Tatjana von Rosen Professor
%20TvR_2026.JPG)
Contact
Name and title: Tatjana von RosenProfessor
Workplace: Department of Statistics Länk till annan webbplats.
Visiting address Room A 4603Albanovägen 12, Hus 4, plan 6
Postal address Statistiska institutionen106 91 Stockholm
About me
I defended my PhD in 2004 at the University of Tartu, Estonia, with the thesis "Permutation Invariance and Reparameterizations in Linear Models". I have been a postdoctoral fellow at the Department of Medical Epidemiology and Biostatistics at Karolinska institutet (2006). Since 2007 I have been working at the Department of Statistics, Stockholm University.
Associate Professor in Statistics since 2008. Professor in Statistics (2026). My research interests comprise univariate and multivariate linear models (estimation, testing and various applications), methods for model validation (detection of influential observations in linear and nonlinear models) and matrix algebra.
I also colalborate with the researches from linguistics, special education, biology and medicine.
Academic degrees
- Ph. D (2004) in Mathematical Statistics, University of Tartu, Estonia.
- M. Sc. (1994) in Biostatistics, Limburg Universitair Centrum, Belgium.
- B. Sc. (1994) in Mathematical Statistics, University of Tartu, Estonia.
My research interests comprise univariate and multivariate linear models (estimation, testing and various applications), methods for model validation (detection of influential observations in linear and nonlinear models) and matrix algebra.
I also colalborate with the researches from linguistics, special education, biology and medicine.
Completed Reserach Projects
As the project leader
2018-2021: Swedish Research Council, 4 400 000 SEK: Development of statistical diagnostic methods and data explorative tools for multivariate multilevel models.
2011-2013: Swedish Research Council, 1 572 000 SEK: Statistical Diagnostics in Multilevel Models.
2011: The Royal Swedish Academy of Sciences, Stiftelse Söderströms donationsfond, 175 000 SEK: Statistical Diagnostics in Linear Models .
2006-2008: Estonian Research Council, 28 000 EUR: Invariant random factors in linear mixed models.
Supervision
Main supervisor
2024: Azadeh Chizarifard, On Estimation and Prediction in Linear Mixed Models: A new approach to studying equal BLUEs and BLUPs.
2015: Karin Stål, Identifying Influential Observations in Nonlinear Regression: a focus on parameter estimates and the score test.
2015: Yuli Liang, Contributions to Estimation and Testing Block Covariance Structures in Multivariate Normal Models.
2014: Chengcheng Hao, Explicit Influence Analysis in Crossover Models.
Supervisor
2021: Ilf Jederlund, Trustful Relationships and School Development, the Department of Special Education.
Topics for future PhD projects
The focus is on statistical modeling of repeated measures data where the response variables are categorical. For example, multivariate binary repeated measures data comprise several binary characteristics (response variables) measured on each individual at several time points.
(i) Modelling of categorical repeated measures data. The development of a hybrid approach to the analysis of categorical repeated measures data where Machine Learning is combined with traditional statistical models e.g. Generalized Linear Models with random effects.
(ii) The inference in Generalized Linear Models with random effects. For example, studying the effects of misspecification of the covariance matrix of random effects and error on estimation and prediction, the model validation, e.g. identification of outliers and influential observations.