Contents
Meaning of Hypothesis
Hypothesis refer to an assumption made for investigation on basis of some evidence. It is inference that is drawn from a sample of data for the entire group or population. Hypothesis comprises of distinct components such as variables, relation in between the variables and population. It is the starting phase of every investigation which facilitate the making of observations and experiments. This assist in verification of observations and carrying out enquiries into right direction. Hypothesis are made from a variety of sources like past studies observation, present day experience, competitors, resemblance in between phenomenon, scientific theories and general patterns influencing people’s thinking process.
Characteristics of Hypothesis
Various characteristics of Hypothesis are as follows: –
- A Hypothesis need to be clear and precise for considering it a reliable one.
- It need to be specific and must have scope for carrying out more tests.
- Presence of evidence is must for construction of hypothesis in carrying out an investigation.
- In case of relational hypothesis, it must state the relationships in between the variables.
- Hypothesis must explain everything in simple and easily understandable manner. Its simplicity should be related to its significance.
Procedure of Hypothesis Testing
There are various steps involves in testing of hypothesis which need to be followed in a systematic manner. All these steps are discussed in detail as given below: –
Setting up Hypothesis
Firstly, the hypothesis to be tested is established. For unknown parameter, an assumption is made by setting statistical hypothesis. This hypothesis provides few numerical values or values range for parameter. Null hypothesis and Alternative hypothesis are two hypothesis constructed for the population.
H0 is used for denoting Null hypothesis which asserts that there is no true difference in between the actual and assumed value of parameter. According to this hypothesis, difference if any is accidental that occurs due to sampling fluctuations.
Whereas, H1 is used for denoting the Alternative hypothesis about population. This hypothesis stands true and is accepted if the null hypothesis (H0) is rejected.
Setting up suitable significant level
Next step after constructing hypothesis for population, researcher need to decide the significance level i.e. confidence level with which null hypothesis to be accepted or rejected. An ‘α’is used for denoting the significance level that is generally defined prior to drawing samples in a way that results obtained do not affect the choice. Generally, 1% or 5% are significance level which are taken.
5% significant level denotes that there are 5 chances out of 100 of rejecting the null hypothesis when it should have been accepted. It simply says that there is 95% confidence of making a right decision.
Suitable test statistic is determined
Once the construction of hypothesis and deciding of significance level is done, now the suitable test statistic and its distribution is determined. Generally, the following form is assumed by statistic tests: –
Test Statistic = Sample Statistic – Hypothesized Parameter
Standard Error of the Statistic
Determining critical Region
Critical region refers to the value that leads to rejection of H0 (Null Hypothesis). Values to the test statistic that will cause acceptance of H0 should be decided prior to drawing the samples.
Performing Computations
After identifying the critical region, different values for random sample of ‘n’ size is computed. Now, the test statistic formula denoted in Step.3 will be applied for verifying whether the results of samples falls in acceptance region or the rejection region.
Decision-making
Now, the statistical conclusions are drawn after all of the steps are performed. Management can take decisions now with respect to acceptance or rejection of null hypothesis (H0).
All these steps need to be followed in a systematic manner for testing the Hypothesis. It will ensure that results obtained are accurate and did not suffer from any of the statistical error.
