*2.6. Statistical Data Elaboration*

Data from GC-IT-MS, UPLC-DAD, and sensory analysis were subjected to one-way analysis of variance (ANOVA), and average values were compared by Least Significant Difference (LSD) test at the level of *p* < 0.05. Data were further processed by multivariate techniques, such as forward stepwise linear discriminant analysis (SLDA) and partial least squares discriminant analysis (PLSDA). The main goal of SLDA was to find the most useful variables (volatile compounds) for the mutual differentiation of all the six monovarietal EVOO. SLDA was applied on mean-centered data of a reduced dataset including six groups (varieties) and 50 variables with the highest *F*-ratios obtained in one-way ANOVA. Wilk's lambda was used as a selection criterion with an *F* statistic factor to establish the significance of the changes in Lambda when a new variable is tested (*F*-value to enter = 1). The main goal of PLSDA was to find the most useful variables (volatile compounds) for the differentiation of each of the six investigated monovarietal EVOO from all the other (five) monovarietal EVOOs. For this reason, PLSDA was applied on mean-centered data of six separate datasets each including two groups (a single vs. other five monovarietal EVOOs) and all the 197 variables. Variable Importance in Projection (VIP) scores were determined as the weighted sums of the squares of the weight in the PLSDA. ANOVA and SLDA data elaboration were performed by Statistica v. 13.2 software (StatSoft Inc., Tulsa, OK, USA), while PLSDA analysis was conducted using MetaboAnalyst v. 4.0 (http://www.metaboanalyst.ca) created at the University of Alberta, Canada [43].
