This talk focuses on new ways for investigating sensory-instrumental data relationships using component-based methods. The proposed approaches are related to manuscripts that are in preparation at the time of this abstract submission. The proposed methods could be used in a wide range of studies, but the talk will emphasize studies involving one reference treatment and many test treatments, which is
We investigate supervised principal component regression (SPCR) its sensory and instrumental results from a data set of eight pinot noir wines. Data include measured concentrations of volatile organic compounds from headspace—solid phase micro-extraction—gas chromatography—mass spectrometry (HS-SPME-GC-MS) and sensory descriptive analysis results on two selected sensory attributes evaluated orthonasally. We show the solution from SPCR as conducted conventionally is equivalent to
This tutorial surveys some of the most frequently used methods for exploring multivariate data sets fromsensory and consumer science. Component-based methods are often applied for data reduction andvisualization of results.The tutorial is divided into three parts.In Part 1, we contrast principal component analysis (PCA) with principal variable analysis, then discussmultiple factor analysis (MFA), which is often used to investigate multiblock
Test-control paired comparisons can be investigated after principal component analysis (PCA). We show that a centred PCA of test-control paired comparisons is equivalent to an uncentred PCA of control-centred results. We demonstrate how to conduct these analyses. We illustrate key properties of centred and uncentred PCA. We justify why coordinates of control-centred results can be visualized in principal components obtained
Supervised principal component regression (supervised PCR; SPCR) of sensory and instrumental results is conducted in new ways focusing on paired comparisons. Conventionally, SPCR is conducted on all objects. SPCR of all objects is shown to be equivalent to SPCR of all paired comparisons but different from SPCR of a subset of paired comparisons. SPCR of test-control paired comparisons is useful when the
This study examines how to carry out test-control paired comparisons after performing principal component analysis (PCA). Different approaches are proposed here, involving either centred or uncentred PCA, and their respective key properties are highlighted. In particular, we show centred PCA of test-control paired comparisons is equivalent to uncentred PCA of control-centred paired differences. It is customary to use a column-centred
This paper discusses the advantages of using so-called component-based methods in sensory science. For instance, principal component analysis (PCA) and partial least squares (PLS) regression are used widely in the field; we will here discuss these and other methods for handling one block of data, as well as several blocks of data. Component-based methods all share a common feature: they
A panel of trained sensory assessors often evaluates samples by quantifying the intensities of sensory attributes. In some cases, samples are instances of in-market or prototype products. To explore results from the panel, it is conventional to obtain a products-by-attributes table of means, center and variance-standardize its columns, then conduct principal component analysis (PCA). The principal components that extract variance
Principal component analysis (PCA) is a popular technique for summarizing and exploring multivariate data sets. We propose how to conduct PCA of results from sensory studies that have a special structure, where only a subset of the product paired comparisons are of interest. We illustrate the proposed approach with two data sets, both from trained sensory panels. In the first
A common task for trained sensory assessors is to evaluate samples by characterizing and quantifying the intensities of sensory attributes. In some cases, the samples are instances of in-market or prototype products. To explore results from the panel, it is conventional to obtain a products-by-attributes table of means, center and variance-standardize its columns, then conduct principal component analysis (PCA). The