BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260512T055328EDT-1855kN32jA@132.216.98.100
DTSTAMP:20260512T095328Z
DESCRIPTION:Leonardo Grilli\, PhD\n\nProfessor - University of Florence\n\n
  \n\n**This talk concerns research financed by the\n	Next Generation EU Pro
 ject Age-It (Ageing Well in an Ageing Society)** \n\nNote: Meet & Greet Pr
 of Grilli from 3-3:30pm in Room 1140\; Prior to seminar 3:30-4:30pm\n\nWHE
 N: Wednesday\, August 27\, 2025\, from 3:30 to 4:30 p.m.\n	WHERE: Hybrid | 
 2001 McGill College Avenue\, Rm 1140\; Zoom\n	NOTE: Leonardo Grilli will be
  presenting in-person\n\nAbstract\n\nLarge-scale assessment data\, such as
  those collected in Italy by Invalsi\, typically include several student b
 ackground variables\, which can be exploited as predictors in modelling st
 udent achievement. Unfortunately\, the student background variables are us
 ually affected by missing values\, posing serious challenges to the model 
 selection procedures. As a further complication\, many of the predictors a
 re variables with unordered categories. This paper proposes combining mult
 iple imputation and variable selection methods in a setting with categoric
 al predictors. In particular\, we implement multiple imputation by chained
  equations (MICE). At the same time\, for variable selection\, we exploit 
 a recently proposed method based on the knockoff filter\, where the knocko
 ff copies are generated using a sequential procedure that properly handles
  both continuous and categorical predictors. A simulation study shows that
  the proposed approach performs well\, also in comparison with other knock
 off-based approaches and the classical lasso. In the application to the In
 valsi test data\, once the student background variables have been selected
 \, we fit a random intercept model to analyse the determinants of the math
  score at grade 5. The proposed approach is computationally feasible and h
 ighly flexible.\n\n\n	Speaker Bio\n\nLeonardo Grilli is a Full Professor of
  Statistics at the University of Florence. He earned a PhD in Applied Stat
 istics from the University of Florence in 2000. He has been the Director o
 f the Master's program in Statistics and Data Science. Currently\, he is a
  member of the board of the PhD Program in Development Economics and Local
  Systems and an elected member of the steering committee of the Italian St
 atistical Society. The teaching activity focuses on introductory statistic
 s and statistical modelling\, including generalized linear models and mult
 ilevel models. The research activity mainly concerns random effects models
  for multilevel analysis\, with methodological advances about the specific
 ation and estimation of models in complex frameworks such as multivariate 
 responses\, informative sampling designs\, and sample selection bias. He a
 lso made contributions in causal inference\, IRT models\, latent growth cu
 rve models\, mixture models\, and quantile regression. The methodological 
 work is driven by applications in the social sciences and medicine.\n
DTSTART:20250827T193000Z
DTEND:20250827T203000Z
SUMMARY:Combining Multiple Imputation and the Knockoff Filter for Variable 
 Selection\, with an Application to Large-Scale Assessment Data
URL:https://www.mcgill.ca/epi-biostat-occh/channels/event/combining-multipl
 e-imputation-and-knockoff-filter-variable-selection-application-large-scal
 e-366247
END:VEVENT
END:VCALENDAR