Präferiertest

Präferiertest is a term used in statistics, psychology and market research to describe procedures for determining which among a set of alternatives is preferred by respondents. The tests aim to uncover explicit preferences or to infer underlying preferences from observed choices, ratings, or rankings.

Common formats include paired comparisons, where participants choose the preferred item in each pair; ranking tasks, where items are ordered from most to least preferred; and rating scales, where respondents rate each option on a scale of preference intensity. More elaborate approaches such as conjoint analysis or discrete choice experiments present bundles of attributes and ask participants to select or rate their preferred bundle, enabling estimation of the contribution of individual attributes to overall preference.

Data from Präferiertests are analyzed with models designed for choice and preference data. Paired comparisons can be analyzed with logistic models or the Bradley–Terry framework; rank data with nonparametric tests like Friedman or with rank-ordered logit models; rating data with ANOVA or ordinal logistic models when assumptions hold. Conjoint and discrete choice analyses typically use conditional logit, multinomial logit, or hierarchical Bayes methods to estimate part-worth utilities or attribute importances.

Applications span product development, marketing and branding research, transportation and urban planning, and consumer psychology. Design considerations include sample size, experimental design quality (e.g., orthogonality and balance), order effects, context effects, and the reliability and validity of the measures. Limitations involve potential biases in self-reported preferences, hypothetical bias in stated choices, and the generalizability of results beyond the experimental setting.

See also: preference testing, paired comparison, conjoint analysis, discrete choice. Etymologically, the term derives from the notion of preferring one option over another and the assessment of such preferences through a test or structured evaluation.