Home Last parution Authors Topics Keywords Parutions Videos Search Reading committee Contact Subscription Article submission   French version The issues from 1 to 148 are also available on line at NUMDAM and at Revues.org for the following issues

Results for criterions: Keyword: Statistical inference Modify search criterions 3 matches Title Statistical inference, exchangeability and multilevel approach Author COURGEAU Daniel Keywords Epistemological probability, Exchangeability, Frequence-type probability, Multilevel approach, Simpson paradox, Social sciences, Statistical inference Topics Epistemology, Mathematical Statistics, Modelling, Social Sciences, Statistics Abstract This paper is concerned with the problems of statistical inference in social sciences. In order to pass from a population to a new individual by inference, de Finetti's idea of exchangeability is useful. This is shown in a simple logit model and in a multilevel model, with a larger number of observed groups, where the Simpson's paradox arises depending on whether you work on aggregate data or on group data. The concept of exchangeability permits, by using the appropriate probabilities, to solve a number of problems arising in statistical inference. But it is necessary to use not only the data but also information inferred by other means in order to make the final inference. Number 179, Fall 2007 Language   French Read the article Title The use of significance tests by psychologists, from normative, descriptive and prescriptive viewpoints Author POITEVINEAU Jacques Keywords Bayesian methods, Confidence intervals, Inductive reasoning, Methodology, Significance test, Statistical inference Topics Mathematical Statistics, Methodology, Probabilities, Psychology, Statistics Abstract At a normative level, the significance tests appear to be ill-suited and the main criticisms are reported. At a descriptive level, both examination of statistical textbooks, re-analyses of published papers and experiments about the use of significance tests by psychologists clearly reveals many misuses. At a prescriptive level, alternative solutions are considered, especially the Bayesian methods which appear to be especially attractive. Number 167, Fall 2004 Language   French Read the article Title Probability, demography and social sciences Author COURGEAU Daniel Keywords Demography, Epistemological probability, Frequence-type probability, Social sciences, Statistical inference Topics Demography, Epistemology, Mathematical Statistics, Modelling, Probabilities, Social Sciences Abstract This paper considers the different approaches to probability, that has been developed from the XVIIth century, compared to the simultaneous development of social sciences. If the frequency-type approach, for which the probability is the mathematical limit to which the relative frequency of an outcome converges, the epistemological approach, for which the probability concerns the available evidence it is reasonable to have a degree of belief in the truth of a given proposition, seems to be more suitable for social sciences. A presentation and a discussion of these two approaches are made here, with numerous examples of application to social sciences. Number 167, Fall 2004 Language   French Read the article Users rights : The entire journal is licensed under a Creative Commons license