Chasing the ghost: When data gets noisy, scientists find creative ways to clean it up

Telling the difference between good and bad information is at the heart of the struggle for knowledge in science. While looking for one thing, any number of stray signals can mislead scientists into thinking they've found something important when they haven't. And sometimes scientists nearly miss making an important discovery because they threw out useful information, thinking it was mere noise. Scientists can be left chasing ghosts---following noise instead of signal---if they have not carefully filtered out the noise from the information they gather.;Separating noise from signal is also a survival skill that people use daily. Every time you cross a street or open a door your brain is judging which parts of the incoming information are important and which parts can safely be ignored. And human brains seem to ignore the unimportant signals almost effortlessly, focusing on the edge of a curb without dwelling on the texture of the sidewalk, or picking out a familiar face in a crowd.;This piece explores some of the fundamental problems scientists face when trying to identify noise and remove it from their experiments. It also examines some sources of noise inside people's heads: the less-than-ideal communication between neurons in the brain, and the ways in which the brain overcomes this noise to give humans the ability to distinguish between useful information and spurious signals.