Introduction to data management and causal inferences, conducting your own power analyses

Level

PhD students of the Doctoral Schools of Life Sciences & Medicine and Social and Behavioural Sciences of Ghent University, VUB and KU Leuven. 

Phd students need more information on those statistical techniques that are highly relevant in their fields (e.g. causal diagrams). Many scientific journals expect PhD students to have conducted a priori power analyses, or require post-hoc analyses when this was not done beforehand. With more complex study designs, researchers often struggle on how to conduct these analyses. Also, when PhD students want to apply for a mandate with the FWO, power analyses are requested.

Organising Committee

  • Leen Haerens (Faculty of Medicine and Health Sciences, Department: Movement and Sports Sciences, E-mail:
  • Greet Cardon (UGent; Department of Movement and Sports Sciences)
  • Jan Bourgois (UGent; Department of Movement and Sports Sciences)
  • Jan Boone (UGent; Department of Movement and Sports Sciences)
  • Lieze Mertens (UGent; Department of Movement and Sports Sciences)
  • Eva D’Hondt (VUB; Department of Movement and Sports Sciences)

Topic

This course will provide an introduction to the topics of data management and causal inferences, and will allow for PhD students to conduct power analyses for their own research. It will provide a general background on each of these topics, while simultaneously allowing for PhD students to ask specific questions regarding their ongoing research.

The course consists of three topics: Data management, casual inferences, and power analyses.
1. DATA MANAGEMENT: It will be explained why we, as researchers, should be care about Research Data Management. It will be explained how it could help you to take care of your data. It will be explained how to write a data management plan and how to familiarise yourself with Research Data Management.
2. CAUSAL INFERENCES: Popular tools such as causal diagrams, standardisation, propensity score methods, marginal structural models for evaluating the effect of time-varying exposures, mediation analysis for inferring causal pathways and finally instrumental variable methods will be introduced.
3. POWER ANALYSES: Sample size calculations are an essential part of study design. A well-designed trial is large enough to detect clinically important differences between groups with high probability (= Power). An Introduction to sample size calculations for some types of randomized controlled trials will be given.

Objectives

  • PhD students understand the concepts concerning power and sample size calculations
  • PhD students can perform sample size calculations for simple statistical tests
  • PhD students can report the results of a sample size calculation in a protocol, grant application, or methods section of a publication.

 

Programme & Venue

The sessions on data management and causal inferences are organized jointly by VUB, KULeuven and Ghent University. The sessions on power calculations are organized by Ghent University together with the Biostatistics Unit.

The course will consist of two parts:

- The first part includes an introduction to the topics of data management and causal inferences.

- The second part is organized in two sessions in June and September 2019 and considers the topic of power analyses.

Date Time Activity Lecturers Venue
Thursday 9 May 2019 8:30-11:00

SESSION 1

Part 1 – Causal inference: “Inferring causal pathways from data: challenges and some solutions”

 

Prof. dr. Stijn Vansteelandt

 

campus HILO, Middenzaal, Watersportlaan 2

11:00-12:00 Part 2 –  Practical application causal inference:
“Analyzing the risk for injuries among elite soccer players.”

 

Nikki Rommers

 

campus HILO, Middenzaal, Watersportlaan 2

13:00-14:15

SESSION 2

Introduction to research data management

 

Dr. Myriam Mertens

 

campus HILO, Middenzaal, Watersportlaan 2

Friday 14 June 2019 or Wednesday 3 July 2019 (in English)
13:00-18:00

SESSION 3

POWER ANALYSES: Introduction to sample size calculations for some types of randomized controlled trials (superiority trials, individually or cluster randomized controlled, with normally distributed or discrete outcome measures)

 

Dr. Ellen Deschepper

Dr. Stefanie De Buyser

Cel Biostatistiek

 

campus HILO, PC lokaal, Watersportlaan 2

Tuesday 3 September 2019 or Friday 6 September 2019
13:00-18:00

SESSION 4

POWER ANALYSES: Practical session with hands-on exercises based on case-studies

Dr. Ellen Deschepper

Dr. Stefanie De Buyser

Cel Biostatistiek

campus HILO, PC lokaal, Watersportlaan 2

 

Instructors/Lecturers

  • Part I: Data management

- Dr. Myriam Mertens  is Research Data Officer of the Ghent University and is expert in the field of Research Data Management. She is involved to help people to understand why Research Data Management could help you to take care of your data. Research data management is the compilation of many small practices that make your data easier to understand, less likely to be lost, and more likely to be usable during a project or ten years later.
Affiliation: UGent - Contact details:

  • Part I: Causal inferences

- Prof. Stijn Vansteelandt graduated as Master in Mathematics at Ghent University, and obtained a PhD in Mathematics (Statistics) at the same university. After postdoctoral research at the Department of Biostatistics of the Harvard School of Public Health, he returned to Ghent University in 2004, where he is now Full Professor (80%) in the Department of Applied Mathematics, Computer Science and Statistics. He is furthermore Professor of Statistical Methodology (20%) in the Department of Medical Statistics at the London School of Hygiene and Tropical Medicine.
Affiliation: UGent - Contact details:

- Nikki Rommers graduated as Master in Rehabilitation Sciences and Physiotherapy in 2015 at Vrije Universiteit Brussel, and obtained her Postgraduate in the Musculoskeletal Physiology Field and Manual therapy at the Ghent University in 2016. Currently, she is working as PhD-student supported by Research Foundation Flanders. This PhD is a collaboration between the Ghent University and Vrije Universiteit Brussel.
Affiliation: VUB - Contact details:

  • Part II: Power analyses

- Dr. Ellen Deschepper and dr. Stefanie De Buyser are working at the Biostatistics Unit of the Faculty of Medicine and Health Sciences. The primary purpose of the Biostatistics Unit is to deliver statistical advice and services to researchers of this faculty and the Ghent University Hospital. This implies advice with respect to study design, choice of appropriate statistical methods, sample size and power issues, support in data analyses, interpreting results, etc...The Biostatistics Unit further organizes internal courses and training sessions dedicated to specific statistical methods and software.

Affiliation: UGent - Contact details: 

Registration fee

Free of charge for members of the doctoral schools of  Ghent University, VUB and KU Leuven.

Registration & Information

To register, please follow this link: https://eventmanager.ugent.be/dmci

Evaluation criteria (doctoral training programme)

100% attendance
-Attendance and active participation
-Delivering a personal research case and conducting power analyses for a personal case