This following example illustrates how the ingredients of a tobacco product would be represented and quantified using 3 datasets:

• Tobacco Ingredients (IT) is used for representing the tobacco ingredients and their attributes within the tobacco product under study.
• Non-tobacco Ingredients (IN) is used for representing other ingredients/additives and their attributes as used in the tobacco product under study.
• Ingredient Quantities by Component (IQ) is used for representing the quantity and function of each ingredient (both tobacco and non-tobacco) by component of the tobacco product in which they are included. These values represent the design/formulation specification of the product, not experimentally determined values (as the latter would be represented in the Product Testing (PT) dataset, not shown here).

All 3 datasets use the SPTOBID variable to identify the tobacco product under study, and the IGDCMPID (ingredient or component identifier) variable to represent the ingredient name. IGDCMPID must be assigned by applicants to be unique with regard to the attributes that describe it. For instance, if the same ingredient varies on 1 or more attribute (e.g., 2 different manufacturers supplied the same ingredient that is used in 2 different components of the product), IGDCMPID must be unique for each use to reflect this difference, even if they share the same common name, and/or the same IUPAC name and CAS registry number. See IT dataset example, which illustrates this, below.