Missing Data

Missing data are values that should be present but aren't — an empty cell, an “NA,” or a placeholder like -999. They arise constantly: a survey question skipped, a sensor offline, a record never joined.

How values went missing matters enormously. If they're missing at random, simple fixes are safe; but if missingness is systematic — say high earners skip the income question — then dropping or naively filling those rows biases every result. Always ask why a value is missing before deciding what to do, and beware tools that silently drop incomplete rows.