ABSTRACT
The goal of this chapter is to determine how multivariate statistics can play a role in determining botanical quality and authenticity. The most common approach for determining sample quality is to combine tests for known degradation products with a tasting panel’s sensory evaluation (Figure 8.1). Some examples of degradation products used for sample quality are pyropheophytin, a degradation product of chlorophyll, and 1-octen-3-ol, which is known to contribute to a musty odor. Neither chemical tests nor sensory panel results are ideal on their own. Most of the chemical tests only screen for a small number of well-characterized attributesfermented, rancid, musty, and so on. Tasting panels can capture nuanced information
Introduction ............................................................................................................ 107 Materials and Methods ........................................................................................... 109
General Workow ............................................................................................. 109 Cranberry Juice Experimental (GC/MS Workow) .......................................... 110 EVOO Experimental (GC/QTOF Workow) .................................................... 111 Identication of Compounds ............................................................................. 112 Multivariate Data Processing: Alignment ......................................................... 112 Multivariate Data Processing: Data Filtering .................................................... 115 Multivariate Data Processing: Statistical Analysis ............................................ 116 Developing Classication Models .................................................................... 119
Conclusions ............................................................................................................ 122 References .............................................................................................................. 123
about sample characteristics that are difcult to determine analytically, but are expensive and slow: expensive because it requires convening trained experts to evaluate sample quality and slow because a tasting panel is only capable of working for a limited time before their acuity begins to suffer. Multivariate statistics can offer a cheaper, more efcient approach for determining botanical quality and authenticity.
