Titel:

"Kernel regression in the presence of size-bias"

Abstract:

  We introduce kernel regression estimators of Nadaraya-Watson and
  local linear type for data where the probability of observing a
  particular value depends on its size or the size of another
  observable variable. Special attention are given to the case of
  length-bias. Asymptotic properties are compared with the
  unbiased case and potential applications described.