Linking Product Features to Perceptions: Self-Stated versus Statistically Revealed Importance Weights
比较了自述法和统计法在将产品特征与消费者感知联系起来时的效果，实证表明方差分析模型在预测效度上更优，支持使用统计方法。
The author compares self-stated and statistical techniques for linking product features to perceptions. An empirical application shows an analysis of variance model to be superior with respect to predictive validity, and thus supports the use of statistical rather than self-stated methods for linking features to perceptions.