Sample bias

Any way in which a sample is not representative or any factor decreasing the likelihood of achieving or securing such a sample, a sample is biased if it does not enable us to rule out some alternatives which we ought to consider.

For example, in social science research some college professors conducting research may rely upon student ‘volunteers’ for their studies. Some findings may be quite generalizable to the larger segments of the population, while others may be much more specific to the student population, or even to college students on a particular campus. Student populations generally have a limited age grouping and may be more politically active or radical, or politically more liberal or conservative than the general population. Take the example that one were to study a sample from only one class at one college, say, political science, it is very likely that such a sample would not even be representative of that college as a whole, let alone any wider group of people.