systematic errors in measurement, or infrences made from measurements, that are related to different identifiable group membership chracteristics such as age, sex or race.
ex) if clock had added 5 min only to arrival times of female employees, the clock is biased in measuring arrival times of female employees.
predictive bias occurs when errors in prediction are made for members of a subgroup.
the principle that every test taker should be assesseed in an equitable manner.
fairness in measurement refers to the value judgements people make about the decisions based on measurements.