$150 – $400

Multiple Dates

NB-IRDT Workshop Series: Using Regression Models to Understand Populations

Event Information

Share this event

Date and Time

Location

Location

Wu Conference Centre

6 Duffie Drive

Fredericton, NB E3B 0R6

Canada

View Map

Refund Policy

Refund Policy

Contact the organizer to request a refund.

Eventbrite's fee is nonrefundable.

Event description
Using Regression Models to Understand Populations

About this Event

Workshop Series

March 12 - Workshop 1: Describing and Comparing Populations
April 16 - Workshop 2: Selecting a Regression Model
May 14 - Workshop 3: Building a Regression Model

Series Description:

In this workshop series, learners will gain an understanding of how regression modelling can be used to describe and make inferences about populations of interest. Each workshop uses population-based data for guided examples; but methods taught are applicable to data collected for many purposes, such as in clinical research or survey studies. Bring your laptop – with your preferred statistical software. In the first half of the day we will review key material using guided examples; and in the afternoon you will have the opportunity to apply the methods learned to your own data (datasets will be provided to those who don’t have their own). This workshop series is led by Dr. Sandra Magalhaes, a researcher at the New Brunswick Institute for Research, Data and Training (NB-IRDT). Dr. Magalhaes is an epidemiologist with a strong interest in research methods and population health, and her research has been primarily focused on conditions of the brain such as dementia and multiple sclerosis.

Each workshop is catered. Seating is limited. This session is being offered in English only.

March 12, 2020

Workshop 1: Describing and Comparing Populations

Learning Objectives:

  • Define statistics used to describe and compare populations
  • Identify key components of effect measures
  • Select an appropriate effect measure for the research design
  • Interpret confidence intervals around effect measures

At the heart of regression modelling are the measures used to describe and compare populations of interest. Key measures include incidence rate, prevalence and probability, as well as comparative measures such as risk difference, odds ratio and hazard ratio, among others. We will focus on developing a conceptual understanding of the purpose and key components of effect measures and their application in helping us describe and compare the world around us. This one-day workshop is the first of a three-part series intended for individuals who are familiar with statistics and want to learn about using regression modelling.

April 16, 2020

Workshop 2: Selecting a Regression Model

Learning Objectives:

  • Define different types of outcomes
  • Identify key components of regression models
  • Select an appropriate regression model for the research design
  • Interpret regression model estimates and corresponding confidence intervals

A first step in conducting a regression analysis is selecting the right model. Regression models are available for different types of outcomes such as presence/absence, time at risk or time to occurrence of the outcome of interest. Linear, logistic, Poisson and survival models are a few examples of regression models that are commonly used. We will focus on developing a conceptual understanding of common regression models and effect measures they estimate to help you in selecting a model that fits the research design. This one-day workshop is the second in a three-part series intended for individuals who are familiar with statistics and want to learn about using regression modelling. Participation in Workshop 1 is not a prerequisite for Workshop 2.

May 14, 2020

Workshop 3: Building a Regression Model

Learning Objectives:

  • Define variables for modelling purposes
  • Identify key components of confounding and interaction identification
  • Select the ‘best’ regression model using appropriate model fit criteria
  • Interpret adjusted regression model estimates and corresponding confidence intervals

Once a regression model is selected, available data are then used to build a model to help us learn about relationships of interest. Important considerations for model building include variable coding, identifying confounding and interactions and using model selection criteria. We will focus on developing a conceptual understanding of the process of model building and interpreting adjusted effect measures and interaction terms. This one-day workshop is the last of a three-part series intended for individuals who are familiar with statistics and want to learn about using regression modelling. Participation in Workshops 1 and 2 is not a prerequisite for Workshop 3.

Share with friends

Location

Wu Conference Centre

6 Duffie Drive

Fredericton, NB E3B 0R6

Canada

View Map

Refund Policy

Contact the organizer to request a refund.

Eventbrite's fee is nonrefundable.

Save This Event

Event Saved