How do you decide what to do with outliers in your data?
What the interviewer wants to hear
- Finds outliers with plots and rules like IQR or z-scores
- Investigates whether they are errors or real events
- Fixes or removes errors; keeps real extreme values
- Uses robust measures like median or capping when needed
- Reports results with and without them
Model answer
I never remove outliers automatically. First I find them, using a box plot or histogram and a simple rule like 1.5 times the IQR. Then I investigate: is a 500,000 AED order a typing error, a test transaction, or a real corporate customer? Errors get fixed or removed, and I note how many. Real extreme values usually stay, because they can matter a lot to the business, for example a few big clients driving much of revenue. If they distort a summary, I switch to robust measures such as the median, or cap values at a high percentile for modelling. For important conclusions I run a quick sensitivity check: show the result with and without the outliers. If the conclusion flips, that is something the stakeholder must know rather than a detail I hide.
Can you answer it under time pressure?
Take a timed medium round and get scored on every answer.
More medium Data Analyst questions
- Daily orders dropped 20% yesterday. Walk me through how you would investigate.
- The product team wants to test a new checkout button. How would you design the A/B test?
- What are SQL window functions and when have you used them?
- How would you choose the key metrics for a food delivery app?
- When would you use Excel, SQL or Python for an analysis task?
- What is cohort analysis and when would you use it instead of looking at overall totals?
- Three managers send you urgent data requests on the same morning. How do you decide what to do first?
- Checkout conversion has fallen. How would you analyse the funnel to find where users are dropping off?
- How would you design the data model for a new Power BI sales report?