How to contact us

Department of Statistics

Columbia House

London School of Economics

Houghton Street

London

WC2A 2AE

 

MPhil/PhD and Research Queries

+44 (0) 20 7955 7511

Email: i.marshall@lse.ac.uk|

 

BSc Queries

+44 (0) 20 7955 7650

MSc Queries

+44 (0) 20 7955 6879

 

Online query form|

BSc and MSc queries

People

Academic staff in social statistics

Bergsma

Dr Wicher Bergsma
Associate Professor

Research interests: categorical data analysis; statistical learning; testing independence; multivariate analysis; graphical modelling

Contact: w.p.bergsma@lse.ac.uk|
Personal page|

 
geneletti001

Dr Sara Geneletti
Lecturer

Research interests: causal inference; graphical models; Bayesian inference; evidence synthesis

Contact: s.geneletti@lse.ac.uk|
Personal page|

 
Kalogeropoulos

Dr Kostas Kalogeropoulos
Associate Professor

Research interests: Bayesian inference; Markov chain Monte Carlo; sequential Monte Carlo and inference on models with stochastic differential equations; infectious disease modelling and evidence synthesis

Contact: k.kalogeropoulos@lse.ac.uk|
Personal page| 

 
Kuha

Dr Jouni Kuha
Associate Professor

Research interests: model selection; measurement error, misclassification and missing data; latent variable modelling; analysis of cross-national survey data; social statistics

Contact: j.kuha@lse.ac.uk|
Personal page|

 
Moustaki

Professor Irini Moustaki
Professor

Research interests: latent variable models; structural equation models; categorical data; missing values; outliers; composite likelihood estimation methods; applications to social science and health

Contact: i.moustaki@lse.ac.uk|
Personal page|

 
skinner#3 

Professor Chris Skinner
Professor

Research interests: sample surveys; measurement error; missing data; statistical disclosure control; official statistics; social science applications

Contact: c.j.skinner@lse.ac.uk|
Personal page|

 
steele

Professor Fiona Steele
Professor

Research interests: statistical methods for social research, including multilevel modelling, event history (survival) analysis and structural equation modelling (SEM), with applications in demography, psychology, education and public health.

Contact: f.a.steele@lse.ac.uk|
Personal page|

 

Research staff in social statistics

karouzakis

Dr Nikolaos Karouzakis
Research Officer

Research interests: asset pricing; term structure of interest rates; yield curve modelling; pricing financial derivative (interest rate derivatives); Bayesian inference; Monte Carlo Markov Chain (MCMC) 

Contact: n.karouzakis@lse.ac.uk|
Personal page

 
Katsikatsou

Dr Myrsini Katsikatsou
Research Officer

Research interests: latent variable models; structural equation models; composite likelihood methods; categorical data; ranking-preference data; applications of latent variable models to behavioural and social sciences

Contact: m.katsikatsou@lse.ac.uk|
Personal page

 

Research students in social statistics

Dayan Yehuda 1

Yehuda Dayan
Research topic: Statistical inference in finite populations through web based sample surveys

Abstract:

Supervisors: Dr Jouni Kuha / Dr Wicher Bergsma

Contact: y.dayan@lse.ac.uk|
Personal page

 
Doretti

Marco Doretti
Research topic: Extensions of g-formula framework for handling time-varying confounding

Abstract: Many clinical trials wish to evaluate the effect of a sequence of treatments repeated over time on an outcome measured at a final stage. The problem of time-varying confounding arises naturally in this context, as there are several variables that at each temporal point act as confounders for the relation of interest but are also influenced by the treatment's value at the previous period, lying therefore on the treatments-outcome causal paths. This means that a standard regression model that controls for them in its specification does not provide a consistent estimate of the effect under investigation and is also likely to be affected by collider stratification bias involving unobserved variables. Thus, the well-known g-formula framework is reconsidered in order to overcome usual problems arising with real data-sets such as multicollinearity or poor data quality.

Supervisor: Dr Sara Geneletti

Contact m.doretti@lse.ac.uk|
Personal page

Marco is a visiting research student from Università degli Studi di Perugia| 

 
MaiHafez1

Mai Hafez
Research topic: Modelling multivariate longitudinal data subject to dropout using latent variable models

Abstract: Longitudinal data are collected for studying changes across time. Studying many variables simultaneously across time (e.g. items from a questionnaire) is common when the interest is in measuring unobserved constructs such as democracy, happiness, fear of crime, social status, etc. The observed variables are used as indicators for the unobserved constructs of interest. Dropout is a common problem in longitudinal studies where subjects exit the study prematurely. Ignoring the dropout mechanism can lead to biased estimates, especially when the dropout is non-ignorable. Another possible type of missingness is item non-response where an individual chooses not to respond to a specific question. Our proposed approach uses latent variable models to capture the evolution of the latent phenomenon over time while accounting for dropout (possibly non-random), together with item non-response.

Supervisors: Professor Irini Moustaki / Dr Jouni Kuha

Contact: m.m.hafez@lse.ac.uk|
Personal page

 
Terzi

Tayfun Terzi
Research topic: Cleaning data contaminated by semi-plausible response patterns

Abstract: A new challenge of bias arises through the increasing use of paid participants: semi-plausible response patterns (SpRPs). SpRPs result when participants superficially process the information of (online) experiments or questionnaires and try only to respond in a plausible way. This is due to the fact that participants who are paid are generally interested in earning fast money and efficiently try to overcome objective plausibility checks and process all other items only superficially, if they process them at all. Thus, those participants produce not only useless but detrimental data, because they attempt to conceal the researcher from their malpractice. The consequences are biased estimations and blurred or even covered true effect sizes; contaminating valid models. New methods for both their identification and treatment have to be developed.

Supervisors: Professor Chris Skinner / Dr Jouni Kuha

Contact: t.terzi@lse.ac.uk|
Personal page

 

 

Share:Facebook|Twitter|LinkedIn|