Training of a Dutch and Malaysian sensory panel to assess intensities of basic tastes and fat sensation of commonly consumed foods

Pey Sze Teo, Astrid W.B. van Langeveld, Korrie Pol, Els Siebelink, Cees de Graaf, Christophe Martin, Sylvie Issanchou, See Wan Yan, Monica Mars

Research output: Contribution to journalArticleAcademicpeer-review

4 Citations (Scopus)

Abstract

Taste has a nutrient sensing function and guides food choices. Therefore, investigating taste profiles of dietary patterns - within and across cultures - is highly relevant for nutritional research. However, this demands for accurately described food-taste databases, which are supported with data on the reliability and performance of the sensory panel that determined the taste values.This study aimed to assess the performance of a trained Dutch and Malaysian sensory panel. More importantly, we assessed whether the standardized training procedure in the two countries yielded similar taste profiles with respect to 15 basic taste solutions, and 19 foods differing in tastes.A Dutch (n = 15) and Malaysian panel (n = 20) were trained for 56-63 h, using basic taste solutions and reference foods on 6 scales, i.e. sweetness, sourness, bitterness, umami, saltiness and fat sensation. Performance of both panels was described by discrimination, repeatability (RMSE), and agreement. Nineteen products with different sensory characteristics were profiled in the Netherlands and Malaysia; subsequently the obtained taste profiles were compared.Both panels were able to discriminate between solutions and products (all p < .001). A vast majority of the taste values could be reproduced; the RMSEs of the different taste values varied between 2.3 and 13.3%. Panel agreement was achieved after the training with solutions; however not for all attributes of the reference foods. Some taste values of the 19 foods were significantly different, however most of these differences were small (<10. points).Our descriptive training procedure yielded two panels from different cultures that were similar in panel performance. More importantly, they obtained similar taste profiles for 19 different foods. This implies that food-taste databases obtained with valid and standardized training procedures may be used to quantify the sensory profiles of dietary patterns of populations.
Original languageEnglish
Pages (from-to)49-59
JournalFood Quality and Preference
Volume65
Early online date5 Dec 2017
DOIs
Publication statusPublished - Apr 2018

Fingerprint

Fats
Food
lipids
eating habits
saltiness
Databases
sourness
umami
bitterness
sweetness
Malaysia
food choices
repeatability
Netherlands
sensory evaluation
sensory properties
nutrients

Keywords

  • Cross-cultural
  • Panel performance
  • Spectrum
  • Taste profiling
  • Trained panel

Cite this

@article{f2cb22dbfbe04204ac98503edc7c23bc,
title = "Training of a Dutch and Malaysian sensory panel to assess intensities of basic tastes and fat sensation of commonly consumed foods",
abstract = "Taste has a nutrient sensing function and guides food choices. Therefore, investigating taste profiles of dietary patterns - within and across cultures - is highly relevant for nutritional research. However, this demands for accurately described food-taste databases, which are supported with data on the reliability and performance of the sensory panel that determined the taste values.This study aimed to assess the performance of a trained Dutch and Malaysian sensory panel. More importantly, we assessed whether the standardized training procedure in the two countries yielded similar taste profiles with respect to 15 basic taste solutions, and 19 foods differing in tastes.A Dutch (n = 15) and Malaysian panel (n = 20) were trained for 56-63 h, using basic taste solutions and reference foods on 6 scales, i.e. sweetness, sourness, bitterness, umami, saltiness and fat sensation. Performance of both panels was described by discrimination, repeatability (RMSE), and agreement. Nineteen products with different sensory characteristics were profiled in the Netherlands and Malaysia; subsequently the obtained taste profiles were compared.Both panels were able to discriminate between solutions and products (all p < .001). A vast majority of the taste values could be reproduced; the RMSEs of the different taste values varied between 2.3 and 13.3{\%}. Panel agreement was achieved after the training with solutions; however not for all attributes of the reference foods. Some taste values of the 19 foods were significantly different, however most of these differences were small (<10. points).Our descriptive training procedure yielded two panels from different cultures that were similar in panel performance. More importantly, they obtained similar taste profiles for 19 different foods. This implies that food-taste databases obtained with valid and standardized training procedures may be used to quantify the sensory profiles of dietary patterns of populations.",
keywords = "Cross-cultural, Panel performance, Spectrum, Taste profiling, Trained panel",
author = "Teo, {Pey Sze} and {van Langeveld}, {Astrid W.B.} and Korrie Pol and Els Siebelink and {de Graaf}, Cees and Christophe Martin and Sylvie Issanchou and Yan, {See Wan} and Monica Mars",
year = "2018",
month = "4",
doi = "10.1016/j.foodqual.2017.11.011",
language = "English",
volume = "65",
pages = "49--59",
journal = "Food Quality and Preference",
issn = "0950-3293",
publisher = "Elsevier",

}

Training of a Dutch and Malaysian sensory panel to assess intensities of basic tastes and fat sensation of commonly consumed foods. / Teo, Pey Sze; van Langeveld, Astrid W.B.; Pol, Korrie; Siebelink, Els; de Graaf, Cees; Martin, Christophe; Issanchou, Sylvie; Yan, See Wan; Mars, Monica.

In: Food Quality and Preference, Vol. 65, 04.2018, p. 49-59.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Training of a Dutch and Malaysian sensory panel to assess intensities of basic tastes and fat sensation of commonly consumed foods

AU - Teo, Pey Sze

AU - van Langeveld, Astrid W.B.

AU - Pol, Korrie

AU - Siebelink, Els

AU - de Graaf, Cees

AU - Martin, Christophe

AU - Issanchou, Sylvie

AU - Yan, See Wan

AU - Mars, Monica

PY - 2018/4

Y1 - 2018/4

N2 - Taste has a nutrient sensing function and guides food choices. Therefore, investigating taste profiles of dietary patterns - within and across cultures - is highly relevant for nutritional research. However, this demands for accurately described food-taste databases, which are supported with data on the reliability and performance of the sensory panel that determined the taste values.This study aimed to assess the performance of a trained Dutch and Malaysian sensory panel. More importantly, we assessed whether the standardized training procedure in the two countries yielded similar taste profiles with respect to 15 basic taste solutions, and 19 foods differing in tastes.A Dutch (n = 15) and Malaysian panel (n = 20) were trained for 56-63 h, using basic taste solutions and reference foods on 6 scales, i.e. sweetness, sourness, bitterness, umami, saltiness and fat sensation. Performance of both panels was described by discrimination, repeatability (RMSE), and agreement. Nineteen products with different sensory characteristics were profiled in the Netherlands and Malaysia; subsequently the obtained taste profiles were compared.Both panels were able to discriminate between solutions and products (all p < .001). A vast majority of the taste values could be reproduced; the RMSEs of the different taste values varied between 2.3 and 13.3%. Panel agreement was achieved after the training with solutions; however not for all attributes of the reference foods. Some taste values of the 19 foods were significantly different, however most of these differences were small (<10. points).Our descriptive training procedure yielded two panels from different cultures that were similar in panel performance. More importantly, they obtained similar taste profiles for 19 different foods. This implies that food-taste databases obtained with valid and standardized training procedures may be used to quantify the sensory profiles of dietary patterns of populations.

AB - Taste has a nutrient sensing function and guides food choices. Therefore, investigating taste profiles of dietary patterns - within and across cultures - is highly relevant for nutritional research. However, this demands for accurately described food-taste databases, which are supported with data on the reliability and performance of the sensory panel that determined the taste values.This study aimed to assess the performance of a trained Dutch and Malaysian sensory panel. More importantly, we assessed whether the standardized training procedure in the two countries yielded similar taste profiles with respect to 15 basic taste solutions, and 19 foods differing in tastes.A Dutch (n = 15) and Malaysian panel (n = 20) were trained for 56-63 h, using basic taste solutions and reference foods on 6 scales, i.e. sweetness, sourness, bitterness, umami, saltiness and fat sensation. Performance of both panels was described by discrimination, repeatability (RMSE), and agreement. Nineteen products with different sensory characteristics were profiled in the Netherlands and Malaysia; subsequently the obtained taste profiles were compared.Both panels were able to discriminate between solutions and products (all p < .001). A vast majority of the taste values could be reproduced; the RMSEs of the different taste values varied between 2.3 and 13.3%. Panel agreement was achieved after the training with solutions; however not for all attributes of the reference foods. Some taste values of the 19 foods were significantly different, however most of these differences were small (<10. points).Our descriptive training procedure yielded two panels from different cultures that were similar in panel performance. More importantly, they obtained similar taste profiles for 19 different foods. This implies that food-taste databases obtained with valid and standardized training procedures may be used to quantify the sensory profiles of dietary patterns of populations.

KW - Cross-cultural

KW - Panel performance

KW - Spectrum

KW - Taste profiling

KW - Trained panel

U2 - 10.1016/j.foodqual.2017.11.011

DO - 10.1016/j.foodqual.2017.11.011

M3 - Article

VL - 65

SP - 49

EP - 59

JO - Food Quality and Preference

JF - Food Quality and Preference

SN - 0950-3293

ER -