Whenever we think about a decision-based research we came around the term Hypothesis. And with the term hypothesis, we also get acquainted with p-value, alpha, etc. What do we mean by these?
Imagine two friends are playing roll a dice. The condition is that If rolling the dice gives an odd number one friend will win BDT 500 and if rolling the dice gives an even number another friend will win BDT 500. There is a 50% chance of having an even number on the first roll. Graphically we can present this as below:
What if the dice rolled and gives an even number four times in a row? The probability of that occurring is approximately 6.25% (see above), which is not quite normal. So we may think that the dice is biased.
So we set a limit to what extent we will allow this to happen. We term this as alpha (Level of significance), normally which is taken as 5%.
To decide the fairness we will set the null hypothesis and alternative hypothesis. We will always set the positive scenario as null hypothesis and the opposite of it as an alternative hypothesis in our case null hypothesis can be defined as:
H0: The Dice is fair.
H1: The dice is unfair.
We will reject/accept the Null hypothesis based on P-value. This is defined as the probability of obtaining the observed results of a test, assuming that the null hypothesis is correct; a smaller p-value means that there is stronger evidence in favor of the alternative hypothesis.
We will use a simple rule to determine if we reject the null hypothesis or not. It is known as PGADRN.
If the P-value is Greater than Alpha Do not Reject Null.
So is your null hypothesis getting rejected or accepted?