Back to the lessonPreparing the printout…

Prompt Engineering · Module 3 — Execute: Apply, Compare and Automate

Period 7 · Task Execution — Data Visualisation Case

[video] Lecture video — Period 7: Task Execution — Data Visualisation Case — /api/lesson-media/videos/4d3653ff-a3ca-4f63-bf8e-bb93c553e599-1794842706243-29f096ab85324d2e.mp4

Flashcards · Period 7

Aggregation — Combining records into one value per group, such as the total net sales of each region.

SUM(Net_Sales_Amount) grouped by Store_Region gives four numbers from 200 rows.

SUM versus average — For "how much in total?" use SUM. An average of individual sales answers a different question (typical size of one sale).

East has the lowest total but the highest average per transaction.

Zero baseline — A value axis that starts at zero. Bars encode value by length, so a truncated axis exaggerates differences.

Chart encoding — How a chart maps data to visual features such as bar length and position. Check that the encoded measure is the one you asked for.

Net sales is not profit — Net_Sales_Amount records sales after discount. It contains no costs, so it cannot support claims about profit, margin or efficiency.

Evaluate checklist — Measure, values, scale, labels and units, claims. A correct ranking never explains why regions differ.

Chart specification — A written description of a chart (data, type, axes, labels) that you can build in another tool. Use it when the chatbot cannot draw.

Spreadsheet fallback — A pivot table or SUMIF formula that reproduces the grouped totals independently. State honestly which output came from the spreadsheet.

Chart refinement — Improving readability (order, clutter, emphasis, labels) while keeping values, measure and baseline unchanged, then rechecking.

Verification record — Brief, prompts, input version, verified table, charts, checks and revision notes saved together so a peer can reproduce the totals.

Prompt Engineering · Period 7 · Task Execution — Data Visualisation Case · printed on November 20, 2026