4. Analyze the data

This stage involves statistical analysis of the data collected. This analysis aims to formalize relationships between the product and semantic, sensory or emotional descriptors, in order to generate design rules. Principal Component Analysis (PCA) is the first tool used to reveal correlations between variables, leading to a reduced space of independent variables. Tools such as ANOVA are then used to compare samples and determine which factors (age, gender, etc.) have a significant impact on the variability of evaluations. Other tools, such as Regression Analysis, can be used to highlight causal relationships between specific factors and evaluation results. In addition, Hierarchical Ascending Classification helps to highlight similarities and dissimilarities between products or possible sub-populations. The data analysis stage comprises a number of actions that enable the data to be analyzed...