Abstract

This article illustrates by means of a case study how the usage of multivariate statistical techniques in marketing surveys yields far more meaningful and useful information than is obtained by conventional cross-tabulation. The case study reported here concerns the usage of snack foods by mothers in the Los Angeles, California area. Almost any competitive situation involving several products or brands, either consumer or industrial, could have been chosen for study. Snack foods were selected because their use has increased rapidly in recent years, and especially because the foods vary greatly in characteristics and desirability. Such products tend to make meaningful analysis using conventional techniques more difficult, and they should serve as a better showcase for the power of multivariate statistical analysis. A convenience sample of 150 mothers in the Los Angeles area was asked to rate each of 10 snack foods on 14 features or characteristics. The snack foods selected for study were: apples, oranges, raisins, whole milk, ice cream, potato/com chips, peanut butter sandwiches, snack crackers.