SPPH 504-007: Application of Epidemiological Methods

Ehsan Karim
ehsank.com

Description

This PhD-level course teaches research trainees emerging and advanced epidemiological methods, including statistical approaches for confounding, missingness and complex surveys towards the development of analysis plans, the analyzes and interpretation of real-world epidemiologic data and the communication of findings.

Contact

  • Feel free to email me for comments / suggestions.

Lab schedule

# Part Topic Lab Links Source type
1 1 Data manipulation Lab 1: Introduction to R HTML
2 1 NHANES Lab 2a: Creating Analytic dataset from NHANES HTML
2 2 Lab 2b: Working with Analytic dataset HTML
3 1 Model diagnostics Lab 3a: Regression diagnostics HTML
3 2 Lab 3b: Avoiding overfitting HTML
4 1 Survey data analysis Lab 4a: Creating the analytic survey dataset HTML
4 2 Lab 4b: Checking the analytic survey dataset HTML
4 3 Lab 4c: Analyzing the analytic survey dataset HTML
5 1 Propensity Score Matching Lab 5a: Matching HTML
5 2 Lab 5b: Propensity score Matching in CCHS HTML
5 3 Lab 5c: Propensity score matching in NHANES HTML
5 4 Lab 5d: Propensity score weighting in NHANES HTML
5 5 Lab 5e: Estimating propensity score weighting for multiple treatments HTML
6 1 Missing data analysis Lab 6a: Missing data and Imputation HTML
6 2 Lab 6b: Missing values in survey data (binary variable with missing values) HTML
6 3 Lab 6c: Missing values in survey data (multiple variables with missing values) HTML
6 4 Lab 6d: Missing values in propensity score analysis involving survey data HTML
6 5 Lab 6e: Propensity score sensitivity analysis, when weights are too large - 1 HTML
6 6 Lab 6f: Propensity score sensitivity analysis, when weights are too large - 2 HTML
6 7 Lab 6g: Model performance from multiple imputed datasets HTML
7 1 Machine learning Lab 7: Machine learning and Application in Propensity Score Analysis HTML
8 1 RMarkdown for Scientific Writing Lab 8: RMarkdown Video
9 1 Complex outcomes Lab 9a: Matched data HTML
9 2 Lab 9b: Polytomous and Ordinal Regression HTML
9 3 Lab 9c: Poisson and Negative binomial regression HTML
9 4 Lab 9d: Survival analysis HTML
10 1 Mediation Analysis Lab 10a: Ideas of Mediation analysis on CCHS HTML
10 2 Lab 10b: Mediation analysis on CCHS HTML
10 3 Lab 10c: Mediation analysis (3 category mediator) on CCHS HTML
10 4 Lab 10d: Mediation analysis using various R packages HTML
11 1 Longitudinal Data (optional) Lab 11a: Mixed Effects Models HTML
11 2 Lab 11b: GEE HTML

Acknowledgements

  • Derek Ouyang/SPPH (acted as GTA for 2019-2020),
  • M Atiquzzaman/Pharm (acted as GTA for 2018-2019) and
  • course participants from 2019-2020 and 2018-2019.