Preference Mapping or Landscape Segmentation Analysis attempts to map panelists, products (brand a, brand b), and product features (fruity, minty, chewy,) on the same 2D plot using different techniques.
Internal preference mapping (or MDPref) has been available in SAS Market Research Application for some time. With MDPref, panelists are stored in columns and products in rows and the data represent an overall evaluation score (for each product). A form of PCA is performed on the panelist columns to reduce the space down to a few as two dimensions. Panelists are represented as vectors on the product space. "Clumps" of vectors represent panelist segments. MDPref maps can be extended to also include product attributes through correlating / regressing the mean attribute scores onto the MDPref map space.
External preference mapping proceeds in the opposite direction. It performs PCA on the attribute data, and then relates the overall evaluation scores as a second step using techniques like Ideal Point Model.
Landscape Segmentation Analysis is a term coined by the Institute for Perception but appears to use MD Unfolding on the overall evaluation data and then extend the attribute data onto ideal consumer point and product space.
Altogether, this represents a pretty meaty sized platform for the Multivariate Methods / Sensory-Consumer analysis space.
