Generalized linear models using R

Target group

Engaged researchers and PhD students with basic knowledge on linear regression and some R programming experience.

Organisers

Biostatistics Unit, Faculty of Medicine and Health Sciences, Ghent University (statcel@ugent.be).

Added Value

The participants will gain theoretical and practical insight into complex statistical models, specifically applied to their domain of research. The sessions will be interactive and practical, making it possible for the participants to learn fast to use R in their research studies.

Format

The specialist course will consist of six interactive and practice sessions. During the sessions, specific exercises related to the domain of health sciences will be discussed. During the lectures, practical tips will be shared with respect to the use and interpretation of these models for specific research questions in the domain of health sciences.

Content

The course will offer 6 highly interactive sessions concerning the use of generalized linear models in health sciences. We will start with a basic linear regression model in R and gradually discuss more complex models. Topics discussed are: Linear regression, logistic regression, Lasso regression for variable selection, multilevel models . The last two sessions will focus on the use of generalized linear modeling in the case of excess zero’s, including hurdle and zero-inflated regression models.
All applications and examples will be from health sciences and medicine. Practical skills such as how to conduct the models in R, how to choose the most appropriate model and how to interpret the results, will also be part of the course.

Lecturers

  • Roos Colman and Dr. Stefanie De Buyser are both statistical consultants at the Biostatistics Unit at the Faculty of Medicine and Health Sciences, providing statistical advice and support to researchers. Roos Colman has a biomedical background which she combines with a master in Statistical data analysis. Dr. Stefanie De Buyser obtained a master of science in Drug Development and a PhD in Health sciences.
  • Dr. Jelle Van Cauwenberg is a Postdoctoral Fellow at the Department of Public Health (Ghent University). During the past years he completed several courses of the Master in Quantitative Analysis in Social Sciences and the Master of Statistics at the KULeuven. These courses included ‘Statistical software’, ‘Statistical modelling’, ‘Multilevel analysis’, ‘Mixed and multilevel models’ and ‘Longitudinal data analysis’. He applied generalized linear (mixed) models (including hurdle and zero-inflated models) in R for his own research. He previously taught statistics to bachelor and master students physical education and physiotherapy at the Vrije Universiteit Brussel and he lectured several courses on multilevel modelling and generalized linear models for PhD students. 

    When

    The specialist course will take place on November 5, 12, 19, 26 and December 3 and 10, 2019 from 14h till 16h30.

    Participants will be asked to engage in self-study (2 hours/week).

    Venue

    UZ campus - K3 (entrance 42), first floor- PC room 1.1

    Programme

    • 05/11/2019    Linear regression models (Roos Colman)
    • 12/11/2019    Logistic regression models (Roos Colman)
    • 19/11/2019    Variable selection in regression analysis using Lasso (Roos Colman)
    • 26/11/2019    Multilevel generalized regression models (Roos Colman)
    • 03/12/2019    Generalized linear models for data with excess zero’s (dr. Jelle Van Cauwenberg)
    • 10/12/2019    Hierarchical generalized linear models for data with excess zero’s (dr. Jelle Van Cauwenberg)

    DUE TO ILLNESS of dr. Jelle Van Cauwenberg, the 2 last sessions are delayed to:

    • -February 11th, 2020  at  2 PM
      -February 18th, 2020  at   2 PM

    Registration

    Registration link: https://eventmanager.ugent.be/genlinmodelsr

    If the course is fully booked, you can send an e-mail to and ask to be added to the waiting list. Please mention your student ID nr.

    Evaluation criteria (doctoral training programme)

    Participants are required to be present and participate actively during at least 5 out of 6 sessions. 

    Registration fee

    The course will be free of charge for members of the Doctoral Schools of UGent.

    Language

    The sessions will be in English except when all attendants are Dutch-speaking.

    Number of participants

    A maximum number of 18 participants can subscribe to allow highly interactive sessions in a pc room