AP Computer Science Principles Flashcards: Computing Bias
Written by AP Content Team, Verified for 2026 AP Exams, Last updated: May 2026
Review key ideas with interactive flashcards. This set includes 10 cards to help you master important concepts.
How can a computing innovation reflect existing human biases?
A computing innovation reflects human biases when the algorithms or the data it uses contain and perpetuate pre-existing prejudices or unfair assumptions from its creators or society.
Card 1 of 10
All Flashcards (10)
How can a computing innovation reflect existing human biases?
A computing innovation reflects human biases when the algorithms or the data it uses contain and perpetuate pre-existing prejudices or unfair assumptions from its creators or society.
What is the programmer's responsibility regarding bias in computing?
Programmers should actively take action to identify and reduce bias in the algorithms they create as a way to combat and avoid amplifying existing human biases.
Explain the relationship between biased data and a biased computing innovation.
If the data used to train or operate a computing innovation is biased, the innovation will learn from and replicate that bias, leading to biased results.
An AI tool for screening job applicants consistently favors male candidates. What is a likely cause based on the principles of computing bias?
The likely cause is bias in the data used to train the AI, which may have reflected historical hiring practices where male candidates were favored.
What is 'algorithmic bias'?
Algorithmic bias refers to systematic and repeatable errors in a computer system that create unfair outcomes, stemming from biases written into its code.
At what point in the software development process can bias be introduced?
Biases can be embedded at all levels and stages of software development, from initial design and data collection to algorithm implementation and testing.
Define 'Computing Bias'.
Computing bias occurs when a computing innovation produces outcomes that are unfairly prejudiced or disadvantageous, often by reflecting biases in its algorithm or data.
A developer is building a speech recognition tool. How could they proactively reduce bias in their innovation?
They should train their algorithm on a diverse dataset that includes a wide variety of accents, dialects, and speech patterns to avoid bias against certain groups.
Why is it insufficient to only check the final output of a program for bias?
It is insufficient because biases can be embedded at all levels of software development, meaning the underlying data or algorithms could still be flawed even if one specific output seems fair.
What are two primary sources of bias in computing innovations?
Bias can be introduced through the algorithms themselves (biases written into the code) or through the data used by the innovation (data bias).