Please discuss the purpose of hypothesis testing. provide an example of a null h
ID: 388617 • Letter: P
Question
Please discuss the purpose of hypothesis testing. provide an example of a null hypothesis and alternative hypothesis. Why is hypothesis testing important for researchers?Please discuss the purpose of hypothesis testing. provide an example of a null hypothesis and alternative hypothesis. Why is hypothesis testing important for researchers?
Please discuss the purpose of hypothesis testing. provide an example of a null hypothesis and alternative hypothesis. Why is hypothesis testing important for researchers?
Explanation / Answer
Please discuss the purpose of hypothesis testing. provide an example of a null hypothesis and alternative hypothesis. Why is hypothesis testing important for researchers?
We conduct research to get solutions of problems that an organization faces. The problems are translated into researchable questions. A researcher based on exploration of various secondary sources, interaction with experts, discussing with stakeholhers etc. comes out with possible reasons of the problem. Among the competing reasons one or more could sound more plausible and relevant than others. Now researcher needs to establish that the most plausible reason identified is actually significant or not in context of the problem on hand. Hence he makes an assumption that the problem on hand is an outcome of the reason identified and if solved could end the problem. In other words he makes an assumption about the relation between the problem and its reason. The assumption about their relationship could be cause and effect type, association type, dependence type etc. The assumption is expressed in for of a statement that could be statistically tested. This statement of assumption is Hypothesis. Sine Hypothesis testing in essentially a quantitative technique, both problem and reason needs to be transformed into variable form so that data could be collected from the relevant population and analysis and testing could be done.
The statement of assumption when tested could either be validated (if found true based on statistical analysis) else will be rejected. If the statement made could not be validated then opposite will be true. It means there is actually two statements volte-face to each other expressing mutually exclusive situations (i.e. if one is true then other will not be or any one will be true at a time.) The first statement of assumption about the relationship is Null Hypothesis and other is Alternate Hypothesis. The first statement is called Null because it assumes that NO relation exists ( Statistically it is possible to establish NO relation, if one fails to establish this then automatically other condition prevails.).
In short alternative hypothesis is what we will believe is true if we reject the null hypothesis.
It is important to note that hypothesis testing gives only probability about the statement of assumption being true or acceptable and involves a degree of error; it is up to researcher to define acceptable limits of error and probability
Examples:
Null Hypothesis (Ho): There is no relation between advertising and sales.
Alternate Hypothesis (H1): There exists some relation between advertising and sales.
OR
Ho: Employee training leads to improvement on productivity.
H1: Employee training does not improves productivity.
Hypothesis testing is important because it allows us to reach a conclusion about the problem and its causes/reason with quantitative support. By seeing a data set one could infer something but that could actually differ from reality (due to sampling error), so the inference we draw needs to be supported by evidence (of course with probability), and hypothesis testing just gives that extra evidence before making a inference or conclusion.
For example if we take test of 100 marks in two sections of a class of sixty each. We randomly selected marks of 6 students from each class and averaged. The average marks for one section is 67 and other 69. To decide whether first section is actually poor on performance than other or difference of two marks is negligible and occurred due to sampling error. Hypothesis testing with defined level of acceptable error could help you to decide.