SenseHub: an integrated web application for sensory analyses Muhammad Aswan Syahputra, STP, MSc Kiki Fibrianto, STP, MPhil, PhD Sensory and Applied Food Science Research Group Faculty of Agricultural Technology, Universitas Brawijaya

About me ● Master in Food Technology with Sensory Science specialization, Wageningen University and Research ● Expertise in Advance Sensory Methods and Sensometrics, data analysis, and R programming ● Founder of Sensolution.ID ● [email protected]

What is Sensory Science?

Sensory Science ~ Food + Consumer Perception Sensory food research aims to reach a better understanding of how the senses respond to food and eating, but also how our senses can be used in quality control and product design. - WUR Sensory Science Food science and technology Marketing and Consumer behavior Nutrition & Health

Quantitative (attribute based) Affective Qualitative (not attibute based) Profiling Checklist Specified Discrim. tests Unspecified Discrim. Tests Holistic Preference Profiling + hedonic Acceptance Scale-based Non scale-based ● QDA ● Free-choice profiling ● Flash profiling ● Just-about-Right (JAR) ● Ideal Profile Method (IPM) ● Check-all-that-Apply (CATA) ● Temporal Dominance of Sensation (TDS) ● 2 Alternative Forced Choice (2AFC) ● 3 Alternative Forced Choice (3AFC) ● Specified Tetrad ● Duo-Trio ● Triangle ● Tetrad, ● Two-out-of-Five ● Sorting ● Napping ● Word association ● Paired preference ● Ranked preference Credits: Betina Piqueras-Fiszman

How to analyse them? Statistical tools!

No content

Why do we need another tool? Think of features! Think of learning curve! Think of affordability!

Feel free to access it at: s.id/sensehub_basic

Features ● Based on R statistical language ● Free/Libre Open Source Software (FLOSS) ● Works in clouds, no installation required ● Covers discriminative, descriptive, affective sensory methods (21+ methods) ● Includes both conventional and rapid sensory methods ● Available in Bahasa Indonesia

Features Experimental setup Design of experiments, Panel performance analysis Discriminative test 2-Alternatives Force Choice (2AFC), 3-Alternatives Force Choice (3AFC), Duo-Trio, Triangle, Tetrad, Hexad, Two- out-of-Five Descriptive test Quantitative Descriptive Analysis (QDA), Flash Profiling (FP), Free-Choice Profiling (FCP), Rate-all-that-Apply (RATA), Check-all-that-Apply (CATA), RATA as CATA, Sorting Task, Napping Affective test Acceptance test, Paired Preference, Multiple Paired Preference, Preference Ranking, Hedonic Rating Optimisation Just-about-Right (JAR) method

Evaluation ● 35 participants (bachelor, master, and doctorate students) ● QDA dataset (12 perfumes, 103 panelists, 21 sensory descriptors) ● 10 questions about data analysis and interpretation ● 11 questions about user experience ● Time was recorded

Participants Total number of participants 35 subjects Gender 85.71% Female, 14.29% Male Age Average 25.89 years old, in range of 20-52 years old Education level 62.86% Bachelor, 28.57% Master, 8.57% Doctorate Analytical software ever used 8.57% MS.Excel, 85.71% Minitab, 2.85% PanelCheck, 25.71% SPSS, 8.57% XLSTAT

List of questions Q1 How many panellists participated in the study? Q2 Please list the samples that tested in the study. Q3 What are the sensory descriptors used to evaluate the sample? Q4 How many sensory descriptors show significant difference between sample? (P<0.05) Q5 How many sensory descriptors do not show significant difference between sample? (P<0.05) Q6 Please indicate the four sensory descriptors that are most discriminative. Q7 Which sample has the highest intensity in ‘woody’? Q8 How many dimensions or principal components are extracted from Principal Component Analysis (PCA)? Q9 How much is the cumulative frequency of the first three dimensions or principal components? Q10 Which samples have closest sensory characteristic similarity with L’instant?

List of questions Q11 How easy is it to import a dataset? Q12 How easy is it to set up the parameters? Q13 How easy is it to get basic information about the imported dataset? Q14 How intuitive is the user interface? Q15 How quick is the computation speed? Q16 How concise is the output of the statistical analysis? Q17 How understandable is the output of the statistical analysis? Q18 How easy is it to transfer the results into another document (word document, excel document, etc)? Q19 How good is the quality of the basic plots/graphs produced? Q20 How suitable is the application/software for sensory analyses? Q21 Please indicate the overall score for this application/software.

Rate of correct answers

No content

Time spent

No content

User experience

No content

Take home messages

● Superior in features compared to the software currently being used ● Conveys better user experience ● Significantly quicker ● Available in Bahasa Indonesia Benefits and limitations ● Requires internet access ● Documentations are not yet completed ● Evaluation only carried out for QDA methods ● Only one statistical software included for comparison

Any questions? [email protected]