Papers by Brigitte Le Roux
Revue française de science politique, 2000
L'espace politique des électeurs français à la fin des années 1990. Nouveaux et anciens clivages,... more L'espace politique des électeurs français à la fin des années 1990. Nouveaux et anciens clivages, hétérogénéité des électoratsé In: Revue française de science politique, 50e année, n°3, 2000. pp. 463-488.
Multiple Correspondence Analysis
ABSTRACT In this chapter, I present the study of Euclidean clouds, taking into account the struct... more ABSTRACT In this chapter, I present the study of Euclidean clouds, taking into account the structures they are equipped with. The motivation is as follows: In ``Analyse des Données", the basic object of methods is a two-way table. Data Analysis brings out the relations between the two sets, that is, the structures of the table; such methods have the remarkable ability to identify these structures —this is one of their most valuable assets— but they do not take into account the structures the basic sets themselves may be equipped with. By structuring factors, we mean relevant variables describing the two basic sets that do not serve to construct the Euclidean clouds. By structured data, we designate data tables whose basic sets are equipped with structuring factors; and by Structured Data Analysis, we mean the embedding of structuring factors into Data Analysis, in the line of Analysis of Variance --- including its multivariate extension (MANOVA}) --- while preserving the principles of the construction of clouds (Le Roux \& Rouanet, 1984). Clearly, structured data constitute the rule rather than the exception. The Data Analysis approach, including structured data analysis and its inductive extensions is called Geometric Data Analysis (The phrase “geometric data analysis” was suggested to us in 1996 by Patrick Suppes (Stanford University).).
HAL (Le Centre pour la Communication Scientifique Directe), 2006
International audienc
Empirical Investigations of Social Space, 2019
More than 15 years after his passing, Pierre Bourdieu continues to be one of the most important f... more More than 15 years after his passing, Pierre Bourdieu continues to be one of the most important figures in the field of social sciences. His oeuvre, his analytical concepts, and his methods of conducting empirical research are still gaining in relevance. Much of this success, both within and outside the field of social sciences, can be traced back to the unique way Bourdieu combined (or rather: related) traditionally distinct scientific cultures and tools (such as spatial concepts and geometric data analysis, qualitative and quantitative techniques, as well as empirical and critical research, etc.).
Multiple Correspondence Analysis
The purpose of this study is to reveal the similarities and differences between the universities ... more The purpose of this study is to reveal the similarities and differences between the universities in Turkey with regard to the numbers of academicians and departments. Within the scope of this objective, we included the first 100 universities among the total 185 universities in Turkey according the URAP data by the criterion sampling method. Conformity analysis, one of the multivariate statistical methods, was used in the data analysis. A two-dimensional graphic is obtained including the points representing the variable categories in the multi-dimensional space by using mathematical models to determine the structures of crosstabs consisting of the rows and columns through multiple conformity analysis. As a result of the study, it was determined that there was a conformity between the Ankara, Odtü, Marmara, Gazi, Hacettepe, İstanbul, Atatürk and İnönü Universities. In addition, a conformity was determined between the Ankara, İstanbul, İstanbul Teknik, Ege and Başkent universities according to the professor numbers within the academician categories. It was determined that there was no conformity between the Özyeğin, Bilkent, Kadir Has, Sabancı, Bahçeşehir, Koç and İstanbul Ticaret Universities regarding the variables addressed within the study. Based on the study, it is suggested that similar studies of multiple conformity analysis should be repeated with the different data groups.
Empirical Investigations of Social Space, 2019
Combinatorial Inference in Geometric Data Analysis, 2019
Combinatorial Inference in Geometric Data Analysis, 2019
Combinatorial Inference in Geometric Data Analysis, 2019
Multiple Correspondence Analysis
Combinatorial Inference in Geometric Data Analysis, 2019
Journal of Statistical Computation and Simulation, 2019
This mathematical book is all about performing geometric data analysis without assuming underlyin... more This mathematical book is all about performing geometric data analysis without assuming underlying distributions for the data, based on permutation procedures in a combinatorial framework. The readers need to have calculus knowledge and familiarity with mathematical statistics. The unique features of this book include multidimensional procedures, and R computing language. There are seven chapters covering a range of topics motivation for combinatorial inference, geometric versus inductive approaches, cloud of points in a geometric space, combinatorial typicality tests, coheris SPAD software, homogeneity permutation tests, repeated measures design, case studies about Parkinson illness, French Parliament versus globalization in 2006, cognitive versus education, matrix calculus, vector space, multidimensional geometry, and orthogonal projection among others. The bibliography is done thoroughly, and it is up-to-date. The list of symbols and abbreviations is helpful in the readability of this book. I enjoyed reading this book. I recommend this book to the statistics and computing professionals.
Monde des Util. Anal. Données, 2010
Depuis une dizaine d'annees, la taille des donnees croit plus vite que la puissance des proce... more Depuis une dizaine d'annees, la taille des donnees croit plus vite que la puissance des processeurs. Lorsque les donnees disponibles sont pratiquement infinies, c'est le temps de calcul qui limite les possibilites de l'apprentissage statistique. Ce document montre que ce changement d'echelle nous conduit vers un compromis qualitativement different dont les consequences ne sont pas evidentes. En particulier, bien que la descente de gradient stochastique soit un algorithme d'optimisation mediocre, on montrera, en theorie et en pratique, que sa performance est excellente pour l'apprentissage statistique a grande echelle.
We present Geometric Data Analysis applied to a barometer data survey in order to construct the F... more We present Geometric Data Analysis applied to a barometer data survey in order to construct the French “trust space”, define groups of French citizens as regards trust and study the evolution of trust over time. The data source stems from the survey entitled “Le barometre de la confiance politique” (“Barometer of political trust”) carried out by the CEVIPOF (Center for political studies in Science Po, Paris). The data have been collected since 2009 on a yearly basis. The samples (about 1,500 persons) are designed to be representative of the French citizens registered to vote. A set of 22 questions that are relevant to 5 themes (trust in political and non-political institutions, economic institutions, interpersonal and personal trust) is selected for this study. The questions are bipolar (19 are coded on a Likert scale with 4 or 5-point and 3 are dichotomous). For constructing the space, we perform correspondence analysis (CA) after doubling the data and we weight the questions in or...
Uploads
Papers by Brigitte Le Roux