Ceiling Effect

The ceiling effect is a measurement problem in which an independent variable stops producing any further change in a dependent variable once that variable has reached the upper limit of the scale used to measure it, so that scores cluster near the maximum and the true differences between high-scoring cases become invisible. It is one of two scale attenuation effects, paired with the floor effect at the opposite end of a scale, and appears both in pharmacology, where an analgesic drug stops relieving more pain past a certain dose, and in survey data collection, where an income question capped at a top category such as $100,000 per year or more makes it impossible to distinguish a respondent earning exactly that amount from one earning far more. A test whose items are too easy for the people taking it shows the same pattern: most test takers cluster at or near the highest possible score, which compresses the resulting scores' variance and undermines any statistical analysis built on the assumption of a normal, spread out distribution.