AP® Computer Science A glossary

Bias in data

Also called: algorithmic bias

Course: AP® Computer Science A Shows up in: Unit 4: Data Collections

Bias that comes from how data is collected or used; algorithmic bias is systemic, repeated errors in a program that create unfair outcomes for a specific group.

Appears in the course and exam description (CED): 4.1.B.14.1.B.2

If a data set is collected in a way that over-represents or leaves out some group, conclusions drawn from it can be biased even when every value is accurate. When a program built on such data keeps treating one group of users unfairly, that is algorithmic bias. Programmers should know how data was collected before using it to draw conclusions.

Related terms

• Data set