A research hypothesis is expected to be a straightforward and concise statement of prediction, predicting the expected result from an academic research project. It identifies the expected results in concrete (rather than theoretical) terms. However not all studies are given hypotheses, some studies are designed to be exploratory, leaving out the need for a formal hypotheses, while some studies are carried out to develop hypothesis for the purpose of further research, such studies do not have a predicted hypothesis but the final result of such studies helps to formulate hypothesis to be tested in future research. But let it be noted that every true experimental design must have this statement at the core of its structure, as the ultimate aim of any experiment.
Prior to writing research hypotheses it becomes very important to evaluate the general research question posed in the academic research paper. This can sometimes be very difficult to pin down as a lot of students often have an unstated sense of what they want to achieve in a study this makes it challenging to clearly state a research question. Glenn Firebaugh (2008) identified two key criteria for research questions:
- Research questions must be researchable: Researchable implies that a question can be answered through empirical research (that is, something that science can address) and also limited enough that a study could actually hope to answer the question in a reasonable period of time.
- A research question should be interesting: The requirement that the research question be interesting implies primarily that the question be important in the context of the ongoing scientific discussion of the topic (that is, interesting to other researchers).
Hypotheses in Quantitative Studies
Academic Research Hypothesis in quantitative studies takes a unique form: one independent variable, one dependent variable, and a declaration about the expected relationship between them. Generally the independent variable is mentioned first followed by language implying causality (terms such as explains, results in) and then the dependent variable; the ordering of the variables should be consistent across all hypotheses in a study so that the reader is not confused about the proposed causal ordering. When both variables are continuous in nature, language describing a positive or negative association between the variables can be used (for example, as education increases, so does income). For a research hypotheses with categorical variables, a statement about which category of the independent variable is associated with a certain category of the dependent variable can be made (for example, men are more likely to support PDP than women). Continuous variables can also be spoken about it categorical terms (those with higher education are more likely to have high incomes).
Most researchers prefer to present research hypotheses in a directional format, meaning that some statement is made about the expected relationship based on examination of existing theory, past research, general observation, or even an educated guess. It is also appropriate to use the null hypothesis instead, which states simply that no relationship exists between the variables; recall that the null hypothesis forms the basis of all statistical tests of significance. A compromise position is to present a research hypothesis which states a possible direction for the relationship but softens the causal argument by using language such as “tends to” or “in general.”
Hypotheses in Qualitative Studies
Hypotheses in qualitative studies are quite different from that of quantitative study. Due to the inductive nature of qualitative studies, the generation of hypotheses does not take place at the outset of the study. Instead, hypotheses are only tentatively proposed during an iterative process of data collection and interpretation, and help guide the researcher in asking additional questions and searching for disconfirming evidence.
Qualitative research is guided by central questions and sub questions posed by the researcher at the outset of the study. These questions usually employ the language of how and what in an effort to allow understanding to emerge from the research, rather than why, which tends to imply that the researcher has already developed a belief about the causal mechanism. In general a qualitative study will have one or two central questions and a series of five to ten subquestions that further develop the central questions. These questions are often asked directly of the study participants (through in-depth interviews, focus groups, etc.) in recognition of the fact that developing an understanding of a particular phenomenon is a collaborative experience between researchers and participants.
Types of Hypothesis
we have two basic types of research hypotheses and they are: Null and Alternative. Null hypothesis is stated in the negative form e.g. there is no significant relationship between employee motivation and job satisfaction of librarians in the university library. It is being denoted with Ho sign. The Alternative hypothesis is stated in the positive or affirmative form e.g. there is significant relationship between employee motivation and job satisfaction of lecturers in the university library. It is being denoted with Hi sign.
Sources of Hypotheses
Hypotheses could be derived from the following sources (Akinade&Owolabi, 2009):
- Directly from research questions
- Directly from research designs
- Directly from research variables
- Review of literatures
- Directly from research objectives
- Directly from research designs
- Directly from findings of other studies
- From personal or practical experience or observation and analogy
- Personal Judgement influenced by your careful evaluation of the research question, research objective, research design and a review of past studies
- Logical reasoning
- Suggestions from students or colleagues or subject experts
A testable hypothesis makes a statement about a presumed or theoretical relation between two or more variables (Kerlinger, 1986) – that can be observed or measured. Thus, a testable hypothesis should state or imply that the variables are observable and measurable. It should also specify the relation among the variables, e.g. higher grades are obtained after studying in the library than studying in the hall of residence.
Non-testable hypotheses may include, “abortion is wrong“, “stigmatizing ebola patient is bad”, “I seem to study better in the library than in your room” (Akinade&Owolabi, 2009, p. 19).
Testing of Hypotheses
The following steps should be observed when testing any research hypothesis (Awoniyi et al, 2011):
- Formulate a null hypothesis (Ho). Hypothesis can either be accepted or rejected.
- Set up a suitable level of significance. It means the level at which to accept or reject null hypothesis. It is conventional among researchers to accept or reject hypothesis at 0.05 or 0.01 level of significance. This implies that the researcher is allowed some margin of error in his research result. For instance 0.05 level of significance implies that the researcher allows 5% error margin and he is 95% confident of whatever result or conclusion is drawn from his study while for 0.01 level of significance the researcher allows only 1% error margin and he is 99% confident of the result or conclusion drawn from his study.
- Select appropriate statistical techniques. There are many techniques from which a researcher can choose the most appropriate to test his hypothesis such as in the following cases:
- When hypothesis has to do with a large sample (i.e. more than 30) and it is testing for a significant difference between two variables, the z – test implying normal distribution is used and when a sample is small (less than 30), t – test is employed.
- When the hypothesis is started to test for relationship between two variables, the Pearson product moment correlation co-efficient is used, when there is equal distribution of subjects for non-parametric variables, the Spearman rank order correlation coefficient is used.
- When the hypothesis is stated to test differences or relationship between two or more independent variables and one dependents variable, the analysis of variance (ANOVA) or analysis of covariance (ANCOVA) could be employed.
- Multivariate Analysis of Covariance (MANCOVA) is used when there are two or more independent variables against two or more dependent variables.
- Chi-square test is applied to data derived from normal scale of measurement, that is, data obtained in form of frequency counts. It is used for testing hypothesis concerning the difference between sample frequency observed within certain categories and those expected with the categories and those expected with the categories. Chi-square can only indicate whether or not a set of observed frequencies differ significantly from the corresponding set of expected frequencies.
Characteristics of Good Hypotheses
According to Awoniyi et al (2011); Akinade and Owolabi (2009); Cherry (2014), each acceptable hypothesis should have the following features:
- It must provide a specific and adequate answer to a problem that is limited in scope (i.e. give only one answer to a problem at a time).
- It must be based on the research topic.
- It must include independent and dependent variables. It should be framed in a way that the expected relationship between independent and dependent variables are clearly stated.
- Hypotheses should be testable and measurable. One of the most common sources of difficulty for the student who embarks on research project is the selection of hypothesis that is not really testable.
- It should be correctly, clearly and precisely stated.
- It should state expected relationships between research variables. a research hypotheses should be appropriate as a basis for research.
- It should be limited in scope. A common error of the research student in planning research is to develop hypotheses of global significance. Students should seek hypotheses that are relatively simple to test, and yet are highly significant.
- It should be consistent with most facts. Any hypotheses formulated as a basis of research must be consistent with a substantial body of established facts.
- The hypotheses selected should be amendable to testing within a reasonable time. Before any research is undertaken, the student should ask himself practical questions about whether he has the resources and time to undertake the investigation.
- Hypotheses must be testable or verifiable by independent researchers using identical conditions, so as to ascertain its validity.
- Hypotheses should be tested without violating ethical standards.
- Hypotheses should be operationally defined.
Understanding some myths surrounding research hypotheses
Myth: A hypotheses is an educated guess
Everyday use of the word ‘hypotheses’ means an intelligent guess. For science, it can be misunderstood to mean an assumption made before doing an experiment or an idea not yet confirmed by an experiment. A better definition of a hypothesis in science is ‘a tentative explanation for a scientific problem, based on currently accepted scientific understanding and creative thinking’. Research hypotheses are supported by lines of evidence and are based on the prior experience, background knowledge and observations of the scientists.
Myth: Hypotheses become theories that, in turn, become laws
Hypothesis, theory and law are three terms that are often confused. This myth says that facts and observations produce hypotheses, which give rise to theories, which, in turn, produce laws if sufficient evidence is amassed – so laws are theories that have been proved true.
Actually, hypotheses, theories and laws are as unalike as apples, oranges and bananas. They can’t grow into each other. Theories and laws are very different types of knowledge. Laws are generalisations, principles, relationships or patterns in nature that have been established by empirical data. Theories are explanations of those generalisations (also corroborated by empirical data).
Akinade, E.O. &Owolabi, T. (2009).Research Methods: A Pragmatic Approach for
Social Sciences, Behavioural Sciences and Education. Lagos: Connel Publications.
Awoniyi, S.A., Aderanti, R.A. &Tayo, A.S. (2011).Introduction to Research Methods.
Ibadan: Ababa Press.
Cherry, K. (2014).What exactly is a Hypothesis? Retrieved from http://psychology.about.com/od/hindex/g/hypothesis.htm on 22/12/2014.
Glenn, F. (2008). Forming a Good Hypothesis for Scientific Research Retrieved from https://www.verywellmind.com/what-is-a-hypothesis-2795039 on 2019-02-14
Guevera, L. (2010) Myths of the nature of science Retrieved from https://www.sciencelearn.org.nz/resources/415-myths-of-the-nature-of-science on 2019-02-14
Kerlinger, F.N. (1986). Behavioural Research: A Conceptual Approach. New Delhi: Sterling Publisher.