Data Science interview questions and answers

Duplicates inflate counts, double-weight users in means, and can be real repeated events or pipeline bugs.

Cardinality is the number of distinct values.

MCAR means missingness is unrelated to observed and unobserved values.

Median imputation is more robust on skewed numeric columns because a few giants do not dominate the fill value.

The fact of missingness can itself predict the outcome, for example users who skip billing fields.

Listwise deletion drops a whole row if any analysis variable is missing.

Values below Q1 minus one-and-a-half IQRs or above Q3 plus one-and-a-half IQRs are conventional box-plot outliers.

A common rule flags |z| greater than 3 under an approximate normal model.

Keep them when they are valid rare events you care about, such as whale spenders, outages, or fraud.

Winsorization caps extreme values at chosen percentiles instead of dropping rows, so sample size stays the same.

A box plot compactly compares medians, IQRs, and flagged extremes across groups, which is handy for many categories.

Heatmaps work well for correlation matrices, confusion matrices, or intensity over two discrete axes such as hour-by-weekday.

Lines preserve temporal order and make trends, seasonality, and shocks easy to see.

A confounder is a variable that influences both the treatment-like factor and the outcome, opening a back-door association.

Simpson's paradox is when a trend that appears in several groups reverses or vanishes when the groups are combined.

Reverse causality means the outcome is driving the supposed cause, not the other way around.

Stratified sampling draws separately within subgroups, such as country or plan tier, often in proportion to size.

Cluster sampling selects groups first, such as cities or schools, then observes units inside those groups.

The bootstrap resamples the observed dataset with replacement to approximate the sampling distribution of a statistic.

Sampling error is random fluctuation from drawing a finite sample, and it shrinks with n under good design.