Financial analysts work with noisy data. As a consequence, it is often difficult to determine whether observed results are due to a real effect or reside within the realm of noise. This column reviews the methodology referred to as hypothesis testing, used to distinguish real effects from noise.