What is non response bias?
Who didn't respond? Then we already have an underrepresented trait!
The purpose of a sample is to say something about the population without having to ask everyone in the population.
However, to say something meaningful about the population, the sample needs to be representative of the population. In other words, the sample must be a smaller version of the larger group.
Say you wanted to know something about how the students at your school felt about the food served on campus. You will need to ask some of the students and then infer what the rest of the school thinks.
This sample of students should include the youngest age group to the oldest age group, boys and girls, most of the ethnicities attending, ect. The best way to do this is to take a simple random sample. Everyone has an equal chance of being selected and you select some of them. This usually results in a group that looks like the larger population.
Then you ask everyone in your selected sample to answer your question about the food. But some do not respond or refuse to give an answer!! Now, part of your sample is missing and they all have a common trait. This is bad, because now your sample probably does NOT look like the population and now you CANNOT infer.
The problem is it can be hard to tell how or sample differs from the population. Maybe the older students didn't answer and they all dislike the food. Then the younger students who do like the food are OVERREPRESENTED. But how do you know? You never got a chance to ask the older students.
There are ways to deal with this issue but it is hard to solve
There is your not so short answer. And that makes us happy :)
if this helped let me know in the comments section.