Statistical modelling
Linear mixed and multilevel models, fixed-effect selection and efficient computational strategies for analysing data with complex dependence.
Assistant Professor · Department of Statistics
Hanna
WdowickaPhD
Statistical modelling.
Probability. Graph algorithms.
Exploring the structure of complex data through statistical methods, mathematical reasoning and computation.
01 /
About
I study how mathematical structure can help us understand complex data and design better computational methods.
I am an assistant professor in the Department of Statistics at Poznań University of Economics and Business. My research interests include statistical modelling, linear mixed models, probability theory and the analysis of graph algorithms.
My work on variable selection focuses on the computational challenges of multilevel models. I also pursue separate research in random trees and graph algorithms, and apply statistical modelling to pedestrian movement.
02 /
Research
Probability & random structures
Probabilistic analysis of growing trees, total path length, martingale limits and the fluctuations of global structural properties.
Graph algorithms & optimisation
Theoretical guarantees for online matching and local search, including the effects of uncertainty and the limits of locally optimal assignments.
03 /
ORCID record Selected work
7 selected works
2026Journal article
Applied Sciences · 16(20), 10010
Adaptive Re-Screening and Pair-Screened Joint Rescue for Scalable Stepwise Selection in Cross-Classified Linear Mixed Models
Adaptive screening strategies that reduce expensive model fitting while retaining predictors that become informative together.
2026Preprint
arXiv:2610.01648
Exact Locality Gaps for Matchable Semi-Matchings
Exact worst-case guarantees for local search in task-to-server assignments, including tight examples on trees.
2026Preprint
arXiv:2610.00614
Total Path Length in Power-Weight Recursive Trees: Martingale Limits and Global Fluctuations
Martingale methods for total path length in weighted recursive trees, covering logarithmic, critical and summable regimes.
2026Preprint
arXiv:2609.40077
Robust and Learned Online Matching in Growing Trees
Online matching under uncertain tree-growth laws, with robustness guarantees and learning from the growing tree itself.
2025Journal article
Mechanical Systems and Signal Processing · 222, 111793
Determinants of pedestrian mediolateral foot placement in walking on laterally-oscillating structures and their consequences for structural stability
Multilevel modelling of pedestrian foot placement and its implications for the lateral stability of structures.
2023Journal article
Przegląd Statystyczny. Statistical Review · 70(4), 73–85Published online in 2024
Computational problems in variable selection for multilevel models using stepwise regression
A computational simplification of forward stepwise multilevel regression, illustrated with employment-related population flows.
2019Journal article
Human Movement Science · 66, 541–553
Topological assessment of gait synchronisation in overground walking groups
An experimental study of gait synchronisation in a nine-person walking group, examining how walking speed and relative position relate to movement coordination.
04 /
Teaching
From statistical concepts to data analysis.
I teach statistics and data analysis with R at Poznań University of Economics and Business and supervise degree dissertations on local labour markets in Poland.
Descriptive statistics
Describing and visualising data, measures of central tendency and variation, and relationships between variables.
Introduction to R
Getting started with R: working with data, performing basic calculations and presenting results.
Mathematical statistics
Parameter estimation, confidence intervals and hypothesis testing as tools for drawing conclusions from sample data.
Statistics
Methods for describing and analysing data, and interpreting results in social and economic research.
Data analysis with R
Preparing and analysing data in R, visualising results and interpreting them in the context of a research question.
Dissertation seminar
Statistical analysis of local labour markets in Poland: formulating research questions, choosing data and methods, and interpreting the results.
05 /
Contact
Let’s connect
through research.
For questions about my work or potential research collaboration, please get in touch by email.
hanna.wdowicka@ue.poznan.plDepartment of Statistics
Poznań University of Economics and Business
Poznań, Poland