Statistics Interview Questions for Data Science

Data-science interviews may ask candidates to explain statistical concepts in the context of decisions. Microsoft Careers’ technical-interview guide lists probability, statistics, hypothesis testing, and p-values among possible data-science preparation areas. It also says interviewers may assess how a candidate analyzes, clarifies, and investigates a result. This is general public guidance for Microsoft, not a guaranteed question list for every employer. A p-value is calculated under a null hypothesis: it is the probability of a test statistic at least as extreme as the observed one, assuming that null hypothesis is true. It is not the probability that the null hypothesis is true, nor a measure of effect size or business value. A significance threshold should be chosen before examining results. A small p-value can be evidence against a null model, but the decision should also consider design quality, practical impact, uncertainty, and the consequences of errors. Multiple outcomes or repeated comparisons require care. NIST describes procedures such as Bonferroni for simultaneous inference and warns that repeating unadjusted pairwise comparisons does not generally preserve the intended overall confidence level. For an experiment, identify the primary outcome, analysis population, and decision rule in advance. A candidate should also ask whether observations are independent, how assignment occurred, and whether the result is large enough to matter. Explain assumptions rather than asserting certainty from a single threshold.