[Data organization in statistics applied to biology]

In health statistics data collection may be associated or not to the variable time, that is the moment of data acquisition. The variable may be necessary, if the characteristics of the statistics problem require it. It may be a useful information, particularly in the study of multivariable systems, in order to increase the selectivity of the mutual relations between the variables involved. The main subject is the relation between presumed pathogenous agents and biological damage. In order to identify a mathematical model of a system when variations in the time are present, drawn the distinction between deterministic an statistics methods, an intepretation of significance, validity, and applicability of statististical methods is proposed. For the purpose of showing affinities, the model is compared with other models, that is linear regression and cohort studies. The quantity of data, the frequency and the length in time of the sampling are considered. Characteristics and qualifications of a software for data processing are discussed. The importance of sampling, transmission and organization protocols of the data is also shown.