You are on page 1of 14

Section : 9

Page : 1 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

LIST OF CONTENTS

SECTION 9

TASTE EDUCATION : FURTHER DEVELOPMENT PAGE

9.1 Levels Tasting: Five Glass Test 2

9.1.1 Introduction 2

9.1.2 Taste Test 2

9.1.3 Method and Materials 2

9.1.4 Taste Test Form 2

9.2 Candidacy for Brewers’ Taste Central Panel 10

9.2.1 Entry Requirements 10

9.2.2 Criteria for Successful Candidacy 10

9.2.3 Statistical Evaluation of Tasters’ Performance 13


Section : 9
Page : 2 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

9. TASTE EDUCATION: FURTHER DEVELOPMENT

9.1 LEVELS TASTING (FIVE GLASS TEST)


9.1.1 Introduction
The objective of the five-glass Levels Tasting exercise is to assess a taster’s
ability to distinguish varying levels/strengths of a doped flavour in a beer, and
rank them accordingly. This provides an ongoing record of individual tasters’
strengths and weaknesses. However, note that It is not designed to assess
overall tasting ability, eg it does not cover all the flavour attributes found in
beer, nor does it assess the ability to differentiate brands, provide flavour
profiles or give a suitable drinkability rating.
It is, however, a very valuable exercise for junior tasters, and for experienced
tasters who may not have been exposed to taste education for a long time.
It should be pointed out to tasters that this is not an examination. Remember
that everyone’s individual tasting abilities differ. It is therefore not a means of
accepting or rejecting tasters in any way.

9.1.2 Target
Divisional Levels Tasting:
* All Brewers’ Taste (Central Panel) members.
* All tasters who take part in Regional and/or Central Office tasting
activities.
* Trainee tasters within Central Office or the Regions are also
encouraged to participate.

Note that Regions are encouraged to carry out their own in-house Levels
Tasting at a higher frequency, eg 2 x per annum.

9.1.3 Frequency
Minimum once per annum at Divisional level.

9.1.4 Beer for Doping


Castle Lager in 750 ml bottles, fresh, sampled from the middle of a BBT run. A
selection (eg of five) BBTs is tasted by an experienced panel and the three best
selected beers (free of any obvious taste defects) are then analysed. The beer
whose analytical results most closely meet specification and which has the best
taste rating is used for the doping process. It is essential that sufficient bottles
are collected for doping.

For the purposes of the annual Divisional Levels Tasting exercise, the beer for
doping will be distributed from one source. The selection and distribution of
the beer will be managed by Central Office Tasting Services.
9.1.5 Doping Procedure
The doping procedure as outlined in Section 7.4 is used.

9.1.6 Compounds Used for Doping, and Doping Levels


Section : 9
Page : 3 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

No more than six dopes may be assessed during any one Levels Tasting
exercise. For the purposes of the annual Divisional exercise, Central Office
Tasting Services will advise which six of the dopes in question must be used
for the exercise in question.
The levels for doping are depicted in Table 3. Make sure that all the doped
bottles are clearly marked and kept in the fridge until tasting.
Shelf life of doped samples: 2 weeks.
Where possible, it is recommended that doping recovery studies are carried out
by having the beer analysed for the dopes present.
Section : 9
Page : 4 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

TABLE 3. DOPING LEVELS FOR LEVELS TASTING, FIVE GLASS TEST


NB: The bottles which are “Controls” (no dope added) must also be opened, fobbed and re-
crowned
COMPOUND CONC. OF VOLUME OF THRESHOLD FINAL LEVEL
STOCK STOCK SOLN. LEVEL ADDED PRESENT
SOLUTION ADDED TO 750 ml IN GLASS
(SECTION 7.4) BEER
Sweet 64% w/v 0 0 Level in control beer
0,45 ml ¼ + 0,04 g/100 ml
0,9 ml ½ + 0,08 g/100 ml
1,8 ml 1 + 0,15 g/100 ml
3,5 ml 2 + 0,30 g/100 ml
Bitter 0 0 Level in control beer
(Iso-alpha acid) ¼ + 1,25 mg/l
½ + 2,5 mg/l
1 + 5,0 mg/l
2 + 10,0 mg/l
Acetaldehyde 7 800 ppm 0 0 Level in control beer
0,18 ml ¼ + 1,9 mg/l
0,35 ml ½ + 3,8 mg/l
0,70 ml 1 + 7,5 mg/l
1,4 ml 2 + 15,0 mg/l
Dimethyl 2 000 ppm 0 0 Level in control beer
sulphide (DMS) 0,005 ml (5 µl) ¼ + 12,5 µg/l
0,01 ml (10 µl) ½ + 25 µg/l
0,02 ml (20 µl) 1 + 50 µg/l
0,04 ml (40 µl) 2 + 100 µg/l
Diacetyl 100 ppm 0 0 Level in control beer
0,02 ml (20 µl) ¼ + 2,5 µg/l
0,04 ml (40 µl) ½ + 5,0 µg/l
0,08 ml (80 µl) 1 + 10,0 µg/l
0,16 ml (160 µl) 2 + 20,0 µg/l
Ethyl acetate 9 000 ppm 0 0 Level in control beer
0,23 ml ¼ + 2,5 mg/l
0,45 ml ½ + 5,0 mg/l
0,9 ml 1 + 10,0 mg/l
1,7 ml 2 + 20 mg/l
Phenolic 400 ppm 0 0 Level in control beer
0,2 ml ¼ + 1 µg/l
0,4 ml ½ + 2 µg/l
0,8 ml 1 + 4 µg/l
1,6 ml 2 + 8 µg/l
Metallic 803 ppm 0 0 Level in control beer
0,025 ml (25 µl) ¼ + 0,025 mg/l
0,05 ml (50 µl) ½ + 0,05 mg/l
0,1 ml (100 µl) 1 + 0,1 mg/l
0,2 ml (200 µl) 2 + 0,2 mg/l
Iso-amyl acetate 750 ppm 0 0 Level in control beer
0,4 ml ¼ + 0,4 mg/l
0,8 ml ½ + 0,8 mg/l
1,5 ml 1 + 1,5 mg/l
3,0 ml 2 + 3,0 mg/l
Section : 9
Page : 5 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

9.1.7 Sample Presentation

Each dope is presented in a set of six glasses. The contents of the glasses are as
follows:

GLASS CONTENTS MARKING OF GLASS


1 Control (nothing added) Random three-digit code
2 Control + ¼ Threshold added Random three-digit code
3 Control + ½ Threshold added Random three-digit code
4 Control + 1 Threshold added Random three-digit code
5 Control + 2 Threshold added Random three-digit code
6 Control + 1 Threshold added (same “T”
as Glass 4)

Make a note of the three-digit codes and their contents for each dope.
In order to conserve sufficient doped samples, pour each sample out of only one
bottle.
Immediately replace the opened bottle in the fridge until required for the next
taster. Do not use opened bottles for more than one taste session.
Dopes which are subject to carry-over (e.g. bitter, metallic) should be served
last.

9.1.8 Taste Sessions


As many tasters as can be accommodated should be asked to attend each session
in order to conserve doped sample.
Not all six of the doped series need to be presented at one session, as this
constitutes 36 glasses of beer. Tasters may, if they wish, complete the exercise
during a second or third session depending on an individual’s fatigue threshold.
Do not allow tasters to confer with each other or “crib” from the booth next door.
A taste co-ordinator must therefore be present in the taste room at all times
during the session.
Tasters should be allowed to taste at their own pace. All the tasters need not
taste the same dope at the same time.
Each taster is provided with a set of six glasses for each dope (9.1.7). He is told
the dope up-front. The doping threshold levels may also be divulged if required.
The taster is then asked to rank the five coded glasses according to the strength of
the dope, using the glass marked with a “T” (which is doped at control plus 1 x
threshold) as a reference.
When he has finished his ranking, he writes the glass codes in the spaces
provided on the result sheet from “lowest level” on the left, to “highest level” on
the right. An example of the result sheet is appended at the end of this section
(9.1.11).
9.1.8 Taste Sessions
The co-ordinator then tells the taster immediately whether his ranking was
correct (see 9.1.9 below). If it was not, the taster has the option to repeat the
exercise on a fresh set of glasses. This can be done immediately, later on in the
Section : 9
Page : 6 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

session, or during a subsequent session. Do not divulge the glass codes to the
taster, as he will be presented with the same codes again.
It is not recommended that a taster tries a third time. Results from previous
exercises has shown that tasters who cannot rank a dope correctly on the second
try, will seldom get it correct on the third try. In any case there may be
insufficient doped samples available to allow tasters three attempts.

9.1.9 Interpretation of Results


All five glasses ranked in the correct order: Correct
Any one adjacent pair in the ranking swapped around
(eg Control and ¼ strength swapped around and ranked
as ¼ and then Control): Correct

More than one pair swapped around, or any other


arrangement: Incorrect

9.1.10 Reporting of Results


Each taster is notified in writing of his/her individual results for their personal
records.
Tasters may compare results with each other if they so wish.
A summary of all the tasters’ results is submitted to Central Office Tasting
Services for the purposes of updating the Divisional Taster Database.
Section : 9
Page : 7 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

9.1.11 Levels Tasting Form


LEVELS TASTING

Name ______________________________ Date ___________________________

Brewery/Region _____________________

For each dope you are provided with 5 coded glasses, and one marked with a “T”. Rank the 5 coded
glasses from lowest to highest level of dope added, using “T” as a reference. (“T” = control (base
beer) + 1 x threshold of the dope in question).

Write the glass codes in the spaces provided from “lowest level” on the left to “highest level” on the
right.

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code

Dope: ______________________________

1 (Lowest) 2 3 4 5 (Highest)
Glass Code
Section : 9
Page : 8 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

9.2 CANDIDACY FOR BREWERS’ TASTE PANEL


9.2.1 Entry Requirements
The following are the recommended entry requirements for tasters who wish to
be considered for candidacy on the Brewers’ taste Panel:
a) Successful completion of Taste Education; Advanced (Section 8).
b) Regular (at least weekly) participation in in-process and other Regional
tasting activities.
c) A good overall knowledge of the brewing process, including the relationship
between analytical and process parameters and flavour attributes.
d) The ability to accurately assess the flavour profile of a beer.
e) A good knowledge of the methodology of profile taste tests and the
Brewers’ Taste system (including correct use of the descriptors, scales and
intensity ratings).
f) Personal commitment to and interest in tasting activities.
Due to the level and quality of commitment required, it is recommended that
tasting is included as one of the individual’s formalised goals for performance
review purposes.
Prospective candidates who meet the above criteria must:
a) Be proposed for candidacy by a member of the Brewers’ Taste Panel who is
familiar with his/her tasting capabilities, who will submit the proposal to the
Brewers’ Taste Committee via their Regional representative.
b) Obtain prior Management approval to sit as a candidate for a period of up to
six consecutive months.

and

c) Obtain prior Management approval, in the event of his/her candidacy being


successful, for participation in Brewers’ Taste on an ongoing basis, as and
when required (time and cost commitments must be made clear to
Management).

9.2.2 Criteria for Successful Candidacy


The candidate must participate in Brewers’ Taste for a period of up to six
consecutive months (at the Brewers’ Taste Panel Committee’s discretion).
During this time, his/her performance will be assessed by the Brewers’ Taste
Committee according to the following criteria:
a) Statistical evaluation of the candidate’s performance as compared with that
of the panel (Spearman’s Rank Correlation). Refer 9.2.3.

9.2.2 Criteria for Successful Candidacy (Continued)


b) Individual assessment of the candidates’ performance by Brewers’ Taste
Committee members (or any other Panel member appointed by the
Chairman) participating during the same sessions. The standard form can be
found overleaf.
Section : 9
Page : 9 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

c) Candidates’ receptiveness to feedback on his/her performance, and to


coaching, should this be deemed necessary, by a senior taster appointed by
the Committee.
d) Chairman’s assessment of the candidates’ personal style and commitment
during taste sessions.
Candidates’ performance will be evaluated quarterly at the Brewers’ Taste
Committee meetings. The Chairman will communicate the outcome of these
discussions to the candidate/s in writing. Interim verbal communication will
also be made as and when is deemed necessary by the Chairman on behalf of
the Committee.
Section : 9
Page : 10 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

BREWERS’ TASTE (CENTRAL) PANEL


CANDIDATE TASTER ASSESSMENT

DATE: .................................... CANDIDATE: ...............................

ASSESSOR: ............................

1. Is he /she in line with the Panel?

2. Is his/her comments on key issues relevant?

3. Is his/her assessment in line with his/her comments?

4. Does he/she use the scale of descriptors properly?

5. Does he/she over-elaborate?

6. Is he/she a confident taster and does he/she concentrate all the time?

COMMENTS

RATING A B+ B
Section : 9
Page : 11 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

9.2.3 Statistical Evaluation of Tasters’ Performance


The evaluation of any subjective scoring method is always very difficult and
suffers from perception by the panel members that it will force conformity. In the
case of beer taste panel ratings the evaluation method should be aimed at achieving
consistency and not conformity.
What this means in practice is that the magnitude of an individual rating given by
a panel member is unimportant. What is important is that the panel agree, on a
given set of beers, which are the best and which are the worst among the samples.
In order to evaluate a tasters’ performance in relation to the panel, non parametric
statistical procedures are used. These procedures are based on an evaluation of the
correlation of an individual’s ranked been ratings to that of the “true” ranking of
the beer samples.
The problem immediately arises as to what is the “true” ranking of a given set of
beer ratings. This “true” ranking can only be estimated from the average ranking
of a given set of ranked ratings. Kendall’s Co-efficient of Concordance is used in
this case to measure the agreement within a given set of ranked beer ratings.
If the association is found to be statistically significant, then it can be assumed that
the average of a set of beer ratings gives a reasonable estimate of the “true”
rankings.
Thereafter, Spearman’s Rank Correlation Co-efficient, which measures the
agreement of an individual tasters’ ranking with that of the estimated “true” rank
of the set of samples, is used to evaluate an individual tasters’ performance.
Data Collection
The data required to perform the evaluation is obtained from the Brewers’ Taste
(Central Panel) database which is kept at Central Office (Technical Information).
The evaluation on each month’s set of data is carried out electronically using
suitable statistical software. The results are represented graphically by taster and
are updated monthly.
Tasters’ Graphs: Interpretation of Results
The graphs below provide a number of typical examples of graphs generated by
this evaluation method.
The bold line in the centre of the graph represents p = 0,05. All points below the
line indicate that there is a 95 % probability that the individual tasters’ ranking
correlates with that of the panel over the samples set assessed for any particular
month (usually 20 samples).
Points above the bold line indicate that there was no correlation at the 95 %
confidence level.

Tasters’ Graphs Interpretation of Results (Continued)


Historical data has shown that the majority of tasters experience short periods of
inconsistency from time to time. This may be due to ill-health, job changes,
personal stress or lack of concentration. This in itself is not a concern provided
these occurences remain infrequent and correct themselves.
Section : 9
Page : 12 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

However, Tasters who are regularly producing out-of-line results for whatever
reason require to be counselled as their results detract from the performance of the
panel as a whole. If such performance cannot be corrected in the long term the
taster should be removed from the panel.

TASTER 1

Over a five-year period, there were no occasions of this taster’s results not being consistent
with the Panel’s. This shows excellent long-term consistency.

TASTER 3

TASTER 2

This graph depicts the results of a taster who suffers from recurring (in this case naso-
respiratory) health problems, which will affect his ability to provide consistent judgements.
In such a case, this taster would be well advised to discontinue tasting.
Section : 9
Page : 13 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

This taster is generally consistent, but on one occasion his tasting ability was negatively
affected by outside influences. This occurs with nearly all tasters from time to time and is not
a cause for concern provided the frequency of such occurrences remains low.

TASTER 4

This taster has a tendency to be erractic, with sporadic occurrences of results inconsistent with
the panel. This is a cause for concern, and if there is no improvement in the long term, this
taster should be excused from tasting.

TASTER 5

This taster started off by being very inconsistent, but with practice, commitment and
concentration his results improved and are currently very consistent.

TASTER 6
Section : 9
Page : 14 of 14
BEER DIVISION BREWING MANUAL 1st Issue : JUL 1995
Vol XI - TASTING Rev. Date : JUL 1997

This taster’s results, which showed reasonable long-term consistency, became erractic in
recent months. This was found to be due to a job-change with attendant stress, which should
correct itself in the long term provided conditions improve.

You might also like