Hypothesis research tries to determine what is happening in some situation by evaluating various conjectures. The goal is to determine which hypothesis is more likely to be true.
A scientific hypothesis is the initial building block in the scientific method. Many describe it as an educated assumption, based on prior knowledge and observation.
What is a Hypothesis?
A hypothesis is a suggested solution for an inexplicable occurrence that does not conform to current accepted scientific theory. The basic idea of a hypothesis is that there is no predetermined result.
For a hypothesis to be called a scientific hypothesis, it has to be something that can be supported or refuted by carefully crafted experimentation or observation.
A key function in this step in the scientific method is to derive predictions from hypotheses about the results of future experiments, and then perform those experiments to see if they support the predictions.
Hypothesis research is made up of three main activities:
- Generation of hypotheses: it is the exposition of the hypothesis as such.
- Hypothesis evaluation: refers to the evaluation of the relative plausibility of the hypotheses given the available evidence.
- Hypothesis testing: That is, the search for more evidence.
Types of Hypotheses
A null hypothesis is the name given to a hypothesis that is possibly false or has no effect. Often during a test, the scientist will study another branch of the idea that may work, which is called an alternative hypothesis. During a test, the scientist can try to prove or disprove only the null hypothesis or test both the null and alternative hypotheses.
If a hypothesis specifies a certain direction, it is called a one-tail hypothesis. This means that the scientist believes that the result will be with or without effect. When creating a hypothesis without predicting the outcome, it is called a two-tailed hypothesis because there are two possible outcomes. The result could be with or without effect, but until the test is complete, there is no way of knowing what the result will be.
During tests, a scientist can find two types of errors.
- A Type I error is when the null hypothesis is rejected when it is true.
- A Type II error occurs when the null hypothesis is not rejected when it is false.
How to write a hypothesis
- To write the null and alternative hypotheses for an investigation, we must identify the key variables in the study. The investigator manipulates the independent variable and the dependent variable is the outcome that is measured.
- Operationalize the variables being investigated. Operational variables (or operational definitions) refer to how you will define and measure a specific variable as used in the study.
- Decide a direction for prediction. If there is evidence in the literature to support a specific effect on the independent variable on the dependent variable, we should write a directional (one-tailed) hypothesis. If there are limited or ambiguous findings in the literature regarding the effect of the independent variable on the dependent variable, we should write a non-directional (two-tailed) hypothesis.
- Write the hypothesis. A good hypothesis is short (that is, concise) and includes clear and simple language.
Testing a hypothesis
The main feature of a hypothesis is that something can be tested and that those tests can be replicated. A hypothesis is often examined by multiple scientists to ensure the integrity and veracity of the experiment. This process can take years, and in many cases the hypotheses do not go further in the scientific method, since it is difficult to gather enough supporting evidence. After analyzing the results, a hypothesis can be rejected or modified, but it can never be proven 100% of the time correct. For example, relativity has been tested many times, so it is generally accepted as true, but there could be an instance, not found, that is not true.
The Evolution of Hypothesis Research
Most formal hypotheses consist of concepts that can be connected and their relationships tested. A group of hypotheses comes together to form a conceptual framework. As enough data and evidence is collected to support a hypothesis, it becomes a working hypothesis, which is a milestone on the road to becoming a theory. Although hypotheses and theories are often confused, theories are the result of a proven hypothesis. Although hypotheses are ideas, theories explain the test results of those ideas.
A related, though subtly different, notion is that of hypothesis-driven research, in which a single hypothesis is selected relatively early in the process, and most of the effort is devoted to substantiating this hypothesis. It is hypothesis-based research with all the attention focused on one conjecture, at least as long as you are not forced to reject it and consider another.
Pitfalls in Hypothesis Research
Hypothesis research fails, in its simplest form, when we take the incorrect hypothesis as true. This can have dire consequences if costly actions are taken. Hypothesis research also fails when there is wrong or excessive confidence in a hypothesis, even if it turns out to be correct. It leads to no conclusion, when more careful investigation might have revealed that a hypothesis was more plausible. There are three main pitfalls that lead to these failures.
That is, without considering the full range of reasonable hypotheses. Much effort is put into investigating one or some hypotheses, usually obvious, while other possibilities are not considered at all. Too often one of those others is, in fact, the right one.
Abusing the evidence
Here the evidence already available is not properly evaluated, leading to erroneous evaluations of the plausibility of the hypotheses. A particular piece of evidence could be considered stronger or more significant than it actually is, especially if it seems to support your preferred hypothesis. In contrast, "negative" evidence, one that directly undermines your preferred hypothesis, or appears to strongly support another, is considered weak or worthless. Furthermore, the entire body of evidence that is based on a hypothesis could be poorly rated. Some fragments of lousy evidence could be considered collectively equivalent to a solid case.
Search in the wrong places
When you search for additional evidence, you instinctively look for information that is actually useless or at least not very helpful in terms of helping you determine the truth. In particular, we are prone to "confirmation bias", which seeks information that gives weight to our preferred hypothesis. We tend to think that by accumulating a large amount of supporting evidence, we are rigorously testing the hypothesis. But this is a classic mistake.
But this is a classic mistake. We need to know not only that there is a lot of evidence consistent with our preferred hypothesis, but also that there is evidence inconsistent with the alternatives. You should look for the correct type of evidence relative to your entire set of hypotheses, rather than just a lot of evidence consistent with one hypothesis. This can have two unfortunate consequences. The search may be ineffective: no evidence is ever found that might have strongly discarded one or more "inside" or "outside" hypotheses.
General guidelines for investigating good hypotheses
Investigate a wide range of hypotheses
Our natural tendency is to hold on to the first plausible hypothesis that comes to mind and start shaking it hard. This must be resisted. From the beginning, you should examine as wide a range of hypotheses as reasonably possible. It is impossible to analyze all the hypotheses and it is absurd to even try.
But a wide selection of hypotheses can and should be considered, including at least some "far shots". In generating this set of hypotheses, diversity is at least as important as quantity. We must continue to search for additional hypotheses throughout the investigation. Incoming information may suggest interesting new possibilities.
Investigate multiple hypotheses
At any moment we must keep a series of hypotheses at stake. In hypothesis tests, that is, in the search for new information, we must look for information that discriminates what will be revealing in relation to multiple hypotheses at the same time.
Look for evidence that does not confirm
Instead of trying to prove that some hypothesis is correct, we should try to prove that it is false. Ideally, we should try to disconfirm multiple hypotheses at the same time. This may be easier if the set of hypotheses is hierarchically organized, allowing you to search for evidence that we eliminate entire groups of hypotheses at once.
Some methodologies have been developed to assist with hypothesis research. Methodologies have some important advantages over proceeding intuitively or spontaneously. They are designed to help us avoid cheating, and they do so by incorporating, to some extent, the above general guidelines. They provide distinctive external representations that help us organize and understand sets of hypotheses and evidence.
The hypothesis is what the researchers predict in relation to two or more variables, but it involves more than an assumption. Most of the time, the hypothesis begins with a question that is then explored through background research. It is only at this point that researchers begin to develop a testable hypothesis.
Kim, J.O. y Mueller, Ch. (1978 a). Introduction to factor analysis. Sage University Paper. Serie: Quantitative Applications in the Social Sciences, no. 13. Beverly Hills y Londres: Sage Publications.
Mora y Araujo, M. (1971a). El análisis de relaciones entre variables, la puesta a prueba de hipótesis sociológica. Buenos Aires: Nueva Visión.
Mora y Araujo, M. (1971b). Medición y construcción de índices. Buenos Aires: Nueva Visión.