Deep dive · module 78 of 100
AI for data analysis
Use AI to describe data and gain initial insights.
After this lesson
You have AI describe a dataset and then verify the conclusions.
In view · Along the same yardstick
Figures require their own check
Use a small fictitious dataset with clear columns and units.
Have possible patterns described and assumptions named.
Verify important calculations with a spreadsheet or code.
Part 1 of 3 · Raw data
Your preference for less movement is on. Choose each part yourself.
Describe, do not calculate
AI can describe patterns, but it is better to perform exact calculations in a spreadsheet or code.
Ask for assumptions
Have AI indicate which assumptions it is making.
Check figures
Compare important numbers with the source data.
New practice example
From explanation to a usable result.
Use AI to explore an analysis, but verify definitions, units and calculations.
Fictional practice situationFictitious count: January has 10 requests and February 15. The column measures requests, not customers or revenue.
View a possible solution
This is how it can be done
The number of requests increases by 5, or 50% compared to January. Nothing follows from this regarding customer numbers or turnover without additional data.
Why this works
Keep absolute change, relative change and meaning separate. Verify the calculation independently.
Make it slightly more difficult
Let January contain zero requests. Why can you then not calculate a percentage increase in the same way?
Extra practice, at your own pace. You do not need to submit anything. This does not count towards an automatic assessment.
Your workplace · learning by doing
Make something. See what changes.
Checking the AI connection… You can already view the written example.
Present your assignment to Ende.
Read the answer, change one thing, and compare.
Previously held
New version
You remain in control
What do you want to improve in your next attempt?
On this device only. AI answers remain only in this page memory. This is self-review, not a grade.
Why this approach works
Keep absolute change, relative change and meaning separate. Verify the calculation independently.
A step further
Let January contain zero requests. Why can you then not calculate a percentage increase in the same way?
A prompt to get you started
Describe the main patterns in this data. Do not provide exact calculations unless you can substantiate them. Mark uncertainties. Data: [data].
Apply it
A spreadsheet contains duplicate rows and empty values. AI presents a ranking without reporting these issues.
Which approach best fits this situation?
- Define how you handle duplicates and missing values and recalculate in a verifiable way.
- Treat every empty value automatically as zero.
- Use the ranking without checking the data definition.
With JavaScript, you can check your approach and receive immediate feedback.
Test your insight
What is a safe way to use AI for data analysis?
- Letting AI do the calculations and trusting them immediately.
- Having AI describe them and checking the figures yourself.
- Never using AI for data.
Extra practice
Describe how you would use AI to explore a dataset, including how you would verify the conclusions.
Have AI describe which products are mentioned most often, then count the numbers yourself in the spreadsheet.
