Biometrical Letters Vol. 49(1), 2012, pp. 67-79
|
COMPARISON OF MULTIVARIATE ANALYSIS METHODOLOGIES IN A PALLIATIVE CARE SETTING Isabel Pinto Doria1, Ana Sousa Ferreira2, Otília Dias3, Helena Bacelar-Nicolau4, Georges Le Calvé5 1 Universidade de Lisboa, FP-UL, LEAD and CEAUL Portugal, irpdoria@fp.ul.pt 2 Universidade de Lisboa, FP-UL, LEAD and UNIDE, CEAUL, Portugal, asferreira@fp.ul.pt 3 IPS, ESTBarreiro and CEAUL, Portugal, otilia.dias@estbarreiro.ips.pt 4 Universidade de Lisboa, FP-UL, LEAD and CEAUL, Portugal, hbacelar@fp.ul.pt 5 Société Française de Statistique, g.lecalve@free.fr |
This study is focused in measure the quality and the satisfaction with the palliative cares provided to onchological patients in domicile. The SERVQUAL methodology adapted for the Portuguese context was used to evaluate the quality of palliative care and patient satisfaction. Portuguese SERVQUAL Questionnaire is composed by five Perception scales and two questionnaires, one about the patient and another about the caregiver. Data analysis presented is the analysis of the answers to the five Perception scales, composed by partial ordered variables, evaluating different aspects of quality and satisfaction. Data was analysed comparing metric and symbolic approaches, using Principal Component Analysis Methods and Agglomerative Hierarchical Cluster Analysis Models. Results suggest that symbolic approach provides a more comprehensive analysis for this kind of data.
Agglomerative Hierarchical Cluster Analysis, Multivariate Analysis, Palliative Cares, Principal Component Analysis, Symbolic variables
