RESUMO
In order to develop a genetic study of human laterality, we conducted an exploratory study concerning one aspect of this phenotype: lattice analysis was used to determine whether the structure of manual preference was the same for right- and left-handers. The study highlights the links between two sets - participants and actions - describing binary data, by ordering them "dually" along a "Galois lattice": participants were ordered according to subsets of actions for which they used only their writing hand, while actions were ordered according to sub-groups using their writing hand to perform them. The twelve item questionnaire of Annett was analysed in two samples of 94 adult right-hand and 31 left-hand writers. The items did not have the same categorical impact for the two groups of left- and right-hand writers. The behaviour of right-handers appeared globally more stereotyped. On the contrary, left-handed profiles were nearly all distinct. To explore these conclusions more thoroughly in the general population would certainly require greater samples. Nevertheless in both cases the observed structures were highly dimensional, a result that would grow stronger as the group sizes increase. Hence whereas some questionnaires purport to evaluate laterality along an unidimensional continuum, the present analysis questions such a strong assumption providing evidence to the contrary.
