Pendant le semestre d’été 2012 : le lundi de 14h00 à 15h00 ou sur rendez-vous
Pendant les vacances de semestre : seulement sur rendez-vous
Cours au semestre d’été 2012 :
Séminaire traitement des données manquantes dans des enquêtes - Seminar Behandlung fehlender Werte in Erhebungen
Cours magistral (CM) de Méthodes de Statistique I (Statistique descriptive) - Vorlesung zu Methoden der Statistik I (Deduktive Statistik)
Cours magistral (CM) de Méthodes de calibration et de pondération - zu Kalibrierungsmethoden und Gewichtung
Cours magistral (CM) de Méthodes d'estimantion de variances - Vorlesung zu Varianzschätzmethoden
Centres d’intérêt de recherche :
Analyse de données avec valeurs manquantes
Méthodes empiriques de Bayes
Nearest Neighbour Algorithme (Algorithme du plus proche voisin)
Publications :
Rässler, S., Koller, F., Mäenpää, C. (2002). A Split Questionnaire Survey Design applied to German Media and Consumer Surveys, Proceedings of the International Conference on Improving Surveys, ICIS 2002, Copenhague.
Koller, F. (2006). Getting the most from auxiliary information in personal interviews, White Paper, Institute for Social Research, University of Michigan.
Koller-Meinfelder, F., Raghunathan, T.E., Rässler, S. (2008). Strategies for Multiply Imputing Missing Integer Values or Rounded Variables, soumis au “Survey Methodology”.
Koller-Meinfelder, F. (2009). Fragebogensplitting – Reduzierung der Antwortbelastung durch Datenergänzung, erschienen in Schnapp, .K.-U., Behnke, N., Behnke, J. "Datenwelten – Datenerhebung und Datenbestände in der Politikwissenschaft", Nomos-Verlag, Baden-Baden, p.152-173.
Koller-Meinfelder, F. (2009). Robust Multiple Imputation of Discrete Data, thèse de doctorat, université de Bamberg.
Conférences :
ICIS (International Conference on Improving Surveys), Copenhague 2002. “A Split Questionnaire Survey Design applied to German Media and Consumer Surveys”.
GfK Media’s Integrated Intelligence Meeting, Vienne 2003. “In-In Data Fusion in Switzerland”.
International Meeting of GfK Statisticians and Methodology Experts (IMSM), Nuremberg 2003. “Evaluation Problems of Data Imputation Techniques”.
International Conference on Logic and Methodology (RC 33), Amsterdam 2004. “Robustifying imputation model estimation”.
International Meeting of GfK Statisticians and Methodology Experts (IMSM), Copenhague 2005.”Correcting for zero-probability measures: A Multiple Imputation experimental design”.
Joint Statistical Meeting (JSM) of the American Statistical Association, Minneapolis 2005. “Sequential Predictive Mean Matching Method of Multiple Imputation” (F. Koller (conférencier), T.E. Raghunathan, S. Rässler, N. Schenker).
International Meeting of GfK Statisticians and Methodology Experts (IMSM), Haarlem 2006.”Causal Inference for TV Trailer Contacts”.
2. Workshop des Medienwissenschaftlichen Lehr- und Forschungszentrums (MLFZ), Cologne 2007. „Datenfusionen als Hilfsmittel der Mediaoptimierung“.
GfK Media Summit, New York 2008. “You fuse, they lose! – Why data fusion has caught on”.
Multiple Imputation for Descrete Survey Data. 2nd Workshop on "Analysis of Incomplete Data" (AID), Bamberg, 02/10/2009.
Meinfelder, F.: Multiple imputation for discrete survey data – the Bayesian Bootstrap Predictive Mean Matching. EAM-SMABS IV European Congress of Methodology, Potsdam, 23/07/2010.
Multiple imputation for discrete survey data-the Bayesian Bootstrap Predictive Mean Matching. Statistische Woche der Deutschen Statistischen Gesellschaft, Nuremberg, 17/09/2010.