##### Question You have two datasets, sample A and sample B. They might come from two versions of a system, two cities, or two periods of autonomous-driving operations. The variable of interest may be continuous, skewed, heavy-tailed, or categorical, and the two sample sizes may be very different. 1. **How would you test whether the two samples come from the same underlying distribution?** Discuss: - continuous versus categorical variables; - univariate versus multivariate comparisons; - parametric versus nonparametric options, and the assumptions each common test makes; - visual diagnostics, effect sizes, and how you interpret **statistical** significance versus **practical** significance; - multiple-testing issues when you compare many features; - what to do when the two sample sizes are very different. 2. **What sampling methods do you know, and when is each appropriate** for data collection or offline evaluation? Cover at least simple random, stratified, cluster, systematic, and weighted/importance sampling, plus any others you consider relevant (e.g., multistage, reservoir, convenience), and say which you would prefer for product evaluation and why.