Data foundations
Walmart — Purchase Behaviour with CLT & Confidence Intervals
Statistical study of customer transactions: compared spending across gender, age and marital status using bootstrapping, the Central Limit Theorem and confidence intervals to back marketing decisions with evidence.
Bootstrapped CIs
3 customer segments
PythonPandasStatisticsCLTHypothesis Testing
Problem
Do different customer groups really spend differently on Black Friday — or is it just noise?
Approach
- Cleaned and explored the transaction dataset with Pandas.
- Used bootstrapped sampling to build sampling distributions of mean spend per segment.
- Applied the Central Limit Theorem to compute 90/95/99% confidence intervals.
- Checked where intervals overlap to decide which differences are statistically meaningful.
Outcome
Clear, statistically validated recommendations on which segments to target — a foundation case study from my MS program.