How to solve a problem, which needs to be considered is to find the cause of the problem. To search for the causes of these problems, research is conducted. So that research can be directed, so need to be formulated prior estimation of the causes of the problem. Estimation of the cause of the problem is called a hypothesis. The hypothesis consists of two words, namely Hypo (which means doubt), and a thesis (which means truth). So the hypothesis means that the truth is still in doubt.
According to Kerlinger, the hypothesis is temporary or proposition tentative conclusions about the relationship between two or more variables; while according to Bailey, the hypothesis is a proposition that is expressed in a form that can be tested and foresee a certain relationship between the two variables (Malo and Trisnoningtias, 1990: 39). The hypothesis will be rejected if any, and acceptable if the facts justify the study. Therefore, rejection and acceptance of the hypothesis is dependent upon the results of empirical research.
The hypothesis can also be viewed as a temporary conclusion. As a conclusion is certainly hypothesis made in vain, but on the basis of certain knowledge that the majority can be taken from the results of previous studies, and theories that are relevant. The hypothesis has a steering function, which gives the boundaries of the kinds of data that should be collected, methods of data collection, and analysis models (Mantra, 2001: 10).
A scientific research hypothesis must meet certain requirements, which are:
- First, the hypothesis is the result of construction of ideas that can be explained by theories or results of particular observation;
- Secondly, the hypothesis should be formulated in the form of a statement (statement) and should never be in question;
- Third, the hypothesis has always been associated with the state of the population, not just the state of the sample studied, the study sample only serves as a platform or vehicle for hypothesis testing, the results of research on the samples will be generalized to the population of source samples taken;
- Fourth, the hypothesis must be involved at least two variables (changes), a statement of only one variable is not a hypothesis to be tested;
- Fifth, a research hypothesis should be tested, so that a hypothesis can be tested.
At least there are three kinds of hypothesis formulation, which is descriptive (describing the characteristics of a unit start being focused research), correlation (describing the relationship between two or more variables but did not indicate which variable is the cause and which variable becomes due in that relationship), and causality (which have shown variable is the cause and which variable becomes due) [See: Malo and Trisnoningtias, 1990: 40-41]
A good hypothesis criteria:
- Developed with existing theories, logical explanation or the results of previous research.
- The hypothesis shows clearly the intention.
- Hypothesis testable
- This hypothesis is better than the competition hypothesis.
Kinds of Hypothesis
1. Hypothesis Descriptive
Descriptive hypothesis is a conjecture on the value of the variable in one sample in it even though there can be multiple categories.
Descriptive Hypothesis example:
Research issues: Does the acceptance of the "Peace in Manokwari" have a difference in those who come from a particular environment?
- Level of education pursued someone allows an openness to accept the peace process.
- A person's values are the basis for acceptance influence the peace process.
- The level of information that a person can give their views on a peace process.
People who came from open social environment more receptive to the peace process.
- People with higher education are relatively more receptive to the peace process.
- People-oriented modern values more accepting peace process.
- People who have a lot of information more easily accept the peace process.
Correlational hypothesis is a hypothesis that contains a statement about the relationship between two or more variables. If the pattern of relationship between two or more variables are causal (causal), then the hypothesis is called the hypothesis of causality
Example Hypothesis Correlations:
Research Issues: Matters relating to the level of a Company's Production.
- The number of experts in a company related to the level of production
- Experts will be difficult to work under strict work rules.
- Work rules in the company related to the level of production.
The greater the number of experts within a company, the lower the level of stringency of the company's work rules, dealing with h accept peace process increasing production yields.
Correlational Hypothesis Consists of Causal and Correlation Hypotheses
a. Causality Hypothesis
Example Hypothesis Causality:
Research Problems: Why the tendency of committing a crime in a community environment.
- An environment in society have an absorption power, ie absorption or silencer to a social phenomenon that DAPT cause shock
- One can feel frustrated if feel excluded from their communities.
- Someone who feels frustrated more easily stimulated to tend to commit criminal acts.
For those in the community who are very low power absorption if they feel increasingly excluded from society, then they are more easily aroused to tend to commit criminal acts.
b. Hypothesis Correlation
Hypothesis correlations (Correlational hypothesis) is the hypothesis that the two variables occur simultaneously without known which affect the others.
- Ha: There is a positive relationship between the amount of compensation and the company's profit.
3. Hypothesis Association
Measurement of association is a general term that refers to a group of techniques in bi-variate statistics were used to measure the strength of the relationship between two variables
Working hypothesis (Hk) and Zero hypothesis (Ho)
Hypotheses formulated by researchers, whether they are descriptive, relational and causality hypothesis is called the working hypothesis (Hk). So that the working hypothesis that can be tested statistically, it would require a comparison hypothesis. In the comparative social research hypotheses made arbitrary shaped null hypothesis (Ho). The null hypothesis (Ho) is the formulation / formulation of the inverse of the working hypothesis (Effendi, 1989: 43-45).
Examples of Working hypothesis (Hk):
Aggressive action is higher in communities that have a high population density than those with low density.
If the perception of the attitude of a model group controlled, husband and wife who have a fixed income jobs, have a low perception of the economic value of children, and therefore tend to be more accepting of a small family norm. Both lead their high perception of the benefits of modern contraceptive use, so the intent as well as their modern contraceptive use is relatively higher when compared with the husband and wife who have no fixed income jobs.
Examples of the null hypothesis (Ho):
There were no differences between the aggressive actions of people with high population density and people who have high levels of low population density.
If perceptions of peer group attitudes are controlled, there was no significant difference between couples who have a fixed income jobs and income is not fixed in the perception of the value of a child, a small family norms, perceptions about the benefits of modern contraception, and the intention of using modern contraception and behavior.
I think enough the description of Hypothesis that I can share. Hopefully this article is useful for us. Amin.