A source of non systematic deviation of an observed treatment effect or other outcomes from a true one comes from chance variation in the sample of data collected n the study (example: sampling error) can be reduced but not eliminated unlike biased