340.775.01 Measurement Theory and Techniques in Epidemiology
- 3rd term
- 4 credits
- Academic Year:
- 2019 - 2020
- East Baltimore
- Class Times:
- Tu Th, 10:30 - 11:50am
- Lab Times:
Friday, 9:00 - 10:50am
Modern peer-reviewed research opines for advanced, novel statistical methods to address important, increasingly complex scientific questions about the multivariate world. This course provides overviews of common and novel sources of measurement in epidemiology and delves into applications of advanced analytic models that are used to represent constructs from measures. Examples are taken from published epidemiologic research. The course will review necessary assumptions and appropriateness of models of construct representation given available data. From this course, students will learn to interpret, apply, and gauge the appropriateness of particular measurement methods.
Reviews concepts, key assumptions, and published applications of measurement theory, including true scores and counterfactual outcomes, latent variables, and validity. Explores novel applications of item response theory to refinement of measures, assessment of differential item functioning, and calibration of metrics across diverse samples. Topics include analysis of novel types of data (biomarkers, high-dimensional data, administrative records, genetics), item response theory, latent growth curve models for longitudinal data and their extensions, and cross-study statistical harmonization and co-calibration. Draws examples from epidemiologic applications in the behavioral and social sciences. Offers students opportunities for applying lessons from didactic lectures in a laboratory setting using prepared examples.
- Learning Objectives:
- Analyze categorical data using item response theory and interpret results
- Analyze and interpret latent growth curve models and extensions
- Analyze and interpret bivariate dual change score models
- Conduct integrative data analysis of constructs across multiple studies and time points that feature differing measures
- Identify and correct for differential item functioning
- Describe the place of one's own research along the continuum from qualitative to quantitative analysis
- Recognize measurement errors and nuances of measurement intrinsic to biomarker data, high-dimensional data (e.g., accelerometry and MRI data), genetic data, interviews and survey questionnaires, administrative data, and group-level aggregations of data.
- Briefly describe aspects of specific types and sources of measurements common in epidemiology, including genetics, surveys, administrative records, high-dimensional data, biological markers, and census-level data.
- Methods of Assessment:
Weekly computer labs with prepared examples (50%); Final data analysis project (35%); weekly Courseplus quizzes (15%)
- Instructor Consent:
No consent required