gewichtend

Gewichtend is a German present participle and adjective derived from the verb gewichten, meaning to assign weights to data or elements. It describes a process, method, or analysis that incorporates weights, as opposed to an unweighted approach. In statistics, econometrics and data science, weighting is used to reflect differing importance, frequency, or reliability of observations by multiplying contributions by weights and renormalizing.

In technical writing, the more common forms are gewichtete (weighted) rather than gewichtend, but gewichtend can appear when emphasizing the act of weighting, for example in phrases such as gewichtende Mittelung or ein gewichtendes Verfahren. The related noun is Gewichtung (weighting) and common terms include gewichteter Mittelwert (weighted mean), gewichtete Regression (weighted regression), and gewichtete kleinste-Quadrate-Verfahren (weighted least squares).

Gewichtende methods are standard in survey sampling, where weights adjust for unequal selection probabilities, nonresponse, or post-stratification adjustments. They are also used in constructing aggregate indices, in machine learning feature weighting, and in risk scoring. While weighting can improve representativeness and model accuracy, it introduces subjectivity in weight selection and can increase variance if weights are misapplied, necessitating careful normalization and sensitivity analysis.

See also: weighting, Gewichtung, weighted mean, weighted regression.