question: What is the manifestation of bias in machine learning? option 1: Inaccurate assessments that universally inflate or deflate scores independent of group status. option 2: Measurement scores containing systematic error that is not relevant to the construct of interest. option 3: Differences in group means reflecting true differences in the underlying construct. option 4: Systematic influences altering human behaviors or judgments about others based on their group membership. 