A bias is not literally a lie, so it is not data that is faked or made up in any way - instead, it's a real data.
Also, bias can be racial- but it's not always racial, therefore, we can have a "racial bias"
The correct answer is:
systematic error in the way a sample represents a population
A bias changes the results through choosing some part of the population that is not representative. For example, if we choose all the English teachers to represent all teachers, we might find that all the teachers think that English is the most important subject.