N. Gitsakis et al., The use of spreadsheet software programs in performing cluster analysis onagricultural multivariate data, WORK SCIENCES IN SUSTAINABLE AGRICULTURE, 1999, pp. 179-186
Cluster analysis is one of the basic procedures that a researcher must impo
se over its data to ensure possible disallocations among the component vari
ables. When the data include such groups or clusters, it is possible for a
non-regular distribution of variables' values to appear. Under those condit
ions, data cannot be correlated and regression models do not describe possi
ble relations among them in a satisfactory way. During the last years many
statistically oriented programs have become available, offering a variety o
f techniques that handle these kinds of data disorientations.
Commercially available spreadsheet software programs offer convenient and e
ffective means for performing cluster analysis over multivariable experimen
tal data. Among the advantages of the spreadsheet approach to cluster analy
sis are its relative ease of use, its general applicability, a procedural t
ransparency, the ability to alter interactively selected parameters and the
capacity to view the results immediately through embedded graphics.
This paper deals with the use of specific Spreadsheet software and its capa
city to handle biostatistical manipulations over experimental data in order
to produce cluster separation based on several theoretical methods. The pr
esentation rests upon a structure of experimental agricultural data and the
sequence of the procedures is presented along with references to a number
of well known commercial statistically oriented programs in order that its
capacity and effectiveness become clear.