Some empirical functions for use in the parametric modelling of stereological size distributions
Laurence E. Scales,
Department of Computer Science, University of Liverpool, Liverpool, England
Vyvyan Howard,
Department of Anatomy, University of Liverpool, Liverpool, England
Abstract
This paper is concerned with some aspects of the unfolding of sphere size distributions by parametric methods, in particular with a comparison of the performance of three types of empirical distribution upon experimental data obtained from neurological tissue. The results suggest that the log normal function is the most useful in this area. A comparison of the estimated errors involved in the parametric and distribution free unfolding of unimodal systems is attempted, and it is concluded that parametric methods can be superior when the fit of the empirical function is good or the sample size is rather small.
Om dit artikel te citeren:
Laurence E. Scales & Vyvyan Howard, «Some empirical functions for use in the parametric modelling of stereological size distributions», Acta Stereologica [En ligne], Volume 2 (1983), Number 1 - Apr. 1983, 187-191 URL : https://popups.uliege.be/0351-580x/index.php?id=4157.