Posted On :

Testing a hypothesis refers to verifying whether the hypothesis is valid or not.

Testing a hypothesis refers to verifying whether the hypothesis is valid or not. Hypothesis testing attempts to check whether to accept or not to accept the null hypothesis. The procedure of hypothesis testing includes all the steps that a researcher undertakes for making a choice between the two alternative actions of rejecting or accepting a null hypothesis. The various steps involved in hypothesis testing are as follows:

This step involves making a formal statement of the null hypothesis (H

The hypotheses should be tested on a pre-determined level of significance, which should be specified. Usually, either 5% level or 1% level is considered for the purpose. The factors that determine the levels of significance are: (a) the magnitude of difference between the sample means; (b) the sample size: (c) the variability of measurements within samples; and (d) whether the hypothesis is directional or non-directional (Kothari, 1988). In sum, the level of significance should be sufficient in the context of the nature and purpose of enquiry.

After making decision on the
level of significance for hypothesis testing, the researcher has to next
determine the appropriate sampling distribution. The choice to be made
generally relates to normal distribution and the t-distribution. The rules
governing the selection of the correct distribution are similar to the ones
already discussed with respect to estimation.

Another step involved in hypothesis testing is the selection of a random sample and then computing a suitable value from the sample data relating to test statistic by using the appropriate distribution. In other words, it involves drawing a sample for furnishing empirical data.

Another step involved consists of
making a comparison of the probability calculated with the specified value of
α, i.e. The significance level. If the calculated probability works out to be
equal to or smaller than the α value in case of one-tailed test, then the null
hypothesis is to be rejected. On the other hand, if the calculated probability
is greater, then the null hypothesis is to be accepted. In case the null
hypothesis H_{0} is rejected, the researcher runs the risk of committing the Type I
error. But, if the null hypothesis H_{0} is accepted, then it involves some risk (which cannot be specified in
size as long as H_{0} is vague and not specific) of committing the Type II error.

Tags : Research Methodology - Introduction

Last 30 days 1238 views