Know-How Exchange : Research/Metrics

Perceptual mapping offers a unique ability to communicate market structure analysis—i.e., the complex relationships among marketplace competitors and the criteria used by buyers in making purchase decisions and recommendations.

All mapping techniques attempt to show the comparative differences in how products or services are rated on a given set of attributes. The validity of a map depends on both
the overall set of attributes and brands in the study as well as the subset of attributes and brands evaluated by each respondent.

Most studies suffer from too many attributes. Manufacturers and service providers see hundreds of ways in which their products and services differ—or might differ—from those of their competitors. In most studies it is usually desirable (or necessary) to select a subset of attributes from respondents to rate.

You can use multiple discriminant analysis of respondent attribute ratings to produce perceptual maps. This analysis finds the optimal weighted combination of all the attributes which would produce the highest F ratio of between-product/service to within-product/service variation. That weighted combination of attributes becomes the first dimension of the map. Then a second weighted combination of attributes is found which has the next highest F ratio, subject to the constraint that this combination be uncorrelated with the first. The lack of correlation permits the plotting of the two dimensions graphically at right angles.
Once the weighted combination of attributes defining each dimension are determined,it is possible to compute the average score of each product or service on each
dimension. Those scores are used to plot the positions of the products or services in the representational space. The averages for all products are zero on each attribute
and also each dimension. Geometrically, this means that the “center of gravity” of all the product/service points lies at the center of the space.

Each attribute is plotted as a vector from the origin to a point which has as its coordinates the correlation of the attribute with the dimensions. This means that an attribute which contributes heavily to a dimension, and is therefore highly correlated with that dimension, appears on the map as an arrow pointing nearly in the same direction as the dimension.
The length of an attribute vector is equal to the square root of the sum of its squared correlations with the dimensions.

The relative length of an attribute vector in any two dimensional space is an indication of the extent to which that attribute is “accounted” for by those two dimensions
Products or services with vague, undifferentiated images, or those about which respondents disagree, lie near the center of the space. Those products / services with highest averages on an attribute are farthest from the center of the space in the
direction of its vector, and those products / services with lowest averages on an attribute are farthest from the center of the space in the opposite direction.