Split-plot designs: a useful tool for the statistical analysis of real medical data and repeated measurement problems


Koukouvinos, C., Lygkoni, E. and Mylona, K. (2009) Split-plot designs: a useful tool for the statistical analysis of real medical data and repeated measurement problems. Bulletin of Statistics and Economics, 3, (S09), Spring Issue, 61-66.

Download

Full text not available from this repository.

Description/Abstract

In this article, a split-plot design is applied for the collection of real medical data concerning biochemical and vital measurements of 283 patients in Intensive Care Unit, on a repeated measurement problem. Practical limitations and issues related to cost reveals frequently the practical usefulness of split-plot designs. When an experiment is planned, factors that are difficult or time consuming to manipulate appear and make complete randomization impractical. A split-plot structure is used when it is impractical to change the levels of some of the experimental factors. Our dataset, selected with the aid of a split-plot design, consists of 2256 observations 17 factors and 8 repetitions. A statistical analysis on the resulting data is performed and useful biological justification of our results is provided.

Item Type: Article
ISSNs: 0973-7022
Related URLs:
Subjects: H Social Sciences > HA Statistics
Q Science > QA Mathematics
R Medicine > RA Public aspects of medicine
Divisions: Faculty of Social and Human Sciences > Mathematical Sciences > Statistics
ePrint ID: 336716
Date Deposited: 04 Apr 2012 14:19
Last Modified: 27 Mar 2014 20:20
URI: http://eprints.soton.ac.uk/id/eprint/336716

Actions (login required)

View Item View Item