This paper distinguishes bias from statistical bias (systematic error)
, and compares two accounts of bias (as an excess of prejudice, and as
a lack of consideration of alternatives) against a list of eight crit
eria. It argues that the latter account is superior.