternative.

Otherwise, the null hypothesis is accepted. These are the only correct assumptions and it is wrong to reject or accept H 1.

The acceptance of the null hypothesis does not mean that it is true. It remains a hypothesis and must comply with the principle of falsifiability., in the same way, that rejection of the null hypothesis does not prove the alternative.

# Problems perceived with the null

The main problem with H 0 is that many researchers and reviewers feel that accepting the null hypothesis constitutes failure of the experiment . This is very bad science, as well as accepting or rejecting any hypothesis is a positive result.

Although the null hypothesis is not refuted, the world of science has learned something new. Strictly speaking, the term “failure” should only apply to errors in experimental design or incorrect initial assumptions.

# Development of the null

The Flat Earth model was common in ancient times, such as in the Bronze Age and Iron Age civilizations. This could be considered the null hypothesis (H 0 ) of the moment.

Many of the ancient Greek philosophers assumed that the sun, moon, and other objects in the universe circled the Earth. Hellenistic astronomy established the spherical shape of the earth around 300 BC.

Copernicus had an alternative hypothesis (H 1 ) indicating that the world was actually circling the Sun, the centre of the universe. Over time, people became convinced and accepted it as null or H 0.

Subsequently, someone proposed an alternative hypothesis that the same sun was also circling something within the galaxy, thus creating a new H 0. This is how research works: H 0 gets closer to reality. Although it is not correct, it is better than the last H 0.

# Formulation of a null and alternative hypothesis

In order to have a better understanding of the **correlation between null and alternative hypotheses** , an example such as the following can be used:

- Assuming that a null hypothesis indicates that there is no cause and consequence relationship between a certain treatment that is being tested for being new and the decrease in the symptoms of a disease.
- It could be said that according to the hypothesis, the new medicine does not generate the expected improvement in relation to the medicine that has been used up to now, which means that any improvement that occurs would be a fluke.
- In order to verify and if the case should reject this
**hypothesis**, a scientific study on the new treatment would have to be carried out. It is discovered that there is an effective cause and consequence relationship between the new medicine and the improvement of the patient’s disease, it can be shown that the hypothesis is totally false. - In the event that this occurs, an alternative hypothesis can be used, since it can be determined that the new medicine has given a better result than the previous one because it demonstrated good progress in the patient.

# Null hypothesis vs. alternative hypothesis

**A statistical hypothesis test is presented as a decision between two hypotheses.**

The null hypothesis consists of a statement about the population of origin of the sample. Usually, it is simpler (fewer parameters, for example) than its antagonist. The null hypothesis is designated with the symbol H 0 .

The alternative hypothesis is also a statement about the source population. Many times, but not always, it is simply to deny the statement of H 0 . The alternative hypothesis is designated by the symbol H 1 .

At the moment we will deal with the simplest case, in which the two hypotheses refer to a single parameter value. In this general situation, the hypotheses refer to a parameter θ ( theta ). The formulation is:

H 0 : θ= θ 0

H 1 : θ = θ 1

In hypothesis testing theory this type of approach is known as simple versus simple hypothesis testing .

Thus, a simple hypothesis postulates that the parameter θ can only take one value or, more technically, that the set of parameters associated with a simple hypothesis consists of a single point.

## Examples of hypotheses to be tested

# Case 1: Null and alternative hypothesis.

**Data obtained in the study**

Dr. da Souza Fairia postulates the same ratio for males and females. In terms of the proportion of the variable X (number of females in ten nests) this is equivalent to the hypothesis that the proportion ( in the population ) is 0.5 .

On the other hand, according to Dr. Calves the ratio is 6: 4 in favor of females, then it is equivalent to the hypothesis that the parameter p in the Binomial variable is 0.6 .

Thus, if X is the number of females in ten nests, and p is the proportion of females, the final form of the contrast is:

H 0 : p = 0.5

H 1 : p = 0.6

As for the data obtained by da Souza they are :

In summary, he has observed in six of the nests there is a female .

# Case 2: Null against alternative.

**Breakdown of study data**

Sports authorities postulate an average of 7.0 ng / ml, while the ADG indicates an average of 8.5 ng / ml for individuals undergoing this type of diet. Therefore, in synthesis, the contrast will consist of:

H 0 : μ = 7.0

H 1 : μ = 8.5

For both H 0 and H 1 , σ = 2.4 will be taken .

The data of the study that the ADG association has obtained, and that according to them support their thesis, have been the following:

The arithmetic mean of the sixteen athletes is 8.54 ng / ml

*Originally published at **https://volkig.com*