THE 8 TYPES OF RESEARCH HYPOTHESES (WITH EXAMPLES) - SCIENCE - 2025

A hypothesis determines the possible characteristics of the variables and the relationships that exist between these variables. All scientific research must start from one or more hypotheses that are intended to be demonstrated.

A hypothesis is an assumption that can be verified through scientific investigation. In other words, hypotheses are the formulation of the problem: they establish possible relationships between variables.

There are many different ways to classify hypotheses according to different criteria. The most common is the one that distinguishes between null hypotheses, general or theoretical hypotheses, working hypotheses and alternative hypotheses. In turn, within each category, different subtypes are identified.

Hypotheses and the scientific method

During the scientific method an attempt will be made to demonstrate the validity of a main hypothesis. This is known as a working hypothesis. If you want to investigate several plausible hypotheses, you will make alternative hypotheses. Within the working and alternative hypotheses there are three subtypes: attributive, associative and causal hypotheses.

Contrary to the working and alternative hypotheses, which quantify the relationship between the variables, the general or theoretical hypotheses establish a conceptual relationship between them. On the other hand, there is also the null hypothesis, which is one that determines that there is no relevant relationship between the variables under study.

If the validity of the working hypothesis and the alternative hypotheses cannot be demonstrated, the null hypothesis will be accepted as valid. Besides these, there are other types of hypotheses, such as relative and conditional ones. They can also be classified according to other criteria; for example, it is possible to distinguish between probabilistic and deterministic hypotheses.

Types of main hypotheses in a scientific investigation

-Null hypothesis

The null hypothesis assumes that there is no relationship between the study variables. For this reason it is also known as the hypothesis of no relation.

This hypothesis will be accepted if the investigation shows that the working hypothesis and the alternative hypotheses are not valid.

Example

"There is no relationship between students' hair color and their academic results."

-General or theoretical hypotheses

The general or theoretical hypotheses are those that are formulated in a conceptual way, without quantifying the variables.

Normally, these hypotheses are obtained through a process of induction or generalization from the observation of similar behaviors.

Example

"The more hours a student studies, the better grades he gets."

Among the theoretical hypotheses are the difference hypotheses, which are those that determine that there is a difference between two variables but do not measure its magnitude. For example, "in the university the number of national students is greater than the number of international students."

-Work hypothesis

The working hypothesis is one that is to be demonstrated or supported by scientific research.

These hypotheses can be verified experimentally, which is why they are also called operational hypotheses.

In general, they are obtained from deduction: starting from general laws that are particularized in a specific case. The working hypotheses can be attributive, associative or causal.

- Attributive

The attributive or point-prevalence hypothesis describes the facts. This hypothesis is used to describe real behaviors, which are measurable and can be distinguished from other behaviors. The attributive hypothesis is composed of a single variable.

Example

"Most of the university students are between 18 and 23 years old."

- Associative

The associative hypothesis establishes a relationship between two variables. If the first variable is known, the second can be predicted.

Example

"There are twice as many students in the first grade as in the last."

- Causal

The causal hypothesis determines a relationship between two variables. The increase or decrease of the first variable determines an increase or decrease in the second variable. These variables are called "cause" and "effect", respectively.

To prove a causal hypothesis, the existence of a cause-effect relationship or a statistical relationship must be determined. It can also be proved by eliminating alternative explanations. The formulation of these hypotheses is of the type: "If… then…".

Example

"If a student studies an additional 10 hours a week, then his grades improve one point out of ten."

-Alternative hypotheses

The alternative hypotheses try to answer the same problem as the working hypotheses. However, as the name suggests, they look for different possible explanations. Thus it is possible to test different hypotheses in the course of the same investigation.

Formally, these hypotheses are analogous to the working hypothesis. They can also be classified into attributive, associative and causal.

Other types of hypotheses

Some authors identify other types of less common hypotheses. For example:

-Relative hypotheses

Relative hypotheses assess the influence of two or more variables on another.

Example

"The effect of the rise in prices on the number of university students is less than the effect of the fall in wages on the number of university students."

Variable 1: price increase

Variable 2: drop in wages

Dependent variable: number of university students.

-Conditional hypotheses

Conditional hypotheses assume that one variable is dependent on the value of two others. In this case, the hypotheses are similar to the causal ones, but there are two "cause" variables and one "effect" variable.

Example

"If the student does not bring the exercise and is late, he will be expelled from the class."

Cause 1: not bring the exercise.

Cause 2: being late.

Effect: being expelled.

For the variable "effect" to be fulfilled, it is not enough for one of the two variables "cause" to be fulfilled: both must be fulfilled.

Possible alternative classifications

The classification of scientific research hypotheses that has been exposed is the most common. However, it is also possible to classify hypotheses based on other criteria.

For example, it is possible to distinguish between probabilistic and deterministic hypotheses.

-Probabilistic hypotheses

These hypotheses suggest that there is a relationship between variables that is true in most of the population.

Example

"If a student does not study, he will fail."

-Deterministic hypotheses

These hypotheses pose relationships between the variables that are always true.

Example

"If a student does not appear for the exam, he will fail."

References

1. Fernández Guerrero, G. Research methodology. University of London. Available at: s3.amazonaws.com
2. Kumar, R. 1999. Research Methodology. A step-by-step guide for beginners. London: SAGE Publications Ltd. Available at: sociology.kpi.ua
3. Powner, LC 2015. Empirical Research and Writing: A Political Science Student's Practical Guide. Singapore: CQ Press.
4. Sabino, C. 1992. The research process. Caracas: Panapo.
5. Sacramento City College. Research hypotheses: types. Available at: scc.losrios.edu