Applied data analysis
Nine weeks on the part of analysis nobody teaches: understanding what was really being asked, choosing a model the decision can bear, and writing the answer so it gets read.
- Length
- 9 weeks
- Time
- 4 h a week
- Enrolled
- 1,704
- Next cohort
- January 2027 cohort
- Written and marked by
- Mara Okonjo Lead instructor, Evaluation
Four modules, one assignment each
Every module ends with work that is read by the person who wrote it. The assignments build on each other, so the order is fixed.
- 01
SQL beyond SELECT
Window functions, joins that do not silently multiply rows, and the habit of checking a count before believing a sum.
- 02
Modelling for decisions
The model is chosen by the decision, not the data. Matching the method to what someone will do with the number.
- 03
Applied statistics
Uncertainty stated in a sentence a manager can act on. Confidence intervals, base rates, and when a difference is noise.
- 04
Reporting that gets read
One page, one chart, one recommendation. Writing for the person who has four minutes and a decision to make.
Who it is for
People whose job has quietly become analytical: product managers, operations leads, founders, marketers who now own a dashboard. You are asked for numbers, you produce them, and you are not always sure they are right.
It is also the recommended first track for people heading towards the systems or LLM programmes without a data background. The scoping call will say if that applies.
How the nine weeks run
Two taught hours a week on a shared dataset from a partner company, anonymised but messy in the way real data is. Two hours on your own on the fortnightly assignment. The final brief is a real question from a real stakeholder, and the review asks one thing above all: would you bet on this number.
