Deloitte Data Analyst Interview Questions

"An inner join retrieves rows only where there is matching criteria in both tables. A left join returns all records from the left table, and the matched records from the right table; if no match is found, NULL values are returned for the right table columns. A cross join returns a Cartesian product of the two tables, matching every row of the first table with every row of the second table."

"To handle missing data, I first assess the pattern of missingness (MCAR vs. MAR). If missing rows are minimal, I might remove them. For numerical values, I impute using median values to avoid outlier distortion, and for categorical values, I use mode imputation. If the missingness itself contains predictive value, I create a binary indicator column to flag that the data was missing."

"While analyzing churn data for a subscription service (Situation), I was tasked with investigating a sudden 5% rise in churn rate (Task). I ran a cohort analysis segmented by payment options and discovered that 80% of churned users experienced failed auto-renewals on specific debit cards (Action). We implemented a pre-billing email reminder and retry logic, saving over ₹12L in monthly revenue (Result)."

"Correlation means two variables tend to change together, but it does not imply one causes the other. Causation means one variable directly causes a change in the other. For example, ice cream sales and sunscreen sales are highly correlated, but buying ice cream does not cause sunscreen purchases; both are caused by the confounding variable of warm weather."

"To design an A/B test, I start by defining the null hypothesis: that the new design does not affect conversion rates. My primary metric would be the sign-up conversion rate. I use power analysis to determine the required sample size based on our baseline conversion rate, minimum detectable effect, and statistical power of 80% with a 5% alpha. After running the test to gather the required sample, I calculate the p-value; if it is under 0.05, we reject the null hypothesis."

"A cohort analysis involves grouping users who share a common event, such as a sign-up date, and tracking their behavior over time. I would use it to analyze customer retention and identify critical churn drop-off periods, helping the product team understand if updates are successfully retaining newer cohorts."