Start with one question
Choose one question, such as which content group has more incomplete records. Explain the period and units used in this particular exercise.
Concrete output: One question, scope and two assumptions.
FICTIONAL LEARNING EXERCISE
Interpret fictional campaign records
Prepare one measurement question using eight sample content records. Figures are fictional, not advertising advice, revenue forecasts or market benchmarks.
Inspect the sample file
8 sample rows · 6 columns
| record_id | content_type | period | impressions | clicks | applications |
|---|---|---|---|---|---|
| M01 | guide | sample-week-1 | 180 | 12 | 2 |
| M02 | guide | sample-week-1 | 140 | 9 | 1 |
| M03 | event | sample-week-1 | 210 | — | 3 |
| M04 | event | sample-week-1 | 160 | 15 | — |
The table shows the first four rows. The file is fictional; some exercises intentionally include missing or conflicting values. Do not interpret them as real operational outcomes.
Download fictional sample CSV ↓Inspect missing values
Find missing click or application cells. Before treating an unknown as zero, explain how that choice would change the interpretation of the table.
Concrete output: Missing-data list and handling decision.
Build a small summary
Create a table comparing two content types. If using totals or ratios, state exactly which rows enter the calculation and which are excluded.
Concrete output: One table and its calculation scope.
Separate an idea from a result
Write a next research question prompted by the table. Do not claim that this file proves a real campaign works or will generate sales.
Concrete output: A one-page fictional exercise note.
My own project narrative
Write what you actually did and its limits in your own words rather than copying a prewritten achievement.
Ask yourself before saving
- Did I separate unknowns from zero?
- Are scope and calculations clear?
- Are proposals distinguished from measured outcomes?