Optional Activities
Every extra activity in the course, in one place — none of it required
Start here
Scattered through the pre-study pages and labs are a handful of optional activities. This page collects all of them so you can see the whole set at once instead of meeting them one at a time.
The short version
None of this is required. None of it is graded. Nothing here appears on an exam.
Every module’s required work is finished without touching a single item on this page. If you do the pre-study, the module, the lecture, and the lab, you are done — that is the whole course.
So why do these exist? Two different reasons, and the difference matters when you’re deciding whether to spend twenty minutes:
- More practice — the same skill you just learned, on different data. Useful if a technique didn’t quite land, or if you learn by repetition. Skip freely if it did land.
- Goes further — an idea or technique the course doesn’t otherwise cover. These are the ones worth a look if the topic touches the work you actually do.
A rule of thumb: if you’re short on time, skip all of it. If you have time and want more reps, use More practice. If you have time and want to know what’s past the edge of the course, use Goes further.
Module 4 · Data wrangling
More practice
- Reshape wide → long with Pew social-media data — practice pivot_longer() on a real Pew table, then chart the result. Pre-study.
Module 5 · Descriptive statistics
Goes further
- Challenge · Build an index from the SM11 items — combine three survey items into a single score, on data you procured yourself. There is no code on the page — you get the goal, the facts you can’t guess, and the checks your answer must pass. Closest thing in the course to being handed a specification and asked to build to it.
- Worked solution — for after you’ve made a real attempt. Lab.
Module 6 · Probability and distributions
More practice
- Does the MDE–SUD association differ by sex? — rebuild the cross-tab for males, compare the conditional probabilities across sexes, and chart the comparison. Lab.
- Apply the normal toolkit to a variable of your own — run the same pnorm() / qnorm() workflow on a different variable, including the fit check. Lab.
Goes further
- Diagnostic testing and positive predictive value — why a highly accurate screening test still returns mostly false positives for a rare condition. The single most consequential application of conditional probability in clinical work, and the one most often gotten wrong. Pre-study.
- Nested events: MDE and severe MDE — what happens to the probability rules when one event is a subset of another, and the surprising way the conditional behaves. Lab.
- Visualizing the binomial: from bars to a bell curve — plot the binomial and watch it converge on the normal as n grows. This is the Central Limit Theorem, seen rather than asserted. Pre-study.
Module 9 · Conducting hypothesis tests
Goes further
- The same clinician, twice: a paired look at Study 1 — what a within-subjects design buys you, on real repeated-measures data where the within-person correlation is essentially zero. A useful corrective: pairing helps for two reasons, and this dataset delivers only one of them. Lab.
What is not on this page
Two kinds of extra material live inside the lab and module pages rather than here, because they aren’t optional in the same way:
- Boxes marked Going further sit at the end of a lab and carry code you may well need later — the clearest example is the survey-weighted analysis in the M07 lab, which is how you would actually analyze NSDUH, BRFSS, NHANES, or any complex survey. Nothing there goes in your notebook, but if you work with survey data it is the most useful page in the course.
- The deep-dive guide — Multiple comparisons — is a reference page you’ll be pointed to from a lab when you need it.