← Back to series HSCI 230

Evaluating Epidemiological Research

Office Hours — the companion podcast for HSCI 230

Each episode is a relaxed walkthrough of the week's lesson — audio, summary, and full transcript on every page.

Foundations & evidence synthesis

Lessons 01 — 02

Foundations of EpidemiologyThe history of epidemiology, ways of knowing (epistemology, ontology, axiology, Indigenous knowledge, Two-Eyed Seeing), and research integrity, misconduct, and reform.

Systematic Reviews & Meta-AnalysisHow the field synthesizes across studies before any single study is appraised: the seven steps of a systematic review, fixed- vs. random-effects meta-analysis, forest plots and heterogeneity, publication bias and influence diagnostics.

Observational study designs

Lessons 03 — 06

Introduction to Observational StudiesOverview of observational study designs and their role in epidemiological research.

Case-Control StudiesSelecting cases and controls, measuring exposure retrospectively, and interpreting odds ratios.

Cohort StudiesFollowing exposed and unexposed groups forward in time, calculating incidence and risk, and interpreting relative risks and hazard ratios.

Ecological & Group-Level StudiesEcologic study designs, group-level variables, the ecologic fallacy, and sources of cross-level bias.

Threats to validity

Lessons 07 — 11

Conceptualization, Measurement & Causal SpecificationConstruct validity, measurement non-invariance, causal specification errors using DAGs, the obesity paradox, overadjustment bias, and residual confounding.

Sampling, Selection Processes & External ValidityBerkson’s bias, healthy worker effect, attrition, nonresponse, prevalence-incidence bias, survivorship bias, and transportability of study findings.

Information Bias & Data QualityNondifferential and differential misclassification, recall bias, social desirability, observer and detection bias, regression dilution, and digit preference.

Design-Specific & Temporal BiasesAllocation concealment, placebo and Hawthorne effects, immortal time bias, lead-time and length-biased sampling, time-window bias, and period and cohort effects.

Confounding & Statistical InferenceClassic and time-varying confounding, marginal structural models, model misspecification, multicollinearity, Simpson’s paradox, ecological and atomistic fallacies, the modifiable areal unit problem, and missing data mechanisms.