nonmissing

Nonmissing refers to data values that are observed and recorded for a variable, as opposed to missing values where no data are available. In data analysis and statistics, nonmissing data are those entries that can be used directly in calculations for a given variable. The term is commonly contrasted with missing data, which can arise from nonresponse, collection errors, or data corruption.

Missing data are often discussed in terms of mechanisms that describe why values are absent: missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR). These mechanisms influence how nonmissingness affects analyses. For example, complete-case analysis uses only observations with nonmissing values across the variables of interest, but this approach can reduce precision and may introduce bias if the missingness is not MCAR.

To address missing data, researchers employ methods that leverage nonmissing information in different ways. Imputation techniques fill in missing values, while model-based and multiple-imputation methods use nonmissing data to estimate plausible values and associated uncertainty. In data processing workflows, identifying nonmissing entries is a routine step for producing accurate summaries, statistics, and visualizations.

In practice, the concept of nonmissing emphasizes data quality and the feasibility of analysis. High nonmissingness simplifies analysis and interpretation, whereas extensive missing data necessitate careful handling to avoid biased conclusions. See also missing data, complete-case analysis, multiple imputation, and data imputation.