Data Science interview questions for freshers

Unique counts reveal identifiers, near-constants, high-cardinality categoricals, and unexpected duplicates.

Confirm grain, time range, source, and how the extract was filtered before computing fancy stats.

Missing values are observations that were not recorded, not applicable, or lost in a pipeline, often shown as NaN, None, or SQL NULL.

None is a Python object, while NaN is a floating-point missing marker

Listwise deletion shrinks the sample and can bias results if missingness is related to the outcome, such as churned users skipping a survey.

Mean imputation replaces missing numeric values with the column mean.

An outlier is a point that sits far from the bulk of the distribution, whether by error or by being a genuine extreme.

The mean uses every value, so one huge number pulls it strongly.

A box plot shows the median, quartiles, and whiskers, with points beyond the whiskers often plotted as outliers.

Use a histogram to see the distribution of one numeric variable, including modality, skew, and range.

Bar charts compare discrete categories, with gaps between bars and a meaningful category axis.

A scatter plot shows the relationship between two numeric variables, one on each axis.

People compare angles and areas poorly, especially with many slices or similar sizes.

Correlation measures how two variables move together, often with Pearson's r for linear association.

Two series can move together because of a confounder, coincidence, reverse causality, or a shared trend over time.

Shoe size and reading ability correlate among children because age drives both.

The population is the full set you want to learn about, such as all active users in a market.

Random sampling gives each unit a known chance of selection, often equal chance in a simple random sample.

Full data may be too slow, too costly, or unavailable, as with surveys of millions of customers.

Sampling bias happens when some members of the population are systematically more likely to appear than others.