Agenda

05 Mag 2025 12:00

Directional replicability: When is the factor of 2 necessary?

Aula Delta 2D, Edificio DELTA - Campus Scientifico via Torino

Speaker: Vera Djordjilovic, DEC - Università Ca' Foscari Venezia

Link Zoom: https://unive.zoom.us/j/85153268624?pwd=MzBhdlA2M1B2dThJQ2Y5T0EwUE5PZz09
Meeting ID: 851 5326 8624
Passcode: SanMarco2

Abstract:
Directional replicability addresses the question of whether an effect studied across n independent studies is present with the same direction in at least r of them, for r >= 2 (Bogomolov and Heller, 2023). When the expected direction of the effect is not specified in advance, the state of the art recommends  assessing replicability separately using one-sided p-values for both directions (left and right), and then doubling the smaller of the two resulting p-values to account for multiple testing (Owen, 2009). In this work, we examine whether this multiplicative correction is always necessary, and under what conditions it can be safely omitted.

Bio Sketch:
Vera Djordjilović is an Assistant Professor of Statistics at the Department of Economics of the Ca’ Foscari University of Venice. Born in Belgrade, she moved to Prague after high-school, where she obtained a BSc in General Mathematics from the Faculty of Mathematics and Physics of the Charles University. After completing a Master degree in Mathematical Statistics at the same Faculty, she enrolled in the doctoral program in Statistics of the University of Padova. Under the supervision of Prof. Monica Chiogna, she was awarded a PhD in 2015. In the following years, she worked at the Oslo Center for Biostatistics and Epidemiology and the Department of Biostatistics of the University of Oslo, before moving to Venice in March 2020.

Her main research interests include methodological challenges in the statistical analysis of biomedical and epidemiological studies. In particular, she has investigated  the use of graphical models for modelling and extracting information from complex biological networks. Her research also focuses on various problems of the high-dimensional statistical inference arising, for instance, in genomic applications where the goal is that of identifying potential disease biomarkers among many thousands of candidates.

Lingua

L'evento si terrà in italiano

Organizzatore

Gruppo Statistica (Prosdocimi)

Cerca in agenda