What is Hypothesis in Research?
The word ‘hypothesis’ is derived from the Greek term ‘hypotithenai’ which means ‘to suppose’ or ‘to put under.’ According to Etymology, the word hypothesis is the combination of the two words, i.e., ‘hypo’ and ‘thesis’ that means less than a thesis. A hypothesis is a tentative statement that is assumed for explaining the relationship between the various independent and dependent variables involved in the research. The hypothesis is often confused with the simple guess, but the hypothesis is a logical and well-calculated guess based on the previous research related to the current areas of research. Researchers conducted several studies to understand various phenomena, however, one must have the knowledge of his/her expectations from the study before conducting any experiments related to the research. Hence, a hypothesis is formulated before conducting the experiment, which is then proved right or wrong at the end of the experiment. The formulation of the Hypothesis before conducting the research helps the researchers to focus only on the crucial factors of the research problems under investigation. Hypothesis guides the researchers on what type of sample population should be considered under the research and what type of data needs to collect and what to avoid.
According to the social researchers William J. Goode and Paul K. Hatt, the hypothesis is defined as,
a proposition which can be put to test to determine its validity”
According to an American sociologist, George A. Lundberg,
A hypothesis is a tentative generalisation, the validity of which remains to be tested. In its most elementary stage, the hypothesis may be any hunch, guess the imaginative idea, which becomes the basis for action or investigation”
What are Variables?
Let us understand the notion of variables before discussing the different types of hypotheses. In simple terms, anything that can change, i.e., do not remain constant is known as a variable. A variable can also be defined as an entity that can have different values. For example, the age of the person is a variable as it can take different values according to the age of the same person or the age of the different people. Similarly, the height of the person can be variable as it can vary for the same person (to some extent) and different people have the same or different heights. Hence the variable is an entity that can be expressed in measurable terms. In psychological research, the measurable terms refer to the characteristics, traits, or attributes of the person under investigation, for example, age, height, weight, hair colour are all the different variables associated with the person. As there exist different attributes or variables associated with a person, the experiments conducted by the researchers consist of different types of variables. The variables are broadly classified into two types, i.e., the dependent variable and the independent variables. Let us discuss these two types,
1. Independent Variables
The independent variables refer to the variables that can be controlled or manipulated by the researchers in the study, which is why they are often termed as the manipulated variables. These variables are called independent because they do not depend upon the other factors involved in the study. In the experiments, the researchers usually look for the existence of the cause and effect relationship between the different variables. To determine the relationship, the researchers check the effect of the change in the independent variables on the other variables involved in the study.
2. Dependent Variables
Dependent variables are also known as outcome variables. These variables can be calculated statistically. The variables that change with the change in the independent variables are known as the dependent variables. For example, the marks you score in your exam is a dependent variable because the marks you will score in your exams depends upon several factors such as the amount of time your prepare for the exam, the stress level, the hours of sleep you took on the night before the exam, these all are the independent variables. Hence, we can say that the dependent variables are the result of the changes in the independent variables.
Relationship among the Variables
There can exit three types of relationships among the variables, these are briefly discussed below,
1. Positive Relationship
The relationship between the two variables is said to be positive when the increase in one variable results in an increase in the other variable. For example, the more you revise before your exam, the more you will score in the exam. Here the marks in the exam is a dependent variable, which shows a positive relationship with the independent variable, i.e., revision before the exam.
2. Negative Relationship
The relationship between the two variables is said to be negative if the increase in one variable results in a decrease in the other variable and vice-versa. For example, the more time you spend watching tv before your exam, the lesser marks you will score in the exam. Here, the dependent variable (marks in the exam) shows the negative relationship with the independent variable (watching tv).
3. Zero Relationship
If there does not exist any significant relationship between the two variables it is said to be zero relationships. For example, ‘you will feel less hungry if you read more.’ Here there does not exist any significant relationship between the reading and the hunger hence it is an example of the zero relationships.
Types of Hypothesis
1. Simple Hypothesis
The simple hypothesis assumes the relationship between the one dependent variable and the one independent variable. For example, if you eat more junk food, you will gain weight faster. Here, the independent variable is ‘eating more junk food’ and the dependent variable is ‘gaining weight.’ This type of hypothesis is also known as the basic hypothesis.
2. Complex Hypothesis
The complex hypothesis shows the relationship between the two or more dependent and independent variable. For example, If you add more healthy vegetables and fruits to your diet, it will result in a healthy body, a reduction in the risk of various diseases such as high or low blood pressure, heart attacks, or renal failure. Hence, in a simple hypothesis there exists the relationship between only two variables, while in a complex hypothesis there exist a relationship in multiple variables.
3. Associative and Causal Hypothesis
The associative hypothesis assumes the associative relationship between the dependent and the independent variables, i.e., if the one variable is changed the other variables also changes, but it does not clearly define that the particular variable results in the change of the other variable. The associative hypothesis only examines the simultaneous occurrence of two or more events. For example, a rise in the number of patients in a particular hospital does not mean that hospital is responsible for making the people sick, they might get sick due to the season change or the other factors. On the other hand, the causal hypothesis predicts the changes in the events to occur in the future due to the changes in the different variables. The hypothesis ‘the change in the font style of their webpages will result in higher engagement,’ is an example of a causal hypothesis.
4. Null Hypothesis
According to the Null Hypothesis, there does not exist any direct relationships between the two variables involved in the study, i.e., the dependent variables do not get affected by the changes in the independent variable. The null hypothesis is denoted by the symbol ‘H0.’ For example, ‘the growth of the plant remains the same whether you watered the plant with distilled water or with tap water.’ is a Null hypothesis.
5. Directional Hypothesis (One-tailed)
The directional hypothesis predicts the relationship between the two or more dependent and independent variables and the nature and direction in which the change might occur, i.e., less, more, greater, smaller etc. This type of hypothesis is focused on the research question, and it is validated through various statistical methods. The directional hypothesis is formulated in the research when the researcher is looking for a specific outcome from the experiment. For example, ‘if the teenagers join the moral development classes in the school for a continuous period of three years they will develop a higher moral character in adulthood than the ones who do not join the classes.’ is a directional hypothesis.
6. Non-Directional Hypothesis (Two-tailed)
The non-directional hypothesis predicts that there exists a relation between the two variables but does not anticipate the exact direction of the relationship between the two variables. The non-directional hypothesis is often used by the researchers in their studies if there exist contradictory findings regarding the same concepts under study, in the previous research. For example, one may hypothesize that the exists a difference in the performance of the high school students and the college students in the IQ test, but it does not predict whether high school students will perform better in the test or the college students will perform better.
7. Statistical Hypothesis
Any logical or illogical statement that can be verified through statistical methods is called the statistical hypothesis. For example, to analyse the statement ‘the consumption of Vitamin E capsules is good for the faster hair growth,’ the researchers have to verify it on the number of people with the help of statistical methods. A statistical hypothesis is the analysis of a small part of the target population with the help of statistical information about the targeted population. For example, if you want to examine the IQ level of the people living in a specific state, it would not be practically possible to analyse the IQ level of each member of that area, instead, you will choose a sample population and conduct several IQ tests on that sample population and draws the conclusion about the IQ level of the people belonging to that specific state on the basis of the results gained from the study over the sample population.
8. Empirical Hypothesis
The empirical hypothesis is also known as the working hypothesis. This hypothesis is just an assumption in the beginning, but it becomes true when tested through several experiments. This hypothesis is the opposite of the logical hypothesis as this hypothesis is tested through various scientific methods. One can find several findings of the hypothesise through the trial and error method. Following are some of the examples of empirical hypotheses that can be easily proven by different experiments over time.
- The skin glows more with the vitamin C capsules than the Vitamin E capsules.
- You will lose weight easily by doing exercise A than exercise B.
- The height grows faster with supplement X than supplement Y.
9. Logical Hypothesis
As suggested by the name itself, the logical hypothesis is tested logically. For example, ‘plants need water to survive,’ can be logically verified because living beings require water for their survival and plants are living. The logical hypothesis is widely used in philosophy because the questions related to the philosophy are generally untestable, hence the researchers validate them using the logical theories.
10. Alternative Hypothesis
The alternative hypothesis is also known as the research hypothesis or the maintained hypothesis. First, several hypotheses are formulated by the researchers and out of all these hypotheses, the most suitable one is chosen. The alternative hypothesis is denoted by the symbol H1 or Ha. To prove whether there exists any relationship between the variables or not, the researchers compare the null hypothesis and the alternative hypothesis. If the null hypothesis is rejected, then it means the alternative hypothesis is accepted, but if the null hypothesis is accepted then it means the alternative hypothesis is rejected. An example of an Alternative hypothesis is, ‘the productivity of the person in doing various daily life activities improves if he/she sleeps for eight hours a day rather than 9 hours a day.’
11. Composite Hypothesis
The hypothesis that does not assume the exact attributes of the dependent variable is considered as the composite hypothesis. For example, consider a hypothesis ‘People of the age 20 have the IQ of range 90.’ In this hypothesis, the dependent and independent variables are exactly defined, hence it is not an example of the composite hypothesis. But, it is difficult to exactly hypothesise some phenomenon, in that case, we can predict the occurrence of something but not exactly how it may occur. In that case, the researchers generally use the hypothesis in the format, ‘People of the age 20 do not have the IQ of range 90.’ This is called the composite hypothesis as the exact parameters are not mentioned in this case.
The hypothesis is not just a simple statement, it reflects the expectations and the possible outcomes of the experiments conducted by the researcher, hence it should be carefully designed as even minor flaws can adversely impact your whole study. One should take care of the following points before drafting the hypothesis for an experiment.
- Always do thorough research before formulating the hypothesis, read about the previous studies and theories that may help you in predicting the relationships among the variables under your investigation.
- Clearly understand the researcher problem that you are trying to solve and analyse the research question that could be researchable and focused on the area under your research.
- Clearly mention the dependent, independent, controlled and manipulated variables. As discussed above, independent variables do not depend upon the various factors involved in the study and the dependent variables depends upon the several factors (independent variables ) involved in the study.
- The language of the hypothesis should be clear to the reader and the variables involved in the study, the group under investigation, and the possible outcome of the research should be clearly mentioned.
- One can draft a simple hypothesis using the if-then statements, i.e., if a certain step is taken, then the particular outcome could happen. Here the first part represents the independent variable and the second part represents the dependent variable. For example, if you do exercise daily, then you can easily lose weight. Where ‘exercising daily’ is the independent variable and ‘weight losing’ is the dependent variable.
- If the research under investigation consists of statistical hypothesis testing, one needs to formulate the null hypothesis along with the alternative hypothesis. The null hypothesis represents the default case if there does not exist any correlation among the variables.
Null Hypothesis (H0): The time spent watching tv before the exam does not affect the marks obtained in the exam.
Alternative hypothesis (Ha): The time spent watching the tv before the exam has a negative effect on the marks obtained in the exam.
Characteristics of Good Hypothesis
As discussed earlier in this article, the hypothesis is not just a simple guess. The researchers generally draft the hypothesis based on the results of the previous studies as it is testable and more likely to be true. The hypothesis should be based on an effective research question that helps to understand the problem under investigation. Following are some of the characteristics of a good hypothesis.
- It should be simple and easy to understand so that the researchers that want to do further research on similar areas can easily continue the study.
- The hypothesis should clearly define the main focus of the experiment in a precise and clear manner. If the hypothesis does not focus on the main problem under investigation the results of the study may not be much reliable.
- The hypothesis should be testable through various methods.
- The hypothesis should clearly mention the relationship between the different variables involved in the study.
- The hypothesis should be related to the general and logical known facts, for example, it would be vague to hypothesize that ‘animals do not need water for their survival.’
- The hypothesis should be based on strong empirical research.
- The hypothesis should be designed for a specific problem, rather than drafting a hypothesis that can be generalised. The hypothesis may fail to draw the accurate interference if it can be generalised easily.
- The hypothesis should be helpful in discovering the new phenomenon and future studies. According to a researcher named J.S. Mill,
A hypothesis is the best source of new knowledge it creates new ways of discoveries”