The definition of confirmation bias is the tendency of people to overvalue data that supports their own beliefs.
Nobody likes to be wrong. The typical pattern is to form a theory based on insufficient data (e.g., that students do or don’t understand, that technology is good or bad, that your principal picks on you), and then restrict further analysis to identifying data that fits that theory.
Also factoring in is the challenge of cognitive dissonance–holding two competing beliefs at once. This is uncomfortable, so we cling to one and pitch the other, then scour data to prove that we are right. This can result in a kind of self-fulfilling prophecy, where you essentially will your biases and insecurities into existence.
Which is as crappy as it sounds.
The following video is a great example of confirmation bias. Is this something you’re guilty of as a teacher? How about your students–do they collect data and form theories, or the other way around?