|Document ID (ISN)||77970|
|Convention or series no.
||INRS NS 0191
||Grzebyk M., Chouanière D., Wild P.
||Statistical models applied to neurobehavioural epidemiology
||Modèles statistiques appliqués à l'épidémiologie neuro-comportementale [in French]|
||Institut national de recherche et de sécurité, 30 rue Olivier-Noyer, 75680 Paris Cedex 14, France, June 2000. 99p. Illus. 48 ref.
||Neurobehavioural epidemiological studies use psychometric testing for generating specific multi-variate data. This report summarizes the results of a study aimed at developing a statistical analysis method suited for multi-variate analysis. Upon reviewing published literature describing appropriate statistical methods, the linear latent variable model was selected. It was first tested and validated on simulated variables, then applied to two sets of real data, one from a cross-sectional study on the neurotoxicity of toluene, and the other on a study aimed at estimating the incidence of insanity among aged persons and identifying the risk factors. Through these different applications of the model, it was possible to develop a better understanding of its applicability and its limitations.
||France; mathematical models; neuropsychic effects; epidemiology
||neurotoxic effects; literature survey; mathematical simulation; evaluation of technique; statistical evaluation; toluene
||E - Books, reports, proceedings
|Country / State or Province||France|
|Broad subject area(s)
||Stress, psychosocial factors
Diseases of the nervous system