Session 1: Wednesday October 28th | Session 2: Thursday October 29th | Session 3: Friday October 30th
Overview
For event-time data, ordinary regression analysis methods are not suitable. Regression analysis that includes the element of time has two key problems:
- Censoring: Nearly every sample contains some cases that do not experience an event. If the dependent variable is the time of the event, what do you do with these “censored” cases?
- Time-dependent covariates: Many explanatory variables (like income) change in value over time. How do you put such variables in a regression analysis?
Survival analysis methods are designed to deal with censoring and time-dependent covariates in a statistically correct method. Originally developed by biostatisticians, these methods have become popular in sociology, demography, psychology, economics, political science, and marketing.
This webinar series, "Survival Analysis" describes the various methods used for modeling and evaluating survival data, also called time-to-event data. General statistical concepts and methods discussed in this webinar include survival and hazard functions, Kaplan-Meier graphs, log-rank and related tests, Cox proportional hazards model for time-varying covariates. The course will also require participants to use a convenient statistical package (e.g., SAS, SPSS, or R) to analyze survival analysis data.
These concepts will be covered in three webinars, each 2 hours long, delivered by Shayesteh Jahanfar, a highly trained and experienced webinar instructor. The interactive webinar software will provide remote access for students to view the instructor's screen, listen to the lecture in real time, and ask questions. The instructor will provide lecture slides (PowerPoint) as well as programming code for use with the statistical packages: SAS, SPSS, and R.
Prior required knowledge
This is a beginner to intermediate level workshop therefore participants will be expected to have some introductory knowledge of hypothesis testing, statistical power, correlation coefficients, and simple bivariate regression. Previous knowledge in and/or work with regression analysis is required.
Webinar objectives
By the end of the survival analysis webinar series, participants will be able to:
- Plan for and run survival analysis, checking statistical assumptions and appropriateness of the model results
- Communicate effectively with statistical analysts on survival analysis methods
- Understand the concepts such as censoring, Cox proportional hazards, graphing survival analysis data (Kaplan Meier plots), the Log Rank and related tests
- Work on a data set to produce tangible results to build a parsimonious model starting from descriptive analysis to model fitting
Pre-Webinar submissions
If you are (or will be) working on a project using survival analysis, submit a short overview with related questions to ann.greenwood@popdata.bc.ca prior to the start of webinar series for discussion during the last webinar session.
Instructor biography
Shayesteh has 21 years of teaching and research experience. She is an analytical consultant, an epidemiologist and a UBC instructor supporting health researchers to gain analytical and study design skills. She is a clinician, has a PhD from New South Wales University, Australia and is currently enrolled to obtain her second PhD at University of British Columbia.
Workshop fees
Regular Rate: $250
Student Rate: $150
Workshop Content
Part 1 – Survival analysis: Theory/assumptions
- Concept of Progression/Life-table
- Survival analysis compared with other regression techniques
- What is survival analysis?
- When to use survival analysis?
- Assumptions
- Univariate method: Kaplan-Meier (KM) curves
- How to do KM plots/analysis in three software formats: SPSS, SAS, R
Part 2 – Survival analysis: Multivariate methods
- Cox-proportional hazards (CPH) model
- Parametric models
- Assessment of adequacy of analysis
- How to do CPH analysis in three software formats: SPSS, SAS, R
Part 3 – Survival analysis: Start-to-finish modeling technique
- Case study: Analysing a data set from descriptive statistics to model building
- Review/discussion of students’ related research questions
