Skip to content
BusinessTech AcademyBusinessTech AcademyData Dive

Data 3 — The Shape of the Picture

Rung 3 · what it shows · two periods

Data 1 asked whether the chart's numbers support its claim. This asks a different question:

Same numbers. Different chart. Different answer. Which one is honest?

Uses library/investigations/_protocol/data-visualizations.pptx — nine charts, images only — and supplies the prompt that record says is missing.

Part A — The five levers

Name them, because a student who cannot name the lever cannot see it being pulled.

#LeverWhat it doesThe tell
1The baselinewhere the axis startsa bar chart whose axis starts at 94%
2The denominatorcount or ratea raw total across places with wildly different populations
3The binhow the range is cut"under 25 / 25–64 / 65+" hides everything inside those groups
4The ordersorted, alphabetical, chronologicalsorting by value manufactures a ranking that may not exist
5The encodinglength, area, angle, colourpeople read length accurately and area badly; a circle twice as wide is four times the ink

Two rules that keep this from becoming "never truncate an axis":

A truncated axis is not automatically dishonest. For a stock price it is standard. For a bar chart of counts it is almost always misleading. The lever is neutral. The pairing of lever to question is what is honest or not.

There is no neutral chart. Every one of these five is a choice, and refusing to choose is not available. The obligation is to choose for the reader's question rather than your own conclusion.

Part B — Make three charts that disagree

Give every table one small dataset — twenty to forty rows, two or three columns is enough. Sources in _planning/source-list.md; usafacts.org and worldpopulationreview.com are the fastest to pull something usable from.

Each table produces three charts of the same data that a reader would take three different messages from. They must state, per chart:

Chart ___ makes it look like: ________ The lever I used: ________ A reader who would be misled by this: ________

Then the question that is not rhetorical:

Which of your three is the honest one?

The trap is that more than one usually is. Two charts of the same numbers can both be honest and answer different questions. The answer the lesson is looking for is not "chart 2" — it is:

"Chart 2 is honest for the question 'is it rising', chart 3 is honest for the question 'how big is it', and chart 1 is not honest for any question I can name."

A chart nobody can name a fair question for is the one to throw away.

Part C — Run the deck

Nine charts. For each, one job only: name the lever doing the work, or say "none — this is straight."

Not every chart has a lever pulled. Some are just charts. Saying so is a correct answer and should be worth the same as spotting one, for the same reason Data 1 caps the JUNK bucket at two.

The infographic hinge

This is where the strand hands off to the INFOGRAPHIC block, and the Academy already owns the standard. From the existing infographic material:

"If you took away the visuals, would the viewer lose any information?"

Put that next to the five levers and it reads differently than it did before:

An infographic is not a decorated fact. It is an argument made out of five choices. Students should be told that before they make one, not after it is graded.

Assessment

Critical Analysis rubric.

Where it runs

Year 10, IS1, in the DATA1 block (2/16 – 2/28), immediately before INFOGRAPHIC.

The block currently carries the Handoff's Phase 1 — running the Full Cycle protocol on the Appley feedback the sophomores collected. These two fit inside one block and belong together: Phase 1 analyses their own data, this lesson decides how to draw it, and the INFOGRAPHIC phase is where they have to.

Order within the block: Handoff Phase 1 (analyse) → Data 3 (how a picture argues) → the Vote → INFOGRAPHIC.

Also runs as a second pass in year 11 against a failure-analysis dataset, where product-failure-infographic.md is already the assignment.

Open