Comparison of Canonical Variate Analysis and Principal Component Analysis on 422 descriptive sensory studies

Peltier, Caroline; Visalli, Michel; Schlich, Pascal

doi:10.1016/j.foodqual.2014.05.005

Cited by 43 publications

(20 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For a comparison of MANOVA and CVA to principal component analysis, the usual way of mapping descriptive sensory data, the reader is referred to Peltier et al. ().…”

Section: Methodsmentioning

confidence: 99%

Perceived complexity in Sauvignon Blanc wines: influence of domain-specific expertise

Schlich

Maraboli

Urbano

et al. 2015

Australian Journal of Grape and Wine Research

Self Cite

View full text Add to dashboard Cite

Complexity is a multi-dimensional and poorly-defined term that is employed frequently when wine is characterised organoleptically. The present study's aim was to investigate the sensorial nature of perceived complexity in white wine as a function of domain-specific expertise. Eightyseven French participants (16 wine professionals, 30 wine connoisseurs, and 41 wine consumers) evaluated thirteen Sauvignon blanc wines that had been produced in New Zealand. The wines were part of a wine innovation project aimed at increasing perceived complexity in Sauvignon wines.The within-subject design required each participant to evaluate all wines by two methods, free sorting and by judging complexity qualitatively and quantitatively. The latter involved measurement of perceived complexity and assumed sub-components of wine complexity via a questionnaire (Medel, 2011). Results showed that sorting behaviour across groups was similar qualitatively, with each group classifying the wines in much the same way. On the other hand, between-group differences were observed in variability with greater within-group consensus amongst oenologists than wine consumers. The complexity questionnaire data showed differences in ratings as a function of both participant expertise and wine. In terms of theories concerning cognitive processing associated with perception of wine complexity, the results are in keeping with the notion that complexity is associated with aspects of harmony and wine balance, rather than with perceptual separability of wine components.

show abstract

“…For a comparison of MANOVA and CVA to principal component analysis, the usual way of mapping descriptive sensory data, the reader is referred to Peltier et al. ().…”

Section: Methodsmentioning

confidence: 99%

Perceived complexity in Sauvignon Blanc wines: influence of domain-specific expertise

Schlich

Maraboli

Urbano

et al. 2015

Australian Journal of Grape and Wine Research

Self Cite

View full text Add to dashboard Cite

show abstract

“…This is slightly different from standard CVA since it contrasts the between-samples covariance matrix with the interaction covariance matrix (interaction between assessor and samples) instead of the within-group covariance matrix. By doing so, CVA draws the product map based on product means with consideration of subject variability (Peltier, Visalli, & Schlich, 2015b).…”

Section: Aggregated Data In Time Intervalsmentioning

confidence: 99%

When the choice of the temporal method does make a difference: TCATA, TDS and TDS by modality for characterizing semi-solid foods

Nguyen

Næs

Varela

2018

Food Quality and Preference

View full text Add to dashboard Cite

For describing the evolution of sensory properties during eating, dynamic sensory methods are still being developed and optimised. Temporal Dominance of Sensations (TDS) and Temporal Check All That Apply (TCATA) are currently the most used and discussed. The aim of this study was to compare TDS, TCATA and a variant of TDS, performed by modality (M-TDS) in the outcome of the dynamic sensory description. These methods were applied with the same trained panel (n = 10) for the evaluation of the dynamic properties of yoghurt samples, with identical composition, only varying in textural properties. Based on a design of experiment, the yoghurts varied in viscosity (thin/thick), size of cereal particle added (flour/flakes) and flavour intensity (low dose/optimised dose, by adding artificial sweetener and vanilla). The TDS curves revealed that the variation in viscosity and particle size led to differences in perception mainly at the beginning of the eating process (Thin/Thick and Gritty/Sandy). Additionally, all samples were also perceived as Bitter at the end of the eating process. TCATA and TDS by modality results were, generally, in agreement with TDS, but they unveiled more details of the samples' dynamic profiles in all stages of the eating process, showing the effect of Vanilla and Sweet for the samples with optimised flavour, and the masked perception of Bitter. The duration of the eating process was standardized and split into three time intervals (T0-T40, T41-T80, T81-T100). Panelists' responses were summarized as frequency values in each time interval. Principal Component Analysis was used to visualize sample trajectories over time in the sensory space, with the need to study up to the third dimension to better understand the trajectories. ANOVA models were used to find the attributes which were significantly differences among products. Panel performance was assessed based on MANOVA models for the three methods. The results indicated that TCATA was more discriminative and panelists were more in agreement. TCATA also described samples in more detail in terms of number of discriminating attributes as compared with TDS. The discussion also centers in the different aspects of perception that could respond to different research questions for the three compared methods.

show abstract

“…1). Therefore, the interpretation of the PCA was limited to the first two axes (Peltier et al, 2015) which explained the 63.5% of the total variance. But only the first two contained >20% of the explained variation (Table S3).…”

Section: Sensory Characteristicsmentioning

confidence: 99%

“…But only the first two contained >20% of the explained variation (Table S3). Therefore, the interpretation of the PCA was limited to the first two axes (Peltier et al, 2015) which explained the 63.5% of the total variance. As indicated in Fig.…”

Section: Sensory Characteristicsmentioning

confidence: 99%

Correlations of sensory parameters with physicochemical characteristics of Argentinean honeys by multivariate statistical techniques

Patrignani

Ciappini

Tananaki

et al. 2017

Int J of Food Sci Tech

View full text Add to dashboard Cite

Summary Honey acceptability is mainly determined by its colour, crystallisation degree and aroma. In the present work, the sensory characteristics and physicochemical parameters of Argentinean honeys from different ecoregions were analysed. Moisture content, Pfund colour, diastase activity, hydroxymethylfurfural content, electrical conductivity, sugar profile and volatile compounds were analytically determined in honey samples, while sensory characteristics (crystal size, fluency score, sweetness, persistence, granularity, crystallisation, colour intensity and aroma) were evaluated by a trained panel. Significant correlations were found between honey crystallisation degree and hydroxymethylfurfural content and diastase activity (P ≤ 0.05). It could be confirmed that honey crystallisation interferes with the visual perception of colour. Floral, fresh fruit, ripe fruit, balsamic and wood aromas could be successfully linked to honey volatile profile (P ≤ 0.05). These results demonstrate that the parameters that could best guarantee the consumers’ preference can be successfully associated with the chemical composition of honey by multivariate statistical analysis.

show abstract

Comparison of Canonical Variate Analysis and Principal Component Analysis on 422 descriptive sensory studies

Cited by 43 publications

References 10 publications

Perceived complexity in Sauvignon Blanc wines: influence of domain-specific expertise

Perceived complexity in Sauvignon Blanc wines: influence of domain-specific expertise

When the choice of the temporal method does make a difference: TCATA, TDS and TDS by modality for characterizing semi-solid foods

Correlations of sensory parameters with physicochemical characteristics of Argentinean honeys by multivariate statistical techniques

Contact Info

Product

Resources

About