York College, CUNY · Health & Human Performance
PH 3203 credits · Hybrid
Applied Biostatistics in Public Health
Learn to read health data the way a careful analyst does: what the numbers mean, where they go wrong, and how to explain it to people who need to know. You won't write code. You'll judge what the code produced.
One question, all semester
Why is frequent mental distress higher in Jamaica than in most of New York City — and how would we know that's really true?
- Semester
- Fall 2026
- Meets
- Wednesdays, 10:00–11:50 a.m.
- Room
- Science 203
- Format
- Hybrid
This week · Week 4 · Sep 23
How do you know a study can say that?
Four studies, four claims. We work out how each one was built, and what that lets it say.
- Opener · Start hereFour studies, four claimsRead four real kinds of study, then work out how each one was built.
- LessonThe four designsWhere each design stands on the timeline, and what each one is allowed to claim.
- WorkshopFour more studies, and your portfolio pieceName the design, judge the claim, then publish a short piece to your GitHub site.
- Last weekWhat's actually in a dataset?Levels of measurement, the notebook, and your portfolio page.
- DatasetThe 168 ZIP codesThe file we work from all semester. Sort it, filter it, search it.
Weeks
What we'll cover
Each week's page opens here after we meet in class.
- 01Sep 2What biostatistics is for, and your first answer to the questionIn class
- 02Sep 9Where health data comes from, and who it never reachesOpen
- 03Sep 16What a dataset is; levels of measurementOpen
- 04Sep 23Study design: what a study is allowed to claimOpen
- 05Sep 30Out of whom? Why a count is not a rateComing soon
- 06Oct 7Populations aren't comparable as they are: age adjustmentComing soon
- 07Oct 14Exam 1 · Sessions 1–6, concepts onlyExam
- 08Oct 21A sample is not the population; numbers wobbleComing soon
- 09Oct 28Is the difference real? What a p-value says and doesn'tComing soon
- 10Nov 4Measuring association: relative risk and odds ratioComing soon
- 11Nov 11Association is not causation; confoundingComing soon
- 12Nov 18People report imperfectly: self-report, undiagnosed cases, estimatesComing soon
- 13Dec 2Tests are wrong sometimes: screening and the prevalence trapComing soon
- 14Dec 9Answering the question. Your September papers come back.Coming soon
- —Dec 15–21Exam 2 · Sessions 8–13 · Final memo dueExam
Hybrid course. Some sessions meet in Science 203 and some meet online. The format for each session is posted at least one week in advance, so check the site before every class.
Grading
How you'll be graded
ParticipationEntry answers and workshop sheets, collected every session, graded for completion
25%
GitHub portfolioYour best in-class work, curated across the semester
20%
Exam 1Oct 14 · Sessions 1–6 · no code, no formulas to memorize
15%
Exam 2Finals week · Sessions 8–13
15%
Final memoTwo to three pages answering the course question, for a community board
15%
LinkedIn postsTen posts connecting course concepts to current public health issues
10%
No points are ever deducted for code that does not run.