Loans doubled! Or did they?
An analyst tells the head teacher that loans in the school library have doubled. The head tells the whole school.
Then somebody finds the app logged each loan twice. The numbers were wrong and nobody had checked. Which step was missing?
Key words
| Word | What it means | Example |
|---|---|---|
| Prioritise | Put jobs in order, with the most important first. | Fix the dates before the charts |
| Build | Make what the analysis needs, such as a clean table. | Join loans and borrowers into one table |
| Blend | Combine data from different places into one. | Add the app data to the paper cards |
| Validation | Checking that the results are right and sensible. | Does the total match the app report? |
| Stakeholder | A person who cares about the result. | The head teacher |
| Database | A store of organised data in tables. | The loans tables |
| Loop | Steps that repeat until the result is right. | Fix, analyse, check, fix again |
| Audit | Check each step, with a record that another person can follow. | A log of every change to the data |
| Reiterate | Go back and repeat steps because the result was not right. | Re-clean the data, then analyse again |
| Test plan | A list of checks to run, and what a pass looks like. | Totals must match to the loan |
| Feedback | What people say about your work so you can improve it. | The librarian wants a weekly chart |
The cells students fill in are left blank in this preview.
By the end of this lesson
Note
I can put the stages of a data project in order and say what happens in preparation, analysis, validation, presentation and review.
Starter: last lesson
Have a go
Starter: last lesson
3 questionsYour answers are marked here and are not kept. A college account keeps every student’s marks.
Preparation: plan, then build
Plan and prioritise. Say how you will work. Choose the tools, such as a database or a notebook. Plan the time. Talk to stakeholders so you build what they need.
Build. Make the data ready. Create the tables you need. Blend data from different places into one. A project may take a month to plan and a day to run.
Worked example: each stage in the library
| Stage | What happens | In the library project |
|---|---|---|
| Understanding | Find the problem and size up the data | Do loans go up each month? |
| Preparation | Plan the work, then build the data | Choose the tools. Blend app and card data. |
| Analysis | Do the analysis tasks | Count loans per month and compare |
| Validation | Plan tests, validate and audit, repeat if needed | Check totals against the app report |
| Presentation | Report the result | Show the head a one-page summary |
| Review | Gather feedback on the presentation | Ask the head what was missing |
The cells students fill in are left blank in this preview.
Validation: is the result right?
Analysis uses the built data to do the job. Next comes validation.
- Plan tests. Write a test plan. Ask users to try the result too.
- Validate and audit. Check the results and the improvements. Keep a record of each check.
- Reiterate. If the result is not right, go back, fix the cause and run the analysis again.