Instructor Notes

Overall approach

This module mixes concepts (Cartography Checklists, Fundamentals of Map Design), Python (the Introduction notebooks) and QGIS (Accessibility). The QGIS basics are taught in the Foundations module’s Getting Started with QGIS episode (moved there in October 2026). Make sure learners have both a Google account and QGIS installed before the session; QGIS installation is the most common delay.
Every figure in the concept episodes was generated from open data (U.S. Census Bureau, Bureau of Labor Statistics, Natural Earth, Gapminder) with code, so the examples are reproducible and properly credited.
All data is read from the web. The notebooks use Plotly’s built-in datasets; the Accessibility episode’s QGIS challenge loads the Indiana counties file from the Foundations module by URL.
Suggested timing for a half day: Introduction (30 min plus notebook 1a), Checklists (30 min), Map Design (60 min), Accessibility (30 min). Learners new to QGIS should first do the Foundations module’s Getting Started with QGIS episode (about 75 min). Notebooks 1b and 1c work well as take-home material.

Start with the misleading axis figure; it gets an immediate reaction and makes the point that correct data can still mislead.
Notebook 1a needs no prior plotting experience. Notebook 1c includes an optional Census section that skips itself when no API key is entered, so Runtime > Run all works for everyone.

Ask learners to bring (or find) a map from the news and apply the checklist to it during the discussion.
The “bad map” is deliberately terrible; ask the group to list every problem before revealing the list.

Normalization and classification are the two ideas learners most often get wrong in their own work. Spend time on the count versus percent figure and the six classification maps.
The classification figure uses Indiana population density, which is strongly skewed. Point out how equal interval puts almost everything in one class, and show the histogram figure to explain why.
Use The True Size Of website live; dragging Russia or Greenland to the equator is memorable.

Both build the same finished map of car-free households in Lafayette (tracts, parks, city limits, a locator inset). Natural breaks differ slightly between tools (QGIS first break about 2.4%, mapclassify about 4%).
Python: the basemap uses Esri World Gray Canvas through contextily because OSM and CARTO tiles often fail in Colab; the cell skips the basemap gracefully if tiles cannot be fetched. Notebook 02b installs matplotlib-scalebar.
QGIS: switching Graduated to Rule-based keeps the classes as rules; add the IS NULL rule for the dorm tract. $area uses the project ellipsoid; if label rules select nothing, set the ellipsoid to WGS 84.
QGIS may show the datum transformation chooser (EPSG:4326 to 26916); the first option is fine.
Scripted layout exports that include an online basemap can freeze QGIS on slow networks; for the class, keep the basemap optional.

Turn on View > Preview Mode > Simulate Color Blindness (Deuteranope) on the projector with a red-green map; it is the most convincing demonstration.
Prevalence figures (about 8% of men and 0.5% of women) are for people of Northern European ancestry; rates are lower in many other populations.

Symptom	Likely cause	Fix
QGIS will not open on macOS	Security setting	Right-click the app and choose Open
A web layer loads but nothing appears	Layer is off screen or under the basemap	Right-click the layer > Zoom to Layer; drag it above the basemap
Graduated symbology shows one color	Wrong field (text or an ID) selected	Choose a numeric field such as `density`
Colab `NameError`	A cell was skipped	Runtime > Run before
Interactive map shows “403” or “API key required”	OpenStreetMap or CARTO tiles blocked in Colab	Use `tiles="Esri.WorldGrayCanvas"`
Census section in notebook 1c fails	Key missing, mistyped or not activated	Check the key and its activation email, or skip the section