noun. the excluding of a fixed percentage of cases at both ends of a distribution prior to computing a statistic on the batch of information in effort to remove the impact of extreme scores on the approximation.
TRIMMING: “Some of the problems may require trimming.”