Described statistical were made of the parameters (variable) evaluated and its
correspondent variations during the accomplishment of this study.
Because these variations depend on the initial values of other characteristics called
co-variable, the meaning of the statistical determination needs to be done with a base of
pattern values. The co-variable used was: age, weight, and index of corporeal mass,
besides the initial values of the pad test. The meaning of the statistical determination
was possible through a study using initial values of the co-variables corresponding to the
average of the population of this sample (model patient), meaning patients doing the pad
test of 21g.
To obtain the wanted values of the variations of the pad test in each of the periods that
are in the study, we used the Co-variance of Analysis technique (ANCOVA) for repeated
measures ^{8}. This technique allows us to make comparisons controlling the
differences due to the co-variables that might have an association with the answer. We
used the maximum verisimilitude method to adjust the models and the significance level of
5% for the selection of the co-variables. |