CARVALHO, P.R.MUNITA, C.S.LAPOLLI, A.L.2019-08-092019-08-092019CARVALHO, P.R.; MUNITA, C.S.; LAPOLLI, A.L. Validity studies among hierarchical methods of cluster analysis using cophenetic correlation coefficient. <b>Brazilian Journal of Radiation Sciences</b>, v. 7, n. 2A, p. 1-14, 2019. DOI: <a href="https://dx.doi.org/10.15392/bjrs.v7i2A.668">10.15392/bjrs.v7i2A.668</a>. Disponível em: http://repositorio.ipen.br/handle/123456789/30052.2319-0612http://repositorio.ipen.br/handle/123456789/30052The literature presents many methods to produce data set clusters and the better method choice becomes hardest because the various combinations between them based on different dissimilarity measures can lead to different cluster patterns and false interpretations. Nevertheless, little effort has been expended in evaluating these methods empirically using an archeological data set. In this way, this work has the objective to develop a comparative study of the cluster analysis methods and to identify what is the most appropriate for an archeological data set. For this, 45 ceramic fragments samples data set was analyzed by instrumental neutron activation analysis (INAA). And, five hierarchical methods of cluster were used to this data set: Single linkage, Complete linkage, Average linkage, Centroid and Ward. The validation was done calculating cophenetic correlation coefficient values by a statistical program R and the comparison between them showed the average linkage method was more accurate for the 45 ceramic fragments samples data set. With this, the statistical program R showed be an tool option for other scientists to calculate their cophenetic correlation coefficient and to identify the more accurate methods for their archeological data set.1-14openAccessvalidationneutron activation analysisarchaeologyquality controlcomparative evaluationsceramicsmultivariate analysiscluster analysiselementsArtigo de periódico2A710.15392/bjrs.v7i2A.6680000-0003-0546-1044Sem Percentil