Data from 268 panellists rating each of 20 different products on 7 attributes. It is presented
data.frame with 268 observations on 140 variables. Each observation
represents a different trained panellist. The columns correspond to products and items.
The 20 different products are coded by alphabetic letters from A to T, and the items are coded
from 1 to 7. So item C5 corresponds to product C being rated on item 5.