Research hypothesis | Null and Alternative Hypothesis
Definition of hypothesis
Hypothesis is a tentative gues, assumed for use in devising theory or planning experiment intended to be given a direct experimental test when possibleEric rogers, 1966
Basically, a hypothesis is a statement that is suggested by knowledge or observation but has not, yet, been proved or disproved. .
Enough complicating, In simple small words
- An educated guess
- A tentative point of view
- A proposition not yet tested
- A preliminary explanation
- A preliminary postulate
Why hypothesis is used?
The hypothesis is a clear statement of what is intended to be investigated.
It should be specified before research is conducted and openly stated in reporting the results.
This allows to:
- Ø identify the research objectives
- Ø identify the key abstract concepts involved in the research
- Ø identify its relationship to both the problem statement and the literature review
A problem cannot be scientifically solved unless it is reduced to a hypothesis form.
For proposing hypothesis
- You must make a prediction/ statement/ postulation
- You must identify at least two variables
Qualities of an ideal research hypothesis
- It should have an elucidating power
- It should strive to have an acceptable explanation or accounting of a fact
- It must be falsifiable simply: it must be capable of being refuted based on the results of the study.
- It must be formulated in simple, understandable terms.
- It should correspond with existing knowledge.
- In general, a hypothesis needs to be unambiguous, specific, quantifiable, testable, and generalizable.
Characteristics of hypothesis
A hypothesis must be conceptually clear
- Concepts should be clearly defined
- The definitions should be commonly accepted
- The definitions should be easily communicable
A hypothesis should have empirical reference
- Variables in the hypothesis should be empirical realities
- If they are not, it is not possible to make the observation and ultimately the test
A hypothesis must be specific
A hypothesis must be specific to place, situation, and operation.
A hypothesis should be related to available techniques of research
- Either the techniques are already available or
- The researcher should be in a position to develop suitable techniques
A hypothesis should be related to a body of theory
- A hypothesis has to be supported by theoretical argumentation
- It should depend on the existing body of knowledge
Types of hypothesis
- Based on their formulation
- Null hypotheses
- alternate hypotheses
- Based on direction
- Non-directional hypothesis
- Based on their derivation
- Deductive hypotheses
But for all practical purposes, we all need to know about the first one
Null hypotheses (h0) and alternate hypotheses (h1)
What does null hypothesis (h0) predict?
h0 always predicts that there is no differences between the groups being studied
Null hypothesis is boring, its 'nothing changed after your study'
Let us take an example of two groups, drinking (experimental) vs. non-drinking (control).
The null hypothesis will be no difference between these two groups when the incidence of driving accidents is studied.
No relationship between the variables being studied i.e. drinking and driving don't result in accidents.
What does the alternate hypothesis predicts?
An alternate hypothesis always predicts that there will be a difference in the relationship between the variables being studied
In fact in medical research, frequently “alternative” is actually the desired conclusion of the researcher!
All we want is to establish that the alternative hypothesis is true and that can be done by rejecting the null hypothesis.
We want to reject that there is no association between drinking and driving accidents and hence we want to prove that there is in fact a relationship between drinking and driving.
So the alternate hypothesis would be, that drinking increases driving accidents.
Now to prove that the alternate hypothesis is true, we conduct the study in two groups.
How to formulate a research study?
- Formulate the h0 and h1
- Collect data that is the observation
- Apply statistical tests
- Prove or disprove the null hypothesis and state the conclusion
Remember we don't do it, the other way. Meaning we don't reject or accept the alternative hypothesis.
How strong is the relationship between variables, depends on statistical parameters like p-value.
Drug trial on a new drug
- H0: there is no difference between the two drugs.
- H1: the new drug is better than the current drug.
Ho relates to the statement being tested
H1 relates to the statement to be accepted if the null is rejected
How are we going to conclude?
Accept h0 = no difference between old and new drug
Reject h1 = difference between old and new exist and is a fact
Inductive and deductive hypotheses
Based on how it's derived, the hypothesis can also be classified as an inductive and deductive hypothesis.
An inductive hypothesis is used for scientific experiments. Data is collected through experiments, observations, or surveys. All the data is then analyzed for derivations.
Inductive hypothesis is a generalization based on observation
A deductive hypothesis is based on pre-existing theory to start with
Forming the hypothesis is the first important step in the planning the research study
Pre-requisit for formulating the hypothesis
- The problem/point of proving (or disproving) has been identified and
- The literature study has been conducted for the same (the researcher is totally aware of the theoretical and empirical background of the concerned topic)
The initial idea is just a starting point, often vague or general, and requires refining before research hypotheses can be generated.
Refinement of the initial idea is based on
- A search of relevant research literature
- The initial observations made regarding the topic of study
Then the initial idea is narrowed into a statement of the problem
In simple words, ideas lead to arrow
Statement of problem
Problem statements will then become research hypotheses
Steps in formulating hypothesis/refining hypothesis
Let's start making an hypothesis with variables of breastfeeding and pneumonia
Identify an area of interest to be researched
Exclusive breastfeeding and pneumonia
Identify a research question – make it narrow and specific
Does exclusive breastfeeding reduces the occurrence of pneumonia ?”
Set a definition for the variables and make them specific
Say, for example, Pneumonia in infancy, community-acquired pneumonia
Understand your control group
Here, nonexclusive breastfed babies
Formulate the hypothesis by making it a conditional statement
Exclusively breastfeeding reduces the occurrence of community-acquired pneumonia in infancy.
Refine it by writing a formal hypothesis
If exclusive breastfeeding reduces the occurrence of community-acquired pneumonia then exclusively breastfed babies have a lesser chance to develop pneumonia in infancy
This type of "If-then" hypothesis is considered the most useful.
Verify that a treatment or exposure is included in the experiment
Here exclusive breastfeeding is the exposure.
Prepare for an outcome measure
Rate of exclusively breastfed babies, developing pneumonia
Hypothesis can be adjusted/refined/changed as more information is gathered but before the actual experiment is carried out.
The concept of null and alternate hypothesis is very important for beginers like us, lack of understanding and discussion on statistics during our pediatric training make these topics further boring and difficult for us. COnsider thesis as our mini research learning tool and begin, here are few more things for you!
Jijo Jose | DCH, DNB (Pediatrics)
Jijo completed his pediatric residency training from Aster Medicity, Kochi, Kerala and works as a pediatrician