Journal of Official Statistics, Vol.13, No.4, 1997. pp. 385–399
Estimating Interpolated Percentiles from Grouped Data with Large Samples
Edward L. Korn, Douglas Midthune, and Barry I. Graubard
Abstract:The possible values for the percentiles of a discrete distribution are the same as the possible values of the distribution itself. When data are discrete due to grouping or rounding, there is frequently interest in the percentiles of the underlying continuous distribution. Simple interpolation methods are discussed that smooth the empirical cumulative distribution function to obtain estimates of these underlying percentiles. Some limited simulations are used to examine the properties of these methods. Modifications are discussed for survey data that are obtained with complex sampling designs. Three examples are presented from the second National Health and Nutrition Examination Survey and the 1987 National Health Interview Survey.
Keywords:Histograms; quantiles; smoothing; survey data analysis.
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group