In: Statistics and Probability
Give an example of a situation when you should perform a paired T-test and an example of when you should do an unpaired comparison.
Suppose, in a class, you want to check whether average height of male and female students in a class are same. Here, you will see that a person, whether he is male or female, is not dependent on height of another person. So, here we will do unpaired comparison since, there is no pairing and all the observations are independent.
Now, suppose you are trying to see if the average glucose level in a person's body before and after lunch. Now, for this you need to calculate a person's blood to check glucose level before and after lunch. For this reason, we will select n people and collect their blood before and after lunch to check glucose level. In this case, for a particular person, the observation before and after lunch will be correlated. So, for testing the average glucose level before and after lunch to be same, we have to perform paired t-test.
################################
If you do not get anything in this solution, please put a comment and I will help you out. Do not give a downvote instantly. It is a humble request. If you like my answer, please give an upvote.