On Geometric and Arithmetic Approaches to Attribution Linking: MATLAB Code

6 Pages Posted: 10 Aug 2012 Last revised: 14 Aug 2012

See all articles by Andrei Reztsov

Andrei Reztsov

University of New South Wales (UNSW)

Date Written: August 9, 2012

Abstract

I am following the recent discussion on a new Arithmetic Linking Algorithm gF (which is equivalent to a Geometric one and was presented in the article “Geometric and Arithmetic Approaches to Attribution Linking are Equivalent” by Andrei Reztsov in the Winter 2011/2012 issue of the Journal of Performance Measurement, Volume 16, Number 2). As an answer to numerous requests from Performance Measurement professionals I developed a simple MATLAB code that provides results for large values of L (i.e. number of months to link or number of days, if we are looking for monthly value through the linking of daily values). As our code is free and suppose to published soon in JPM, every Performance Measurement practitioner (even not a computational specialist or the person with no knowledge of MATLAB) could easily use the code and obtain required results in milliseconds. The executable version of this software is also available upon request; in this case any PC without MATLAB could be used for the calculations. The new Arithmetic Linking Algorithm effectively “patches the existing holes” in other methods. Unlike the standard and modified Frongello Algorithms, the new method is independent of the month order. Compared to Carino and Menchero, this is a linear and non-smoothing construction. The new algorithm is simple, easy to use, and hence more appropriate for practitioners without advanced training in mathematics. It should eventually replace whatever Arithmetic algorithm is used by Performance Measurement specialists (i.e. Frogello. Carino, Menchero).

Keywords: Arithmetic Linking Algorithm, Performance Measurement, Carino, Menchero, Frogello, executable software, MATLAB

Suggested Citation

Reztsov, Andrei, On Geometric and Arithmetic Approaches to Attribution Linking: MATLAB Code (August 9, 2012). Available at SSRN: https://ssrn.com/abstract=2127448 or http://dx.doi.org/10.2139/ssrn.2127448

Andrei Reztsov (Contact Author)

University of New South Wales (UNSW) ( email )

Sydney, NSW 2052
Australia
(+61 2) 95173960 (Phone)

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
475
Abstract Views
2,183
Rank
110,600
PlumX Metrics