Definition: Bias means that the results of an experiment are not valid because there's something wrong with the way the experiment was conducted. For example, an experiment to see whether flower seeds need water to grow by varying the amount of water they are given - the first one was not watered; the second was given some water and the third received regular watering. The seed that wasn't watered didn't grow at all; the one that got some water did grow (but not by much), and the third flourished (as expected). The result of this experiment suggests that flower seeds need water but the way the experiment was done had a bias because the seed that received the most water was next to the window and therefore got plenty of sunlight; the second was further away and the third was in the shade. To prevent this bias all three seeds needed to be positioned so that they all had access to the amount of sunlight.