Publications Software programs ROBCOOP4A




Vassiliades, E., 2002. Program ROBCOOP4A for Estimation of Balanced Classical and Robust Analysis of Variance: Instructions for Use and Source Code. IUGS Commission on Global Geochemical Baselines, Athens, Hellenic Republic, Special Publication, 4, vi, 42 pp., 12 Figures, 4 Tables, 2 Appendices, Supplementary material, ISBN: 978-618-85049-3-6.


The new edition of program ‘ROBCOOP4.EXE’ by Ramsey (1998) is available in two different variants, a 32- and a 64-bit version, namely:

  • “ROBCOOP4A_ΕΝ32.EXE”, and

Apart from the English versions, there is also a Hellenic language version, which is described in a separate section of the accompanying eBook:

  • “ROBCOOP4Α_GR32.EXE”, and

Since, the program source code is given, it can be translated to any other language.


A test data set in two formats is provided: "BGRXRFTS".xlsx" and "BGRXRFTS.dat".


The executable ROBCOOP4.EXE programs, and accompanying files are made available in the following two zip files:

ROBCOOP4A can process concurrently up to 500 variables determined on 2000 samples times 4 analyses, i.e., 2 for the routine and 2 for the field duplicate samples. The quadruple analyses means that the program uses a balanced analysis of variance (ANOVA) design (see diagram on the right hand side). The minimum number of duplicated sites for reliable estimates is eight, which are selected at random (Lyn et al., 2007). It is stressed that this minimum number of duplicated sites concerns small projects.


The updated ROBCOOP4A program estimates, apart from the classical and robust sampling, analytical and geochemical variances, the expanded uncertainty factor (Ramsey et al., 2019). For more explanations refer to Chapter 7 (Demetriades et al., 2022a) in the ‘International Union of Geological Sciences Manual of Standard Methods for Establishing the Global Geochemical Reference Network (Demetriades et al., 2022b).


The new version of ROBCOOP4A provides the user with the option of applying a coverage factor of his/her choice, with the condition that the confidence interval used is stated. This option makes the program more robust and in compliance with the recently published EURACHEM/CITAC Guide on measurement uncertainty (Ramsey et al., 2019).


The cited references can be found in the eBook by Vassiliades (2022).