Dr Hal MacFie

Consultant in Sensory & Consumer Science

Hal MacFie Training Services
Company Registration No: 7864651
Company Address: 43 Manor Road
Keynsham, Bristol BS31 1RB, UK
Contact Tel: (44) 0117 9863590

About Us!

Hands on Advanced Sensory Statistics Course

Designed to update the practicing sensory scientist with knowledge and tools to take advantage of recent developments in this fast moving field.
The course is scheduled to run straight after the Pangborn Conference in Rhode Island.

Hal MacFie, Anne Hasted

August  24 (pm) to 25 2017

Venue:New Horizons of Rhode Island
24 Albion Rd
Lincoln, Rhode Island  02865

Course Organiser
Hal MacFie
For details and registration e-mail:

This course tackles recent developments in sensory science, bringing new statistical methods to bear and reviewing recent literature. It covers the analysis of QDA type data, developments in performance monitoring and difference testing. The most recent rapid methods are reviewed and the issue of when naïve consumers can be used instead of sensory assessors is addressed. The use of multi-way methods to relate sensory to other data sets is dealt with. Finally new approaches to dealing with modelling sensory pathways are explained and exercises provided using the XLSTAT Path PLS module. Wherever possible attendees will get hands on experience in analysing data from these methods.


Dr Hal MacFie is a statistician by training, with an international reputation in the areas of product assessment and consumer research. Author of over 100 scientific publications and three books, he was Head of Consumer Science at the Institute of Food Research, Reading England, before becoming Director of the Institute. He was for 20 years European editor of the Food Quality and Preference journal. Dr MacFie is now an independent consultant, and a Visiting Professor at Reading and Nottingham Universities and continues to develop innovative training courses for Sensory Professionals.

Anne Hasted began her career as an academic statistician at Reading University, leaving in 1989 to set up Qi Statistics Ltd offering statistical training, specialised software and consultancy support to research and industry. She has wide training and consultancy experience in both sensory and consumer research and has an international reputation for “user friendly” training.

Course Outline Day 1  

Thursday 24thAugust


Analysis of sensory descriptive data

Mixed Model Analysis
Schlich two way interaction extension
Brockhoff two way scaling – theory and results

Sensory analysis of descriptive data has to take into account replication and the fact that the assessor by sample interaction is often significant indicating a degree of disagreement among the assessors. This is very effectively done by using a mixed model where assessors are treated as so called random effects.
We show how the mixed effects model is set up and calculated in the specialised Sensory program SENPAQ and the general statistics software XLSTAT.
3 month license for SENPAQ will be given free to each attendee
30 day license for XLSTAT will be given free to each attendee
SENPAQ and XLSTAT workshop will train attendees in the use of both packages
Each attendee will have a 6 month window to purchase both packages at a 30% discount.

16.00 Performance Monitoring – current status and recent developments
Review of SenPAQ, XLSTAT EyeOpenR and Timesens capabilities
Recent literature results comparing performance
Results from Schlich SENSOBASE

Several packages include methodology to assess panel performance. We will review the methodologies and indicate their usefulness and discuss ways of setting standards for panel performance. Dr Pascal Schlich has examined the performance of assessors on his enormous data base of sensory data. We will report some of his conclusions.
Workshop on SENPAQ, XLSTAT and EyeOpenR

17.30 Close
18.00 Optional Course Dinner

Course Outline Day 2

Friday 25 August

09.00   New developments in Difference Testing
Thurstonian methods versus Guessing models.
Replacing the Triangle test: Tetrad, Same Different tests
Equivalence testing developments

Statisticians use guessing models based on the probability of getting a specified difference test correct by chance. Psychophysicists prefer to consider the psychological definition of a perceptual distance based on random variation in perception of two stimuli, or their difference.
We will define the two perspectives and show how to use them to calculate the power of a test.
Recently Ennis and Ennis have proposed that the tetrad test in which assessors have to select two pairs of stimuli from a set of 4 samples is preferable to the triangle test. We will review these papers and explain why we also like the reminder A not A test.
Exercises using VPOWER.

10.30  Recent Developments in Rapid Methods
Sorting, Multiple sorting and Clustered sorting analysis
Global versus partial napping
Results from Delholm thesis
UltraFlash Profiling
Analysis and Confidence intervals
Comparisons with QDA

Napping, Ultra-flash profiling, Sorting and Multiple sorting are rapid methods for assessing perceptual properties of food and beverage stimuli. We will define the methods and report recent extensions.
XLSTAT workshops will give attendees the chance to analyse data from each of the methods

12.00 Lunch
12.45 Projection, PCA Regression, PLS for relating sensory to other data
  PCA projection using supplementary variables
Putting a new sample into the space
PCA regression: Pros and Cons
PLS – definition and Pros and Cons

This talk shows how to project new variables, such as sensory or instrumental measures into a PCA space. It also shows how to project a new point into a PCA point plot. This concept will be extended to more formal modelling of the relation between sensory and instrumental data using either PCA regression or Partial least Squares.
Exercises will use XLSTAT PCA module and the PLS module.

13.30 Path Analysis – use of PCA and PATH PLS
Relations between sensory categories and predicting liking
Introduction to Other techniques – Structural Equation modelling  and Path PLS

The process of sensory perception of a food or beverage product often follows a predetermined sequence, starting with vision, smelling, chewing swallowing and aftertaste. The perceptions from one modality often influence those of the next and there is interest in modelling these pathways (Path Modelling). The pathway can then be combined with a wider process where sensory pathways lead to emotional and hedonic responses that can be combined with package expectations to influence choice and repurchase. We will show how these processes can be modelled using a combination of PCA and rotation and XLSTAT path Modelling software PATH PLS.
XLSTAT PATH PLS exercises.

16.00   Close

About the Course Venue:

Training Venue
New Horizons of Rhode Island
24 Albion Rd
Lincoln, Rhode Island  02865
(9 miles from Pangborn Conference venue)

The nearest hotel to the New Horizon Training Center is The Marriot Hotel


Yes: Please enrol me in Hands On Advanced Sensory Statistics Course Rhode Island 24 -25 August 2017

Registration fees:
Include 2 lunches, course dinner, course materials    $1400
XLSTAT Sensory module Contains BASE + SENSORY module Special offer 1 year license at 70% of full price ($580) $400
SenPAQ Software perpetual licence at 70% of full price ($750)  $520
Unfortunately there is no virtual training option available for this course

You will be provided with your own computer at the venue but you are welcome to bring your own laptop with you if you wish and we will give you instructions to preload a trial version of XLSTAT and SenPAQ onto your computers before the course.

Discounts: We offer a 10 % discount on registrations when two or more people from the same company register for the same course, at the same time.

Registration Policy:  Registration is not final until payment is received. Unpaid spaces will be opened to new registrants 30 days ahead of courses.

Payment:  Payment may be made in US dollars, Euros or GB pounds via the Worldpay Gateway or into Currency accounts. Contact for routing and IBAN details of the currency account you require.

Refund Policy: Cancellation of registration can be made up to 30 days ahead, and return of payments, minus reasonable administrative expenses, will be made for these cancellations. Cancellations within 15 to 30 days of the course start will receive a credit for a future course. Registrants who fail to attend or cancel less than 15 days prior to the seminar start date are responsible for the entire fee.

Substitution of another person for the same course may be made at any time.


Course Registration Fees



Number of Places

Course Fee


XLSTAT Sensory License $400

SenPAQ Licence $520

Total: $

    Payments and Refund Policy    

For electronic bank transfers: contact for IBAN details
Mailing address for registration and payment:
Dr H J H MacFie
43 Manor Road
Keynsham, Nr Bristol,
BS31 1RB, United Kingdom

Tel/Fax +44(0)1179863590

Electronic registration forms to