Interactions entre induction et déduction dans le raisonnement humain

by Rui Manuel Da Silva Neves

Doctoral thesis in Psychologie

Under the supervision of Christian George.

defended on 1995

in Paris 8 .

  • Alternative Title

    Interactions between induction and deduction in human reasoning

  • Abstract

    This work is a contribution to the study of the processes of inference implied in causal and dispositional attributions, which underlie a wide spectrum of daily life situations. Our hypothesis is that in order to make these attributions, people extract the regular elements (those which covary and those wich do not covary) in the given cases, and they ascribe a degree of plausibility to the constituted hypotheses given a limited set of contextual informations. An analysis of john stuart mill's system of logic (1843 1988) has provided the basis for the definition of a system of pragmatic schemas of inductive and deductive inferences formalized by several conditional rules. The input of these rules is fully developed by the extraction process that underlies the hypothesis generation and by four types of context ual informations : (i) the exhaustivity of data, (ii) the nature, simple or complex, of the relations between antecedent and consequent elements, (iii) the unicity versus multiplicity of antecedents and consequents, and (iv) the reliability of the datas. The output of the rules is a degree of plausibility attached to the generated hypothesis. In order to test the pragmatic schemas of inference model, four experiments have been conducted to discriminate the effects of the input variables.

  • Abstract

    Ce travail constitue une etude des processus inferentiels sous-jacents aux activites d'attribution de causalite ou de proprietes, activites qui sont impliquees dans un large spectre de situations de la vie de tous les jours. Notre hypothese est que pour produire ces attributions, les individus realisent l'extraction des elements reguliers (cov ariants et coinvariants) dans les cas qui leurs sont donnes, et qu'ils attribuent une plausibilite aux hypotheses ainsi formees etant donne un nombre limite d'informations contextuelles. Une analyse du systeme de logique inductive de john stuart mill (1843 1988) a permis la definition d'un systeme de schemas pragmatiques d'inferences inductives et dedu ctives qui ont la forme de regles conditionnelles avec en entree la nature du processus d'extraction des regularites et quatre types d'informations contextuelles : (i) la completude des donnees, (ii) le caractere simple ou complexe des relations entre antecedents et consequents, (iii) l'unicite versus la pluralite des antecedents et des consequents et (iv) la fiabilite des donnees. La sortie des schemas consiste en l'attribution d'un degre de plausibilite a l'hypothe se reguliere consideree. Afin de verifier le modele de schemas pragmatiques, quatre experiences ont ete realisees qui testent l'aptitude des sujets a discriminer les effets des variables en entree des schemas pragmatiques.

Consult library

Version is available as a paper


  • Details : 343 f
  • Notes : Publication autorisée par le jury
  • Annexes : Bibliogr.: f. 325-343

Where is this thesis?

  • Library : Université Toulouse Jean-Jaurès. Centre de ressources de psychologie Louise Michel.
  • Unvailable for PEB
  • Odds : B10.640
  • Library : Université Paris 8-Vincennes Saint-Denis (Sciences humaines et sociales-Arts-Lettres-Droit). Service Commun de la Documentation. (Saint-Denis) .
  • Available for PEB
  • Library : Université Paris 8-Vincennes Saint-Denis (Sciences humaines et sociales-Arts-Lettres-Droit). Service Commun de la Documentation. (Saint-Denis) .
  • Available for PEB
  • Odds : -
See the Sudoc catalog libraries of higher education and research.