FSA, Quantitative Evaluation of Alternative Food Signposting Concepts, Report of Findings. 2005
FSA, Quantitative Evaluation of Alternative Food Signposting Concepts, Report of Findings. 2005
FSA, Quantitative Evaluation of Alternative Food Signposting Concepts, Report of Findings. 2005
Report of Findings
Synovate (UK) Ltd Registered number: 3824237 Registered office: 43-45 Portman Square London W1H 6LY
Contents
1. Background ..................................................................................................... 1
2. Objectives........................................................................................................ 2
2.1 Research Objectives ....................................................................................... 2
2.2 Objectives of Signposting Scheme.................................................................. 3
3. Methodology .................................................................................................... 4
3.1 The Interview ................................................................................................... 6
3.2 Pilot Study ....................................................................................................... 7
3.3 Statistical Differences ...................................................................................... 7
3.4 Sample Sizes .................................................................................................. 8
4. Executive Summary......................................................................................... 9
9. Information on Signposting............................................................................ 69
9.1 Proportion of Respondents Who Would Like to Know How Information on
Signposting is Calculated .............................................................................. 69
9.2 Where Respondents Would Like Information to be Shown ........................... 69
10. Food Categories on Which Consumers Would Like To See Signposting ..... 70
10.1 Total Sample ................................................................................................. 70
10.2 Among Respondents with Children Living in the Household and Without
Children Living in the Household................................................................... 71
11. Other Questions Asked Regarding Nutrients, Nutritional Panel and GDA .... 73
11.1 Importance of Nutrients When Considering Whether a Product is Healthy or
Not................................................................................................................. 73
11.2 Frequency of Reading Nutritional Panel on Food Products .......................... 73
11.3 Proportion of Respondents Who Spontaneously Asked What GDA Meant
During Interview ............................................................................................ 75
12. Main Ethnic Minority Consumers................................................................... 76
12.1 Individual Product Evaluations ...................................................................... 76
12.1.1 Understanding ...................................................................................... 76
12.1.2 Time to Interpret ................................................................................... 78
12.1.3 Use of Signposting Information When Answering Questions on
Nutritional Content................................................................................ 79
12.2 Comparison of Two Products ............................................................... 80
12.2.1 Understanding ...................................................................................... 80
12.2.2 Time to Interpret ................................................................................... 81
12.2.3 Use of Signposting Information When Answering Questions on
Nutritional Content................................................................................ 82
12.3 Consumer Preference ................................................................................... 83
14. Conclusions................................................................................................... 92
Following on from this work the current phase of quantitative consumer research was
commissioned to investigate consumers understanding of, and preference for, four
signposting concepts:
With the exception of Simple Traffic Lights, the signposts featured the following four
key nutrients: fat, saturated fat, salt and sugar.
Synovate was commissioned to conduct this quantitative research, and this report
presents the findings.
1
This work complements that being taken forward as part of the Scottish Action Plan Eating for
Health, Meeting the Challenge, the Welsh strategy "Food and Wellbeing" and the proposed Food
and Nutrition Strategy for Northern Ireland.
2
www.food.gov.uk/multimedia/pdfs/alternlab.pdf
3
www.food.gov.uk/multimedia/pdfs/signpostingnavigatorreport.pdf
While the primary objectives of the research were to assess the performance of four
signposting concepts, consumer preference was also explored.
1. Determine how well each of the four alternative signposting concepts being
tested help consumers compared to products without additional signposting
information.
2. Determine to what extent the concepts enable consumers to correctly and
quickly identify whether an individual product is high, medium or low in terms
of fat, saturated fat, salt or sugar.
3. Determine to what extent the concepts enable consumers to make
comparisons between similar products and to quickly distinguish between
them in terms of fat, saturated fat, salt or sugar.
4. Assess how easy and clear each of the signposting concepts being tested are
to understand.
5. Determine to what extent the concepts are applicable across different
consumer demographic groups.
6. Determine to what extent the concepts are applicable across different product
categories.
7. Establish consumer preference for each of the concepts
The main aim of this research was to find out which concept would best meet
requirements 1-6 above.
While this research also provided an indicative guide to requirements 5 and 6, further
qualitative research was commissioned to explore these aspects in more detail. A
report of this research has been published separately.
The decision was taken that the signposting concepts and products without
signposting should be tested in a way that replicated reality as far as possible. It was
therefore decided that the signposting should be tested using real products as
stimulus. Products were purchased during April 2005 from retail outlets in central
London. Life-sized photographs of the products on A3 and A4 card were used as
stimulus. The front of the cards bore a photograph of the front of the product and the
rear side of the cards bore a photograph of the rear part or side of the product that
featured the nutritional panel, allowing respondents to refer to this during the
evaluations.
The signposts on the product photographs were consistent with the nutrient content
of the product in question, and were consistent in size across similar products and
were scaled in proportion to the product size.
The four signposting concepts assessed are shown in Appendix 1. In addition, actual
food packaging with no front of pack signposting was included.
The five alternative formats (four signposting and products without signposting) were
applied to five different product categories (three examples of products per category);
Cereal bars/breakfast cereals
Ready meals
Chicken burgers/pizzas
Pasta ready meals/curry ready meals
Cake/crisps
The categories were chosen in order to ensure that the signposting concepts were
evaluated on a broad range of composite products with differing nutrient levels.
Each product category comprised three examples of similar products, the first
product was used in the Individual Product Evaluations and the remaining pair used
in the Comparisons of Two Products. These assessments are explained in section
3.1.
The four key nutrients included in the evaluations were those that featured on all four
signposting concepts except Simple Traffic Lights, namely: fat, saturated fat, salt and
sugar.
Table 1
Products
Signposting Signposting Signposting Signposting
without
Concept 1 Concept 2 Concept 3 Concept 4
Signposting
Product
Category 1 5 4 3 2
A
Product
Category 2 1 5 4 3
B
Product
Category 3 2 1 5 4
C
Product
Category 4 3 2 1 5
D
Product
Category 5 4 3 2 1
E
In addition, any possible order effect was controlled for by rotating the order in which
the concepts were presented to respondents.
In total, therefore, there were 25 test cells, with a separate quota set for each cell so
as to ensure matched sample profiles in terms of demographics. The overall target
quotas, and those achieved are shown in Table 2 below.
Target Achieved
Gender Male 32% 31%
Female 68% 69%
Age 16-34 40% 42%
35-54 40% 38%
55-70 20% 20%
Working Status Working 46% 55%
Non-working 54% 45%
Social Grade ABC1 54% 55%
C2DE 46% 45%
The data was weighted at the analysis stage according to the profile of the UK
population (target quotas) on the above demographics, as well as on country
(England, Scotland, Wales and Northern Ireland) and ethnic origin (white and non-
white).
Details of the products used in this research, and the nutrients people were asked
about are given in Appendix 2. The numerical nutrient content information was used
to determine whether a respondent answered correctly during both parts of the
performance section of the interview.
Each respondent evaluated each of the signposting concepts and the products
without signposting in the performance section before completing the second and
third sections of the interview.
Prior to the start of fieldwork, a two-stage pilot study was conducted to test the
research design and particularly the effectiveness of the questionnaire. Minor
changes were made to the questionnaire following the pilot.
All significant differences referred to in this report are at the 95% confidence level.
This means that if similar independent samples are drawn from the UK population, a
difference of the same magnitude or greater would be expected 19 times out of 20,
and that therefore there is only a likelihood of 1 in 20 (5%) that the difference
occurred by chance.
Significant differences between the different concepts and the products without
signposting are shown on the charts in this report by use of letters denoting each of
the different concepts and the products without signposting. This is explained in the
footnote 5 accompanying the first paragraph below Figure 1. These are differences
between the columns of data.
Significant differences between the demographic groups for each concept or the
products without signposting on the summary charts (for example Figures 4, 5 and 6)
are shown by the use of green and red rings. These are differences between the
different rows of data.
A green ring on summary charts signifies that the score ringed is significantly higher
than at least one other number of that group, while a number ringed in red is
significantly lower than at least one other number of that group.
The sample sizes in all Figures relating to the performance evaluations (Individual
and Comparison of Pairs of Products), namely 6.1.1 to 6.1.3, 6.2.1 to 6.2.3 and all
Figures in sections 12 and 13 except sections 12.3 and 13.3 are the number of
evaluations. The number of evaluations is (exactly or almost exactly) twice as high as
the number of respondents, as each respondent carried out two evaluations for each
of the concepts and the products without signposting: namely an evaluation of two
different nutrients.
The sample sizes for all other Figures in the report are the number of respondents.
The research assessed how well each signpost format performed when respondents
were asked to assess the level of fat, saturated fat, salt or sugar in an individual
product, or to compare levels of two of these nutrients in a pair of products.
Respondents were also asked whether they would find signposting helpful and which
signposting format they preferred. The reasons behind their preference were
explored.
Over 2600 interviews were conducted across the UK. Across both Individual Product
and Comparison of Two Products Performance Evaluations, Multiple Traffic Lights
and Colour-coded GDA performed strongest of the signposting concepts tested, with
Simple Traffic Lights and Monochrome GDA performing significantly less well. In the
Individual Product Evaluations, the Multiple Traffic Lights option produced the highest
level of correct responses across all age, socio-economic and geographical sub-
groups and main ethnic minority groups.
The difference in the performance of the Colourcoded GDA option and Multiple
Traffic Lights in the Individual Product Evaluations was very marked, with Colour-
coded GDA performing significantly less well than Multiple Traffic Lights by a margin
of 21 percentage points. The difference in the level of correct responses, when
comparing two products, between Colour-coded GDA and Multiple Traffic Lights was
smaller at 6 percentage points. On balance, the Multiple Traffic Lights signpost was
therefore considered to most closely meet the objectives of the initiative. It helped
people identify whether an individual product was high or low in a particular nutrient
correctly and quickly, and to compare nutrient levels in pairs of products. The
majority of people found it easy to use and understand, and a wide range of
consumers was able to use it correctly including key demographic sub-groups.
The overwhelming majority of respondents (96%) said they thought front of pack
signposting would be useful in helping them make healthier food choices. A similarly
high proportion of people interviewed said they would prefer a colour-coded
individual nutrient signpost, such as Multiple Traffic Lights or Colour-coded GDA.
Colour-coded GDA was the most favoured signpost.
Overall, across both performance tests, Multiple Traffic Lights and Colour-coded
GDA performed better than Simple Traffic Lights, Monochrome GDA and the
products without signposting.
It is clear that the colour-coding was a major contributing factor to Multiple Traffic
Lights and Colour-coded GDAs strong performance in the Individual Product
Evaluations, as well as for Multiple Traffic Lights in the Comparison of Two
Products.
With respect to this test, Multiple Traffic Lights performed best, both in terms of
eliciting the highest level of correct responses and the quickest interpretation of
the information. Colour-coded GDA performed second best, producing a
reasonable level of correct answers.
When asked to say whether the nutrition content of the various products was high,
medium or low, Multiple Traffic Lights elicited the highest level of correct answers. On
average across all nutrients and product categories, from the total number of
evaluations, a high proportion of responses given were correct (79%), when
answering about the products bearing the Multiple Traffic Lights signposting.
When the Multiple Traffic Lights signposting4 on the products was claimed to have
been used when answering the questions, a very high proportion, 90%, answered
correctly regarding the nutrition content of the products in the evaluations.
These levels of correct response were significantly higher than those for the other
three concepts and the products without signposting.
4
After the respondent stated whether they thought a product was high, medium or low in a
specified nutrient, he or she was asked (without being prompted) what information, if any,
they were using when giving their answer. From this, it was ascertained which
respondents claimed to use signposting.
The proportion of correct answers for sugar and saturated fat was significantly higher
than for fat and salt for all four signposting routes.
Multiple Traffic Lights produced a high level of correct responses across fat,
saturated fat, sugar and salt.
Colour-coded GDA elicited a significantly lower level of correct responses for fat than
the other three nutrients: the level of correct response for Colour-coded GDA was
52% (on average across all product categories). This is attributable to a poor correct
response score for ready meals on fat of 41%.
Multiple Traffic Lights produced a similar high level of correct responses across each
of the five categories of cereal bars, ready meals, chicken burgers, pasta meals and
cakes.
Colour-coded GDA elicited a significantly lower level of correct responses for ready
meals than the other four categories: the level of correct response for Colour-coded
GDA was 55% (on average across all product categories). This is attributable to a
poor correct response score on fat for ready meals of 41% cited above.
Multiple Traffic Lights produced a similar high level of correct responses across all
demographic groups, except amongst consumers aged 55-70 years. Among this
group the level of correct responses was 72%, the lowest level of correct responses
of any demographic subgroup with Multiple Traffic Lights. This is nevertheless a
relatively high score, demonstrating that Multiple Traffic Lights produces a high level
of correct response among all subgroups.
Levels of correct responses among 55-70 year-olds (of the total number of
evaluations) were lower for all signposts than among at least one younger age group.
However, for Colour-coded GDA the level of correct response on average across all
nutrients and product categories was 57%. This score was significantly lower than
the score of 70% among 16-34 year-olds and also 66% among 35-54 year-olds. The
differences in the correct proportions between these age groups when using Colour-
coded GDA was significantly larger than that when using other signposting concepts
or the products without signposting.
Respondents from socio-economic groups C2DE also gave a significantly lower level
of correct response than those from ABC1 groups when using products without
signposting and all signpost formats except Simple Traffic Lights. However, the
With respect to the speed in which the information on the signposting was
interpreted, responses were given most quickly when using Multiple Traffic Lights.
Responses were given in 5.1 seconds on average across all nutrients and product
categories. This was significantly faster than the time taken to interpret and answer
using the other three concepts and products without signposting.
The time to interpret and answer for Colour-coded GDA was also fast: 5.4 seconds
on average across all nutrients and product categories. This was significantly faster
than when Monochrome GDA was used (6.7 seconds), implying that the colour is
helpful to consumers in enabling them to quickly evaluate a product.
When the signposting was claimed to have been used, the time to interpret was even
faster for Multiple Traffic Lights, 4.2 seconds on average, significantly faster than for
all other signposts. Colour-coded GDA elicited the second-fastest average time of 5
seconds.
Consumers were significantly more likely to claim they used signposting when shown
a product with Colour-coded GDA than with the other three signposting concepts
when answering the questions about nutrient content. Interestingly however, the
level of correct responses with Colour-coded GDA was significantly lower than with
Multiple Traffic Lights, indicating that people were less able to use the signpost
appropriately. In 69% of evaluations with products bearing the Colour-coded GDA
signposting, respondents claimed to have used the signposting (on average across
all nutrients and product categories), compared to 63% for Multiple Traffic Lights and
60% for Monochrome GDA. The Simple Traffic Lights signposting was claimed to
have been used in a relatively small proportion of evaluations, 21%.
With respect to this test, Colour-coded GDA performed best, both in terms of
eliciting the highest level of correct responses and the quickest interpretation of
the information. However Monochrome GDA and Multiple Traffic Lights also
elicited a high level of correct response.
When asked to say which product of various pairs of products contained the most of
a specific nutrient or whether they contained the same amount, Colour-coded GDA
elicited the highest level of correct answers. On average across all nutrients and
product categories, from the total number of evaluations, a very high proportion of
responses given were correct (88%), when answering questions about the products
bearing the Colour-coded GDA signposting.
When the Colour-coded GDA signposting on the products was claimed to have been
used when answering the questions, a very high proportion of correct responses,
92%, was given regarding the nutritional content of the products.
These levels of correct response were significantly higher than those for the other
three concepts and the products without signposting.
However Multiple Traffic Lights and Monochrome GDA also performed strongly.
Monochrome GDA elicited a level of response of 85% among the total number of
evaluations and in 88% of the evaluations when the signposting was claimed to have
been used. Multiple Traffic Lights produced a correctness level of 83% among the
total number of evaluations and 86% of those who claimed to have used its
signposting. There was no significant difference between these scores for these two
concepts. Although the difference in the level of correct responses between Colour-
coded GDA and Monochrome GDA and Multiple Traffic Lights is statistically
significant, in numerical terms it is small suggesting that while numbers on the front
of pack helped respondents, they were only slightly more useful than the colour
coded information on Multiple Traffic Lights. Furthermore, among people who said
they used signposting to respond, while the difference in the level of correct
responses between Colour-coded GDA and Multiple Traffic Lights is 6 percentage
points when comparing two products, in the individual product evaluations Multiple
Traffic Lights performed better by a margin of 21 percentage points.
Colour-coded GDA produced a similar high level of correct responses across fat,
saturated fat, sugar and salt (on average across all nutrients and product categories).
The levels of correct responses in respect of Multiple Traffic Lights for salt and sugar
were not significantly lower than those for Colour-coded GDA, although they were for
fat and saturated fat.
Colour-coded GDA produced a similar high level of correct responses across each of
the five categories of breakfast cereals, ready meals, pizzas, curry meals and crisps.
The levels of correct identification for Multiple Traffic Lights for breakfast cereals,
pizzas and curry meals were not significantly different to those for Colour-coded
GDA.
The correct response levels for Multiple Traffic Lights for crisps and ready meals
were significantly lower than those for Colour-coded GDA. The level of correct
response for crisps for Multiple Traffic Lights of 74% was significantly lower than that
of Colour-coded GDA, 86%, and the average across all nutrients for Multiple Traffic
Lights, 83%.
As described in the previous section entitled Across Each of the Four Nutrients, this
significantly lower level of correct responses for fat when using Multiple Traffic Lights
is almost solely attributable to the lower correctness scores for crisps.
With respect to the speed in which the information on the signposting was
interpreted, responses were given most quickly with Colour-coded GDA. Responses
were given in 5.6 seconds on average across all nutrients and product categories.
This was significantly faster than the time taken to interpret and answer using
Monochrome GDA and Multiple Traffic Lights: 6.0 and 6.3 seconds respectively.
When the signposting was claimed to have been used, the time to interpret was even
faster for Colour-coded GDA, 5.1 seconds on average, significantly faster than for all
other signposts. Multiple Traffic Lights and Monochrome GDA elicited the next fastest
average times of 5.5 and 5.6 seconds respectively.
As noted in the Individual Product Evaluations, the signposting was claimed to have
been used in significantly more evaluations with Colour-coded GDA than with the
other three concepts when answering the questions about nutrient content. In 80% of
evaluations respondents claimed to have used the signposting with Colour-coded
GDA (on average across all nutrients and product categories), compared to 76% for
Monochrome GDA and 67% for Multiple Traffic Lights. The proportion of evaluations
in which the signposting of the two GDA concepts were claimed to have been used
when comparing products was significantly higher than that for the Individual Product
Evaluations.
In 28% of evaluations respondents said they used the nutritional panel with Multiple
Traffic Lights. In significantly fewer evaluations respondents said they used it with the
GDA concepts: 13% of evaluations with Colour-coded GDA and 16% of evaluations
with Monochrome GDA. These proportions where the nutritional panel was claimed
to have been used with the GDA concepts were significantly lower than those for the
Individual Product Evaluations.
Those who preferred Colour-coded GDA said they did so mainly because it included
colour coding, as well as exact numbers, amounts and detail, while those who
preferred Multiple Traffic Lights did so because they felt that it was easy and quick to
read, understand and use.
A very large proportion, 96%, said they thought front of pack signposting would be
useful. Of these, 70% said they thought it would be very useful and 26% said it would
be quite useful. 2% said it would not be very useful. No respondents said it would not
be at all useful and 2% had no opinion.
Of all twenty three food categories shown on a list to respondents (copy appended in
Appendix 4), the one on which consumers would most like to see signposting was
chicken burgers, sausages, nuggets and fish fingers (84%).
The highest scores were for meal centre components (chicken burgers, sausages,
nuggets and fish fingers (84%)), chilled and frozen ready meals (83%), breakfast
cereals (83%), pizzas (82%), snacks (cakes and biscuits (82%), crisps (80%)) and
confectionery (chocolate and sweets (78%)).
Twenty nine per cent said they would like signposting on all food and drinks products.
Among the main ethnic minority groups, the findings were consistent with those of
the total sample on the key results of the Individual Product Evaluations:
The findings were also consistent in the Comparison of Two Similar Products
evaluations:
The pattern of the level of correct response in the Comparison of Two Similar
Products was similar among respondents of main ethnic minority groups to
that of the total sample, with Colour-coded GDA eliciting the highest level of
correct response, followed by Monochrome GDA and Multiple Traffic Lights.
However the levels of response among main ethnic minority groups was
However unlike across the total sample, among main ethnic minority groups
there was no significant difference in the time taken to respond for Colour-
coded GDA and Multiple Traffic Lights.
The proportion of respondents from main ethnic minority groups preferring Colour-
coded GDA was lower than among the total sample, but this signpost was still
significantly preferred over the other signposts.
Among these consumers, the findings were consistent with that of the total sample
on the key results of the Individual Product Evaluations:
However, the level of correct response for all of the signposts and products
without signposting was generally lower among C2 and DE respondents
compared to ABC1 groups.
The findings were also similar in the Comparison of Two Similar Products
evaluations:
Colour-coded GDA performed best in terms of the correct response level among
DE socio-economic group consumers.
Again the level of correct response for all the signposts and products without
signposting was generally lower among C2 and DE respondents compared with
ABC1.
For each of the signposting concepts and products without signposting respondents
performed two types of evaluations. The first set of evaluations comprised assessing
the level of nutrient content of five individual products. The second set of evaluations
comprised assessing which of two products contained the higher level of nutrient
content for five pairs of products.
6.1.1 Understanding
The green bars in Figure 1 show the proportion of correct answers across all
evaluations (among all respondents) across all four nutrients and five product
categories when asked whether a product was high, medium or low in the specified
nutrients.
Figure 1
Understanding Proportion of Correct Responses Whether Product
is High, Medium or Low - Across All Nutrients and Categories
ACDE INDIVIDUAL
79%
TOTAL SAMPLE
ADE
66%
63% 62%
57%
AE
Across 43%
all 37% 38%
nutrients 34%
Correct
21% response
Incorrect
response
The proportion of correct responses given for Multiple Traffic Lights was high, whilst
with Colour-coded GDA the level was reasonable and with Monochrome GDA,
products without signposting and Simple Traffic Lights it was poor.
5
Each of the concepts/No Signposting route is labelled with a letter (A, B, C, D or E) below
their label. A letter or letters above a concepts/No Signpostings green (correct) bar on
the chart signifies that that the score for that concept/No Signposting route is significantly
higher (at the 95% confidence or higher) than those concepts corresponding to the
letter(s) in question.
Figure 2
58%
Across
all A
nutrients 42%
Correct
31% 31% response
Incorrect
response
10%
Those respondents who claimed to have used the signposting with Multiple Traffic
Lights and Colour-coded GDA were significantly more likely to give a correct
response than the total sample. This indicates that these signpost concepts were
helpful in informing consumers as to whether products were healthier or less healthy
options in the Individual Product Evaluations.
However, the level of correct responses with Simple Traffic Lights was significantly
lower than that of the total sample, implying that the signposting was not helpful in
informing consumers as to whether products are healthier or less healthy options.
Figure 3 shows the proportion of correct responses for each of the four signposting
routes among the key demographic subgroups.
The significant differences between the level of correct response of the four
signposting concepts among the total number of evaluations when signposting was
claimed to have been used (the first row in the table on the Figure) are shown in
letters above the five columns (as shown in Figure 2). Any variation from these
6
After the respondent stated whether they thought a product was high, medium or low in a
specified nutrient, he or she was asked (without being prompted) what information, if any,
they were using when giving their answer. From this it was ascertained which
respondents claimed to use signposting.
The rings around the numbers in Figure 3 and all other similar summary charts (e.g.
4, 5 and 6) in this report show significant differences between different subgroups. A
green ring signifies that the score ringed is significantly higher than at least one other
number of that group, while a number ringed in red is significantly lower than at least
one other number of that group.
It can be seen that the pattern of response among each of the subgroups is very
similar to that among the total sample, with Multiple Traffic Lights performing best
among all subgroups and Colour-coded GDA performing second best.
Figure 3
Figures 4, 5 and 6 show the proportion of correct responses for each of the four
signposting routes and products without signposting among all of the demographic,
attitudinal and user subgroups.
Figure 4
Figure 6
Need to be careful about sat. fat in diet 35% 80% 65% 43% 35%
Need to be careful about salt in diet 36% 79% 63% 40% NS 37%
Need to be careful about sugar in diet 35% 79% 61% 41% NS 35%
Need to be careful about calories in diet 27% 78% 65% 38% 33%
It is also evident that Multiple Traffic Lights produced a relatively similar high level of
correct responses across all demographic groups, except amongst consumers aged
55-70 years, among whom the correct response level was 72% of the total sample
(on average across all nutrients and product categories). This score was significantly
lower than the score of 81% of evaluations among 16-34 year-olds and 80% among
35-54 year-olds. This proportion of 72% was the lowest level of correct responses of
any demographic subgroup, but nevertheless this score is high, demonstrating that
Multiple Traffic Lights produces a high level of correct response among all
subgroups.
Levels of correct response among 55-70 year-olds were lower for all signposts than
among at least one younger age group. However Colour-coded GDA achieved a
correct response of 57%. This score was significantly lower than the score of 70%
among 16-34 year-olds and 66% among 35-54 year-olds. The differences in the
correct proportions between these age groups when using Colour-coded GDA was
significantly larger than that when using other signposting concepts or the products
without signposting.
From Figure 5 it can be seen that in Scotland the level of the correct identification
when using Colour-coded GDA was significantly lower than in England, Wales and
Northern Ireland. Respondents from Scotland, as well as Northern Ireland, also gave
a significantly lower level of correct responses than those from England when using
Multiple Traffic Lights. However, the difference for Colour-coded GDA was
significantly higher than for Multiple Traffic Lights.
In Figure 6 it is evident that the level of correct responses for Multiple Traffic Lights
and Colour-coded GDA was significantly higher amongst those who were observed
by interviewers to have used the nutritional panel as well as those claiming to use the
signposting. In addition, those claiming to refer to the nutritional panel when
evaluating the products without signposting gave a significantly higher correct
response than those who did not refer to the panel.
Figure 7 shows the proportion of correct responses by each of the four nutrients (as
well as for all nutrients combined) across all categories when using each of the four
signposting concepts and the products without signposting.
Figure 7
Across all four signpost routes and the products without signposting, the proportion of
correct answers for sugar and saturated fat was significantly higher than the other
nutrients.
Multiple Traffic Lights produced a similar high level of correct responses across fat,
saturated fat and salt.
Colour-coded GDA elicited a significantly lower level of correct responses for fat than
the other three nutrients. Fifty two per cent gave a correct response for this nutrient
(on average across all product categories). As described in section 6.1.1.3, this is
attributable to a poor correct response score for ready meals on fat of 41%.
Figure 8
INDIVIDUAL
Understanding Proportion of Correct Responses Whether USED SIGN-
Product is High, Medium or Low - Summary by Nutrient POSTING VS
NOT
ACD AD A
Significant differences same as those for
All Nutrients (above columns). Any Simple Multiple Colour- Mono-
variation from this shown in red. NS = no Traffic Traffic coded chrome
significant differences Light Light GDA GDA
ALL NUTRIENTS 37% 79% 66% 43%
Used Signposting 31% 90% 69% 42%
Did not use Signposting 36% 59% 50% 38%
FAT
Used Signposting 11% 91% 58% 29%
Did not use Signposting 28% 48% 31% NS 32%
SATURATED FAT
Used Signposting 50% 88% 78% 62%
Did not use Signposting 50% 67% 60% 51%
SALT
Used Signposting 25% 89% 66% 24% NS
Did not use Signposting 24% 58% AD 54% 28%
SUGAR
Used Signposting 64% 97% 84% 64% NS
Did not use Signposting 50% 71% 57% 50%
It can be seen that when Multiple Traffic Lights and Colour-coded GDA signposting
were claimed to have been used, a significantly higher level of correct response was
given for each of the different nutrients than when respondents said they did not use
the signposting.
When Monochrome GDA signposting was claimed to have been used, the level of
correct identification was significantly higher for saturated fat and sugar, but there
was no significant difference for fat and salt. The lack of a significant difference for fat
is due to the particularly low correctness level for ready meals, 19% (as shown on
Figure 10). It would seem that this low correctness score is probably due to
respondents incorrectly supposing that the ready meal in question a Chilli Beef
Ready Meal - is not low in fat (which it is), based on their general assumptions about
these types of food. This seems to indicate that with Monochrome GDA, respondents
are significantly more likely than with Multiple Traffic Lights or Colour-coded GDA to
not be able to use the signposting information correctly. It would seem therefore that
the colour-coding on Multiple Traffic Lights and Colour-coded GDA is helpful to
consumers and that consumers are able to use these.
When Simple Traffic Lights signposting was claimed to have been used, the level of
correct responses was significantly higher for sugar when signposting was used than
when it was not.
Figure 9 shows the proportion of correct responses about the nutrient content of a
product by category.
Figure 9
All
37% 79% 66% 43% 38%
Categories
Cereal bar 40% 85% 71% 41% E 35%
Ready meal 27% 79% 55% 34% 31%
Burger 58% 74% 65% 43% 40%
Pasta meal 45% 79% 71% 49% 45%
Cake 33% 78% 67% 46% 39%
It can be seen that Multiple Traffic Lights produced a similar high level of correct
responses across each of the five categories of cereal bars, ready meals, chicken
burgers, pasta meals and cakes, although the correctness score for chicken burger
was significantly lower than that for cereal bar, ready meal and pasta meal. Its level
of correct response was significantly higher on each of the categories than the three
other signposting concepts and products without signposting.
Monochrome GDA also elicited a significantly lower level of correct responses for
ready meals than the other four categories.
Simple Traffic Lights elicited a significantly higher level of correct response on cereal
bars, chicken burgers and pasta meals than on cakes and ready meals.
Figure 10 shows the level of correct responses for each of the different product
categories by nutrient.
Figure 10
SUGAR
Cereal 53% E 91% 81% 60% E 44%
With the exception of all products in relation to saturated fat, Multiple Traffic Lights
produced a level of correct responses that was significantly higher on each of the
nutrients by product than the three other signposting concepts and products without
signposting.
Interviewers timed how long respondents took to give their response to each of the
questions they were asked about the level of nutrient content of a product.
Figure 11 shows the average time in seconds taken by respondents to give the
response for each of the signposting concepts and the products without signposting
across all nutrients and categories.
Figure 11
BCD BCD
BC
B
8.2 8.5
Across 6.7
all 5.1 5.4
nutrients
The time to interpret and answer for Colour-coded GDA was also relatively fast, with
a time significantly faster than when Monochrome GDA was used (as well as than
Simple Traffic Lights and products without signposting), implying that the colour is
helpful to consumers when evaluating a product.
Figures 12, 13 and 14 show the average time taken by the different demographic,
attitudinal and user subgroups.
Multiple Traffic Light elicited responses that were significantly faster than
Monochrome GDA, Simple Traffic Lights and the products without signposting across
all demographic subgroups, and significantly faster than Colour-coded GDA across
about half of the subgroups.
When the signposting was claimed to have been used, the time to interpret was even
faster for Multiple Traffic Lights, 4.2 seconds on average, significantly faster than for
all other signposts and products without signposting. Colour-coded GDA elicited the
second-fastest average time of 5 seconds.
Figure 12
INDIVIDUAL
Time To Interpret (in seconds) Across All Nutrients and BY
Categories Summary by Subgroup (1 of 3) SUBGROUPS
Figure 14
INDIVIDUAL
Time To Interpret (in seconds) Across All Nutrients and BY
Categories Summary by Subgroup (3 of 3) SUBGROUPS
Need to be careful about sat. fat in diet 8.1 5.4 5.4 NS 6.4 8.0
Need to be careful about salt in diet 8.4 5.5 6.6 7.3 B 8.8
Need to be careful about sugar in diet 8.2 5.3 6.2 6.5 8.2
Need to be careful about calories in diet 7.6 5.0 5.1 NS 5.7 NS 8.2
6.1.2.2 By Nutrient
Figure 15 shows the time taken to respond by nutrient, using each of the signposting
concepts and the products without signposting.
Figure 15
BCD BC
B BCD
Saturated
7.7 7.7
Fat 5.0 5.3
6.5
BCD BC BCD
Salt 8.7 9.1
6.3
5.0 5.1
BCD
BC BCD Base: Number of Evaluations: Fat: Simple Traffic
Light 1066, Multiple Traffic Light 1074, Colour GDA
Sugar 10.6 1070, Monochrome GDA 1075, No signposting
9.8 1069; Sat Fat: Simple Traffic Light 2137, Multiple
7.2
5.0 5.4 Traffic Light 2144, Colour GDA 2148, Monochrome
GDA 2140, No signposting 2137; Salt: Simple
Traffic Light 1613, Multiple Traffic Light 1597,
Simple Multiple Colour Monochr Colour GDA 1613, Monochrome GDA 1603, No
No Signpost
Traffic Traffic GDA GDA signposting 1600: Sugar: Simple Traffic Light 542,
Multiple Traffic Light 527, Colour GDA 535,
A B C D E Monochrome GDA 538, No signposting 532; Q2a
The time to respond when using Multiple Traffic Lights signposting was significantly
faster than the other three signposting concepts and the products without signposting
across all four nutrients, except on salt and sugar, for which the times were not
significantly faster than Colour-coded GDA.
After the respondent stated whether they thought a product was high, medium or low
in a specified nutrient, he or she was asked (without being prompted) what
information, if any, they were using when giving their answer. Figure 16 shows the
proportion of evaluations where respondents claimed to use different sources of
information for each of the signposting concepts and products without signposting
when answering the nutrient content questions.
Figure 16
69%
63% 62%
60%
57%
26% 26%
21%
18% 16%
13%
9% 8%
6% 6% 6% 4% 6% 6% 6%
3% 5% 3% 2% 3% 5%
2% 2% 2%
0%
N/A
Signposting Nutritional panel Common Ingredients list I just know about Didn't use
sense/general nutrition of this anything
knowledge type of product
Base: Total Number of Evaluations: Simple Traffic Light 5358,
Multiple Traffic Light 5342, Colour GDA 5366, Monochrome GDA
5356, No signposting 5338; Q2b
It can be seen that of all sources of information, signposting was claimed to have
been used more than any other information when using Colour-coded GDA, Multiple
Traffic Lights and Monochrome GDA. The differences in the use of signposting with
each of the different concepts and with the products without signposting are
addressed in greater detail in section 6.1.3.2.
In evaluations when using Simple Traffic Lights, the nutritional panel was claimed to
have been used significantly more than the signposting, with the signposting claimed
to have been used in only a small proportion of evaluations.
In a relatively large proportion of evaluations, respondents said they had used the
nutritional panel with Multiple Traffic Lights and Monochrome GDA when answering
the nutrient questions, while the Colour-coded GDA was claimed to have been used
by a significantly smaller proportion.
Common sense/general knowledge and the ingredients list were claimed to have
been used in a small proportion of evaluations when using each of the signposting
Figure 17 shows the first set of columns from Figure 16, which indicate the extent to
which signposting was claimed to have been used when answering questions about
nutrient content across all nutrients and product categories.
Figure 17
Across 40%
37%
all 31%
nutrients
21%
Used
Didnt use
A B C D
Figures 18, 19 and 20 show the extent to which respondents claimed to use
signposting on each of the four different signposting concepts by the different
demographic, attitudinal and user subgroups.
Figure 18
Whether Referred to Signposting Across All Nutrients and INDIVIDUAL
-TOTAL
Categories Summary by Subgroup (1 of 3) SAMPLE
AD ABD A
Significant differences same as those for
Total Number of Evaluations (above Simple Multiple Colour- Mono-
columns). Any variation from this shown Traffic Traffic coded chrome
in red. NS = no significant differences Light Light GDA GDA
Total Number of Evaluations 21% 63% 69% 60%
First Seen 22% 78% 75% A 74%
Figure 20
Need to be careful about sat. fat in diet 27% 64% 70% 63%
Need to be careful about salt in diet 21% 61% A 63% A 63%
Need to be careful about sugar in diet 24% 61% A 67% 59%
Need to be careful about calories in diet 27% 62% A 69% A 63%
Claimed use of Colour-coded GDA signposting was significantly higher than that of
Multiple Traffic Lights and Monochrome GDA signposting among all age groups, both
men and women, among both those working and not working and among both ABC1
and C2DE socio-economic groups.
Among the following subgroups, there was no significant difference between the
proportion claiming to have used Colour-coded GDA signposting, compared to either
Multiple Traffic Lights or/and Monochrome GDA signposting:
When the concept was seen first (not significantly higher than both Multiple
Traffic Lights and Monochrome GDA).
Among the AB socio-economic group (not significantly higher than Multiple
Traffic Lights).
Among non-white respondents (not significantly higher than both Multiple
Traffic Lights and Monochrome GDA).
In Scotland and Wales (not significantly higher than Multiple Traffic Lights).
It can be seen from Figure 18 that the Multiple Traffic Lights, Colour-coded GDA and
Monochrome GDA signposting were claimed to have been used significantly less by
55-70 year-olds than 16-54 year-olds. It can also be seen that C2DE respondents
were significantly less likely to use the Multiple Traffic Lights signposting than ABC1
respondents.
From Figure 19 it can be seen that all signposting except that of Simple Traffic Lights
was used significantly less by those with no children in household than those with
children in the household.
Those in Scotland were significantly less likely to say they used Colour-coded GDA
than those from England, Wales and Northern Ireland.
On Figure 20 it can be seen that those who were observed by interviewers to have
referred to the Nutritional Panel were significantly more likely to say they referred to
the signposting when using Multiple Traffic Lights, Colour-coded GDA and
Monochrome GDA.
Those rating the idea of signposting as very useful were also more likely to say they
had used each of the signposting concepts than those who did not rate signposting
as very useful.
Figure 21 shows the proportion of evaluations where respondents said they had used
the three constituent parts of the signposting, namely the colour-coding, the per
serving figures and the GDA figures.
Figure 21
Colours 66%
It can be seen that, in most cases, respondents claimed to have used the colour-
coding. The per serving figures were used significantly less than the colour-coding
and in a reasonable proportion of evaluations. In a small proportion of evaluations,
respondents claimed to have used the GDA figures and in only a very small
proportion did respondents claim to use the both the per serving figures and GDA
figures, as the information was intended to be used.
Given the significantly higher level of correct response when Colour-coded GDA was
used compared to the Monochrome GDA, it is clear that the colour coding on Colour-
coded GDA helped consumers to ascertain the level of nutrient content of products in
the single product evaluation.
As explained in Section 3.1, respondents were shown photographs of the five pairs of
different products (one by one) and asked to state which of the two was higher or
lower in two of the nutrients most relevant to those products (or whether they were
both the same), based on either the front-of-pack signposting or other information
that they chose to refer to. Each pair of products bore a different format of
signposting information: either one of the four signposting concepts or products
without any signposting.
6.2.1 Understanding
The green bars in Figure 22 show the proportion of correct responses of the total
evaluations (among all respondents) across all four nutrients and five product
categories when asked whether a product was high, medium or low in the specified
nutrients.
Figure 22
Information Gathering Which Contains the most?
Across All Nutrients and Categories COMPARISON -
ABDE AE TOTAL SAMPLE
AE 88%
85%
83%
67% 66%
Across
all
nutrients
33% 34%
Correct
response
Incorrect
17% 15% response
12%
When asked to say which product of various pairs of products contained the most of
a specific nutrient or whether they contained the same amount, Colour-coded GDA
elicited the highest level of correct answers across all nutrients and product
categories significantly higher than each of the other signposting concepts and the
products without signposting.
The correctness level of Simple Traffic Lights was fair and was at a very similar level
to that of products without signposting.
Figure 23
Across 60%
all
nutrients
40% Correct
response
Incorrect
response
14%
12%
8%
The level of correct response for Simple Traffic Lights was significantly lower than
that of the total sample, implying that this signposting format was not helpful in
8
After the respondent stated whether they thought a product was high, medium or low in a
specified nutrient, he or she was asked (without being prompted) what information, if any,
they were using when giving their answer. From this it was ascertained which
respondents claimed to use signposting.
This level of correct response was significantly higher than those for the other three
concepts and the products without signposting. Although once again, the level of
correct responses when using Multiple Traffic Lights and Monochrome GDA was also
very high (86% and 88% respectively).
Figure 24 shows the proportion of correct responses for each of the four signposting
routes among the key demographic subgroups.
It can be seen that the pattern of response among each of the subgroups is similar to
that among the total sample, with Colour-coded GDA performing significantly higher
than the other three signposts among most of the key subgroups.
Figure 24
Figures 25, 26 and 27 below show the proportion of correct responses for each of the
four signposting concepts and the products without signposting among all of the
demographic, attitudinal and user subgroups.
Figure 25
It can be seen that Colour-coded GDA produced a high level of correct responses
across all demographic groups, and performed significantly better than Multiple
Traffic Lights and Monochrome GDA among approximately a third of demographic
subgroups.
From Figure 26 it can be seen that in Scotland the level of the correct identification
when using Simple Traffic Lights or Multiple Traffic Light or Monochrome GDA
concepts is significantly lower than in England.
Need to be careful about sat. fat in diet 60% 81% 86% AE 81% 68%
Need to be careful about salt in diet 59% 79% 85% ABE 82% 66%
Need to be careful about sugar in diet 67% 80% 83% AE 81% 71%
Need to be careful about calories in diet 53% 76% A 84% AE 80% A 71% A
It can be seen from Figure 27 that the level of correct response for all concepts and
products without signposting was significantly higher amongst those claiming to use
the nutritional panel than those who said they did not use it.
6.2.1.2 By Nutrient
Figure 28 shows the proportion of correct responses by each of the four nutrients (as
well as for all nutrients combined) across all categories when using each of the five
signposting concepts and the products without signposting.
Colour-coded GDA produced a high level of correct responses across fat, saturated
fat and salt (on average across all nutrients and product categories).
The levels of correct identification for Multiple Traffic Lights for salt and sugar were
not significantly lower than those for Colour-coded GDA. The correct response levels
for Multiple Traffic Lights for fat and saturated fat were significantly lower than those
of Colour-coded GDA. The correct response level for fat for Multiple Traffic Lights
was significantly lower than the average across all nutrients for Multiple Traffic Lights.
This significantly lower level of correct responses for fat when using Multiple Traffic
Lights is almost solely attributable to the lower correctness scores for crisps. Both
these products bore the same traffic light colour (red) for fat.
It can be seen from Figure 29 that when asked which of the two products was higher
in fat or whether they contained the same amount, while in 67% of all evaluations
respondents said which product was higher, in 21% they said that both products
contained the same level of fat.
The level of 21% was significantly higher than the proportion saying both the same
when using Colour-coded GDA, which was 6%, indicating that when the traffic light
colour is the same for a specific nutrient, it is more difficult to establish which product
contains a higher or lower amount of a nutrient when using a signposting concept
which is based on only colour-coded banding and does not included numerical
information.
COMPARISON
Responses Given for Fat TOTAL SAMPLE
FAT Ready Crisps Ready Crisps Ready Crisps Ready Crisps Ready Crisps
meals meals meals meals meals
Product 1 has higher 42% 74% 17% 67% 6% 88% 9% 85% 44% 80%
content
Product 2 has higher 50% 14% 80% 10% 91% 5% 87% 8% 49% 13%
content
Both the same 5% 10% 2% 21% 2% 6% 4% 5% 4% 5%
Cant tell 3% 1% 1% 1% 1% 0% 0% 1% 2% 1%
Dont know 1% 1% 0% 1% 0% 1% 0% 0% 1% 2%
It can be seen from Figure 30 that Colour-coded GDA produced a similar high level
of correct responses across each of the five categories of breakfast cereals, ready
meals, pizzas, curry meals and crisps.
Figure 30
The levels of correct identification for Multiple Traffic Lights for breakfast cereals,
pizzas and curry meals were not significantly different to those for Colour-coded
GDA.
Figure 31 shows the level of correct responses for each of the different product
categories by nutrient.
Colour-coded GDA and Monochrome GDA produced a similar high level of correct
responses that was significantly higher on each of the nutrients within each of the
categories than Simple Traffic Lights and products without signposting.
This is not the case for Multiple Traffic Lights however, which performs less well on
crisps (67% correct responses for fat). The reason for this poorer performance is
addressed in section 6.2.1.2 above.
SUGAR
Breakfast Cereals 81% 93% 94% 93% 78%
Figure 32 shows the speed in which the information on the signposting was
interpreted for the different signposting concepts and the products without
signposting. It can be seen that the responses were given most quickly with Colour-
coded GDA, significantly faster than the time taken to interpret and answer using the
other three concepts and the products without signposting. The times to interpret and
answer for Monochrome GDA and Multiple Traffic Lights were also fast.
BCD ABCD
CD C
9.5 10.0
Across
all 6.3 5.6 6.0
nutrients
Figures 33, 34 and 35 show the average time taken by the different demographic,
attitudinal and user subgroups.
When the signposting was claimed to have been used, the time to interpret was even
faster for Colour-coded GDA, 5.1 seconds on average, significantly faster than for all
other signposts and products without signposting. Multiple Traffic Lights and
Monochrome GDA elicited the next fastest average times of 5.5 and 5.6 seconds
respectively.
Colour-coded GDA elicited responses that were significantly faster than Simple
Traffic Lights and the products without signposting across all demographic subgroups
and significantly faster than Multiple Traffic Lights and Monochrome GDA across
about three-quarters of the subgroups.
Monochrome GDA elicited significantly faster responses than Multiple Traffic Lights
across approximately a quarter of the subgroups.
It can be seen that for all concepts and for the products without signposting,
responses are given significantly quicker by 16-34 year-olds than by those from older
age groups.
Figure 35
Need to be careful about sat. fat in diet 9.5 6.3 C 5.3 6.3 9.1 BCD
Need to be careful about salt in diet 10.0 6.8 NS 6.2 6.8 NS 10.2 BCD
Need to be careful about sugar in diet 9.4 6.7 NS 6.0 6.3 NS 9.6 BCD
Need to be careful about calories in diet 8.6 CD 6.9 NS 5.5 5.6 NS 8.4 CD
Figure 36 shows the time to respond using each of the signposting concepts and the
products without signposting by nutrient.
It can be seen that the time to respond when using Colour-coded GDA signposting
was significantly faster than when using the other three signposting concepts and the
products without signposting for saturated fat. It was faster than all except
Monochrome GDA for fat and salt, and faster than all signposts except Monochrome
GDA and Multiple Traffic Lights for sugar.
Time To Interpret (in seconds) Which contains the most? Across All
Categories
BCD BCD COMPARISON
CD BY SUBGROUP
BCD
CD BCD
C
Saturated
9.4 9.7
Fat 6.6 6.2
5.7
BCD BCD
Salt CD
10.0 10.6
6.0 5.4 5.7
After the respondent stated whether they thought a product was high, medium or low
in a specified nutrient, he or she was asked (without being prompted) what
information, if any, they were using when giving their answer. Figure 37 shows the
proportion of respondents claiming to use different sources of information for each of
the signposting concepts and the products without signposting when answering the
nutrient content questions.
It can be seen that consumers claimed that they used the signposting more with
Colour-coded GDA (28%) than with the GDA concepts: 13% of evaluations with
Colour-coded GDA and 16% of evaluations with Monochrome GDA. This is perhaps
to be expected, as the GDA-based concepts contain much of the same information
as shown on the back of pack.
The differences in the use of signposting with each of the different concepts and the
products without signposting route are addressed in greater detail in section 6.2.3.2.
Figure 37
67% 67%
60%
28%
19%
16% 14% 14%
13%
6%5% 6% 6%4% 4%8% 5%3% 3%6% 4%2% 4%
3% 2% 1%1%
0%
N/A
Signposting Nutritional panel Common Ingredients list I just know about Didn't use
sense/general nutrition of this anything
knowledge type of product
It can be seen from Figure 38 that the signposting was claimed to have been used
significantly more with Colour-coded GDA than with the other three concepts when
answering the questions about nutrient content.
As may be expected due to the nature of the task, the proportion of evaluations in
which people claimed to have used the signposting with the two GDA concepts was
significantly higher than that relating to the Individual Product Evaluations.
Across
all 33%
nutrients
24%
19% 20%
Used
Didnt use
It can be seen in Figures 39, 40 and 41 that Colour-coded GDA signposting was
claimed to have been used significantly more across all subgroups than Simple
Traffic Lights and Multiple Traffic Lights. In addition, it was claimed to have been
used more than Monochrome GDA among most demographic subgroups. The only
subgroups among which there was no significant difference in the extent to which the
Colour-coded GDA signposting was claimed to be used were:
It can be seen from Figure 39 that the signposting of Multiple Traffic Lights, Colour-
coded and Monochrome GDA were used significantly less among respondents aged
55-70 than those respondents aged 16-54. The difference for Multiple Traffic Lights
was larger than the two GDA concepts.
From Figure 40 it can be seen that all signposting except that of Simple Traffic Lights
was used significantly less by those with no children in household than those with
children in the household. This was also the case in the Individual Product
Evaluations.
Those rating the idea of signposting very useful were also more likely to say they had
used the Multiple Traffic Lights signposting, Colour-coded GDA and Monochrome
GDA signposting than those who did not rate signposting as very useful.
Figure 39
Figure 41
Need to be careful about sat. fat in diet 19% 66% 79% 75%
Need to be careful about salt in diet 18% 63% 76% 74%
Need to be careful about sugar in diet 18% 63% 77% 72%
Need to be careful about calories in diet 20% 65% 81% 70% A
If a respondent said they had used the Colour-coded GDA signposting, they were
asked what specific information on the signposting they had used.
Figure 42 shows the proportion of evaluations in which respondents said they had
used the three constituent parts of the signposting, namely the colour-coding, the per
serving figures and the GDA figures.
Figure 42
Colours 66%
The pattern of responses is very similar to that of the Individual Product Evaluations.
It can be seen that, in most cases, respondents claimed to have used the colour-
coding. The per serving figures were used significantly less than the colour-coding
and by a reasonable proportion. In a small proportion of evaluations, respondents
claimed to have used the GDA figures and in only a very small proportion they
claimed to have used the both the per serving figures and GDA figures, as the
information was intended to be used.
Following the Performance section of the interview, respondents were shown a board
displaying the four alternative signposting concepts (not mounted on photographs of
products) and asked which concept they preferred most, second most, third most and
least.
Figure 43 shows the proportion of all respondents that preferred each of the concepts
most. People were not aware whether they had correctly used the signposts during
the earlier performance section.
Figure 43
30%
65% 95%
preferred
a colour-
Base: All respondents
coded
Colour-coded GDA 2673; Q5 nutrient
based
scheme
Multiple Traffic Lights was preferred most by the second largest proportion and by a
significantly larger proportion than those liking the Monochrome GDA and Simple
Traffic Lights, both of which were preferred most by only a very small proportion.
Figures 44, 45 and 46 show the proportion of respondents preferring each of the
signposting concepts among all of the demographic, attitudinal and user subgroups.
Figure 44
Figure 46
Respondents aged 55-70 were significantly more likely to prefer Multiple Traffic
Lights than respondents aged 16-34 but nevertheless respondents aged 55-70 were
still significantly more likely to favour Colour-coded GDA over Multiple Traffic Lights.
It is interesting that those claiming to use the nutritional panel occasionally or never
are significantly more likely to prefer Multiple Traffic Lights than those who claim to
use the nutritional panel always or usually. For Colour-coded GDA, the opposite is
the case: those claiming to use the nutritional panel occasionally or never are
significantly less likely to prefer Colour-coded GDA than those who claim to use the
nutritional panel always or usually. This would suggest that peoples familiarity with
the nutritional panel is a driver of preference for Colour-coded GDA.
7.2 Proportion Preferring Each Concept Most, Most or Second Most and
Proportion Liking Each Concept Least
Figure 47 shows the proportion of the sample that preferred each of the concepts
most or second most and the proportion that liked each of them least.
Figure 47
6%
1st or 2nd 67%
preferred 93%
32%
The vast majority of respondents preferred Colour-coded GDA most or second most,
while the majority preferred Multiple Traffic Lights most or second-most.
From the third set of bars it can be seen that Simple Traffic Lights was liked least by
the majority of respondents and significantly more so than the other three concepts.
Monochrome GDA was liked least by a fair proportion, whilst only a very small
proportion liked either Colour-coded GDA or Multiple Traffic Lights least.
The large proportion preferring Colour-coded GDA on the one hand and the very
small proportion preferring Monochrome GDA most and the relatively large
proportion disliking Monochrome GDA most imply that the colour-coding was a key
driver of preference.
Figure 48
TOTAL
Reason for Preferring Each Concept Most SAMPLE
Monochrome
Simple Traffic Multiple Traffic Colour GDA
GDA
Base size: 41 # 798 1723 90
Like the colour codes 17% 22% 46% -
Multiple Traffic Lights was preferred because it was felt to be easy to read,
understand and use and because the information can be used quickly or at a glance.
The colour coding was also named by a sizeable proportion as a reason for
preferring it.
It can be seen that Colour-coded GDA was preferred above all for its colour coding.
Other main reasons for respondents liking it most were the fact that it gives exact
numbers, amounts and all the information a consumer could need, followed by
respondents citing the high/medium/low guide.
Monochrome GDA was preferred because of its detailed information and giving all
the information a consumer could need and because of its being clear and easy to
read.
Of the very small proportion (and low base size) preferring Simple Traffic most, it
being clear and easy to read was the main reason for liking it most.
Figure 49 shows the proportion of those liking each of the concepts least stating each
of the (grouped) reasons listed.
Figure 49 TOTAL
Reason for Liking Each Concept Least SAMPLE
Monochrome
Simple Traffic Multiple Traffic Colour GDA
GDA
Base size: 1689 132 71# 742
Doesnt explain anything gives too little
77% 61% 10% 23%
information
Only says OK choice 23% 2% - 1%
# = caution:
low base
This was also the main reason given for liking Multiple Traffic Lights least (61%).
The two GDA concepts were liked least because they were perceived as being
difficult to understand (49% for Colour-coded GDA9 and 31% for Monochrome-GDA).
Monochrome GDA was also criticised by 22% for not having a colour-coded guide.
9
Note that the base for this is low: 71 respondents
After the preference section of the interview, a number of attributes were read out to
respondents and they were asked to which of the concepts, if any, the attributes
applied.
Figures 50 and 51 show the findings from these questions. Figure 50 shows the
findings for the positive attributes while Figure 51 shows the results of the negative
attributes.
It is suggested that Colour-coded GDA was affected by the positive halo effect
described above and therefore the relative scores of it compared to the other
concepts are not necessarily a true reflection of the genuine opinions of respondents.
The strongest evidence of a positive halo effect for Colour-coded GDA is the fact that
it is rated by a higher proportion of respondents as being easy to understand and use
than Multiple Traffic Lights, despite the fact that (a). Multiple Traffic Lights performs
better than Colour-coded GDA on performance in the Individual Product Evaluations
in terms of eliciting the highest proportion of correct responses and eliciting them
most quickly, and (b). Multiple Traffic Lights is preferred by a higher number of
respondents (spontaneously) than Colour-coded GDA for being easy to read, use
and understand.
Figure 50
61%
45%
41%
Would help me make Would help me work out Easy to understand and Would help me compare Gives the right amount of
healthier food choices how much fat, salt, sugar use easily info
more quickly in diet
65%
42%
27%
23%
11%
8% 7%
5% 4% 3%
1%
It can be seen on Figure 50 that a high proportion of respondents said that the
Colour-coded GDA possessed all five of the positive attributes. The proportion citing
Colour-coded GDA was significantly higher than the proportion citing all other three
concepts for each of the attributes.
A fair proportion of respondents said that Monochrome GDA would help them work
out how much fat, salt and sugar there is in their diet.
From Figure 51 it can be seen that a very large proportion of respondents said that
Simple Traffic Lights did not give them enough information about fat, salt and sugar
and a large proportion said it was too simple.
A relatively small proportion said that Multiple Traffic Lights did not give them enough
information about fat, salt and sugar and was too simple.
It can also be seen that a fair proportion of respondents said that Monochrome GDA
was too complicated, while only a small proportion said this for Colour-coded GDA,
implying that the colours on the latter make it much easier to use than the former.
Respondents were significantly less likely to say that Multiple Traffic Lights was too
complicated compared to each of the other signposting concepts.
After the performance and preference section, respondents were asked to rate how
useful they thought the general idea of having signposting on food products would be
to them. The findings are shown in Figure 52.
Figure 52
It can be seen that nearly all respondents thought the idea of signposting was a
useful one, with the majority rating it as a very useful idea. Very few rated it as not
being useful and very few did not have an opinion (2%).
The subgroups significantly more likely to rate the idea of signposting as very useful
(compared to other subgroups within the grouping) and the proportion of each of
these subgroups rating it as very useful are as follows:
The proportion of respondents who would like to know more information on how the
signposting data was calculated is shown in Table 4.
Table 4
Proportion of total
sample
%
Would like to know more information 54
Would not like to know more information 46
Base: 2676
It can be seen that around half of the sample would like to know more information,
whilst a significantly lower proportion, also around half, said they did not want to
know more.
Those who said that they would like to know more information about how the
signposting data is calculated were then asked where they would like information to
be provided. The findings are shown in Figure 53.
Figure 53
Where Would Like Information on Signposting to be Provided
On food
72%
packaging
In leaflets
available 37%
in store
On signs
15%
in stores
The majority said they would like information to be provided on food packaging itself.
Respondents were asked on which food categories, if any, they would like front of
pack signposting information to be displayed. A list of 23 categories was shown to
respondents and is appended in Annex 4. Figure 54 shows the findings. It should be
noted that figures for individual food categories account for those respondents saying
they would like to see signposting displayed on either all foods and drinks appearing
on the list shown to them or on all foods and drinks.
Figure 54
Food Categories On Which Would Like to See Signposting - Ranked
Burgers, sausages, nuggets & fish fingers 84%
Chilled and frozen ready meals 83%
Breakfast cereals 83%
Pizzas 82%
Cakes & biscuits 82%
Crisps 80%
Pies 79%
Chocolate & sweets 78%
Soft drinks 78%
Tinned products 77%
Ice cream & desserts 77% Note that figures
Cooked meats 75% for individual
Sandwiches 75%
food categories
Yoghurts 74%
Fruit juices 74% account for those
Butter & Spreads 73% saying they
Fats & oils 71% would like
Cheese 70% Signposting on
Bread including rolls 70% all those listed/all
Pasta, rice & dried foods 69% foods and drinks
Fresh meat and fish 66%
Milk & Cream 66%
Fresh fruit & veg 62%
All those listed 31%
All food & drinks 29%
None of them 4% Base: All respondents 2676 ; Q9c
The category on which consumers would most like to see signposting was burgers,
sausages, nuggets and fish fingers (84%).
The highest scores were for meal centres (burgers, sausages, nuggets and fish
fingers (84%), chilled and frozen ready meals (83%) and also pies (79%)), breakfast
cereals (83%), pizzas (82%), snacks (cakes and biscuits (82%), crisps (80%)) and
confectionery (chocolate and sweets (78%)).
10.2 Among Respondents with Children Living in the Household and Without
Children Living in the Household
Figures 55 and 56 show the proportion of respondents with children living in their
household and those without children living in their household who would like
signposting to appear on the packaging of the different food categories.
From Figure 55 it can be seen that significantly more respondents with children
would like to see signposting on products than respondents without children for the
following categories:
The difference for crisps is particularly marked, placing these in fourth position
among respondents with children, compared to sixth-equal among respondents
without children.
Figure 55
Food Categories On Which Would Like to See Signposting by
Presence of Children in Household (1 of 2)
88%
Burgers, sausages, nuggets & fish fingers
83% *
85% Children in
Cakes & biscuits
81% * household
85% Any
Pizzas
81% * None
83%
Chilled and frozen ready meals
83% Note that figures
83% for individual
Breakfast cereals
83% food categories
Crisps
84% account for those
78% * saying they
81% would like
Chocolate & sweets
76% * Signposting on
81%
Pies
78%
all those listed/all
foods and drinks
78%
Tinned products
77%
79%
A red asterisk (*)
Soft drinks between two
77%
79% numbers denotes a
Ice cream & desserts
76% significant difference
75% at the 95%
Yoghurts
73% confidences level
74%
Sandwiches Base: All respondents: Any children
75% 835, None in household 1841 ; Q9c
Respondents were asked to rate the importance of the four nutrients focused on in
the research (fat, saturated fat, salt and sugar), as well as calories, in terms of their
importance when considering whether a product is healthy or not. The findings of this
question are shown in Figure 57.
Figure 57
Very important Quite important Not very important Not at all important
Mean score where very important = 4 to not at all important = 1 Base: All respondents 2676; D1
It can be seen that around a half of all respondents rate fat, saturated fat, salt and
sugar as very important, while around one third rate calories as very important. Fat
and saturated fat are rated significantly more important on average than salt, sugar
and calories.
Calories are rated significantly less important (on the mean importance score and the
very important score) than the other four nutrients.
Respondents were asked how often they read nutritional panels on food products
when they are shopping.
It can be seen from Figure 58 that around a quarter say they read them always, while
similar proportions say they read them usually or occasionally.
Always 25%
Usually 28%
Occasionally 25%
Rarely 12%
Never 10%
A slightly smaller proportion said either they rarely or never read nutritional panels.
Respondents who said they were careful about certain nutrients when they were
shopping were significantly more likely to say they always read nutritional labels as
shown in Table 5.
Table 5
Interviewers made a note at the end of the interview whether the respondent had
spontaneously asked what GDA meant during the course of the interview.
Eleven per cent of the sample asked what GDA meant at some point during the
interview, while 89% did not.
Subgroups significantly more likely to ask what GDA meant compared to other
subgroups in the same group are shown in Table 6.
Table 6
Proportion of subgroup
%
Respondents from ABC1 socio-economic groups 13
Respondents from AB socio-economic groups 16
Wales 18
South England 14
Those preferring Colour-coded GDA concept 12
Base: 2676
A boost of respondents from the largest broad ethnic minority groups (as identified in
the 2001 census) was recruited, namely Asian and Asian British, Black and Black
British. Within these broad groups a mix of different ethnic groups was recruited. In
total 166 interviews were conducted among these groups, while 289 were conducted
among respondents of any non-white ethnic origin.
At the analysis stage, the data was weighted to be representative of the ethnic profile
(and other demographic data) of the UK, as explained in Section 3. The incidence of
people from an ethnic minority in the UK is 7.9%10. Within this, the largest ethnic
groups are as follows:
12.1.1 Understanding
It can be seen from Figure 59 that, consistent with the results among the total
sample, among the main ethnic minority groups, Multiple Traffic Lights performed
best, and Colour-coded GDA scored second best. The level of correct response for
Multiple Traffic Lights and Colour-coded GDA was slightly (but not significantly) lower
than for the total sample.
10
Census data 2001
25%
21%
In Figure 60 it is evident that respondents from main ethnic minority groups were
slightly more likely to give a correct response when using signposting than the total
sample with Multiple Traffic Lights.
As among the total sample, Multiple Traffic Lights elicited the highest level of correct
response among respondents of main ethnic minority groups, with Colour-coded
GDA eliciting the second-highest correct response level.
Figure 60
Understanding Proportion Answering Correctly Whether Product is
High, Medium or Low - Across All Nutrients and Categories
ACD Main Ethnic Groups Total INDIVIDUAL
ACD
93% Correct sample
90% USED
response
Incorrect SIGNPOSTING
response AD
AD
69% 69% 69%
66%
62%
58%
A
42%
38%
34%
31% 31% 31%
10%
7%
From Figure 61 it is clear that also consistent with the total sample results, among
the main ethnic minority groups, Multiple Traffic Lights elicited responses significantly
faster than the other three concepts and products without signposting. The times to
respond were similar among respondents from main ethnic minority groups
compared to the total sample, although the time to respond among the former was
slower when using Simple Traffic Lights and Colour-coded GDA.
Figure 61
BCD
BCD BCD
BCD
BC
B B
B
8.2 8.5
8.7 8.4
6.7
6.5 5.4
5.9 5.1
5.0
Figure 62 indicates that among the main ethnic minority groups, the extent to which
signposting was claimed to be used with each of the four concepts was similar to that
of the total sample. However, there was no significant difference between the
proportion using Multiple Traffic Lights and Colour-coded GDA, with the former used
slightly more.
Figure 62
Whether Signposting Referred To When Answering Any of Nutrient
Content Questions INDIVIDUAL
Main Ethnic Groups Total sample
79% Used 79%
AD ABD
AD Didnt use AD 69%
67% 66% A
A 63%
59% 60%
41% 40%
37%
33% 34%
31%
21% 21%
12.2.1 Understanding
As is evident in Figure 63, the pattern of the level of correct responses was similar
among respondents of main ethnic minority groups to that of the total sample.
However the levels of response among main ethnic minority groups was generally
lower than among the total sample, with the largest difference evident with products
without signposting. Unlike the total sample, among main ethnic minority groups
there was no significant difference between the level of correct response of Colour-
coded GDA and the high levels for Monochrome GDA and Multiple Traffic Lights.
Figure 63
42%
36% 34%
33%
19% 17%
17% 15%
15%
12%
In Figure 64 it is evident that the level of correct responses when using signposting
among the main ethnic minority groups with Multiple Traffic Lights is slightly lower
than among the total sample, while for Colour-coded GDA it is the same, and for
Monochrome GDA it is very similar.
As among the total sample, Colour-coded GDA performed best among main ethnic
minority groups, producing a significantly higher level of correct response. However,
unlike among the total sample, Colour-coded GDA did not score significantly better
than Monochrome GDA.
60%
56%
44%
40%
17%
14%
11% 12%
8% 8%
From Figure 65 it is clear that among the main ethnic minority groups there was no
significant difference in the time to interpret for Colour-coded GDA, Monochrome
GDA and Multiple Traffic Lights in fact Colour-coded GDA elicited slightly slower
responses than Multiple Traffic Lights and Monochrome GDA, which differed from the
total sample, among whom responses using Colour-coded GDA were significantly
faster than with other signposts and products without signposting.
Figure 65
Time To Interpret (in seconds) Which contains the COMPARISON
most? Across All Nutrients and Categories
Main Ethnic Groups Total sample
CD C
Figure 66 indicates that among the main ethnic minority groups the extent to which
signposting was claimed to have been used with each of the four concepts was
similar to that of the Total Sample. However, the signposting was used slightly less
with Multiple Traffic Lights and Monochrome GDA among respondents from main
ethnic minority groups than among the total sample.
Figure 66
38%
33%
27%
24%
22%
19% 20%
17%
It can be seen in Figure 67 that among respondents of main ethnic minority groups,
the proportion preferring Colour-coded GDA was lower than among the total sample
and the proportion preferring Monochrome GDA was significantly higher. Colour-
coded GDA was nevertheless the most preferred signpost among main ethnic
minority groups.
Figure 67
Multiple Multiple
Traffic Traffic
Lights Lights
30%
30%
60%
65%
Colour-coded GDA
Colour-coded GDA
This section examines the results among the lowest socio-economic groups, D and
E, compared to the higher socio-economic groups of ABC1 combined and C2 group.
13.1.1 Understanding
From Figure 68 it is clear that, consistent with the total sample results (shown in
Figure 1), and ABC1 respondents, among C2 and DE respondents, Multiple Traffic
Lights performed best, and performed to a level similar to that of the total sample.
The level of correct response for all of the signposts and products without signposting
was generally lower among C2 and DE respondents than among ABC1 respondents.
Figure 68
24%
22%
19%
Simple Multiple Colour Mono No Simple Multiple Colour Mono No Simple Multiple Colour Mono No
Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost
A B C D E A B C D E A B C D E
Correct Base: Simple Traffic Light 2264, Multiple Traffic Light 2260, Colour
response GDA 2274, Monochrome GDA 2266, No signposting 2256; Q1
Incorrect
response
In Figure 69 it is evident that, consistent with the total sample (shown in Figure 2),
the level of correct responses among respondents from DE and C2 groups was
significantly lower when using Colour-coded GDA, compared to respondents from
ABC1 groups.
Once again, Multiple Traffic Lights elicited the highest level of correct response
among respondents of C2 and DE socio-economic groups (as well as ABC1), with
Colour-coded GDA scoring second highest.
Figure 69
AD
AD AD 73%
69% 69%
67% 67%
65%
61%
59%
57%
A A A
43%
41%
39%
35%
33% 33%
31% 31%
27%
Simple Multiple Colour Mono Simple Multiple Colour Mono Simple Multiple Colour Mono
Traffic Traffic GDA GDA Traffic Traffic GDA GDA Traffic Traffic GDA GDA
A B C D A B C D A B C D
Base: C2: Simple Traffic Light 249, Multiple Traffic Light 682, Colour GDA 781, Monochrome GDA 691 Correct
DE: Simple Traffic Light 271, Multiple Traffic Light 797, Colour GDA 939, Monochrome GDA 809; response
ABC1: Simple Traffic Light 621, Multiple Traffic Light 1869, Colour GDA 1974, Monochrome GDA 1736; Q1
Incorrect
response
Figure 70 indicates that consistent with the total sample results (shown in Figure 11),
among C2 and DE respondents, Multiple Traffic Lights elicited responses most
quickly. Respondents of C2 socio-economic grade gave responses faster of those of
respondents from ABC1 and, in particular, DE socio-economic grades. However,
while among the total sample and ABC1 respondents, Multiple Traffic Lights elicited
responses significantly faster than the other three concepts and the products without
signposting, among C2 and DE respondents, there was no significant difference in
the time taken to respond for Multiple Traffic Lights and Colour-coded GDA.
Figure 70
C2 DE ABC1
8.5
7.8 7.9 8.9 8.8 8.0
6.8
5.9 6.9
5.7 5.4
4.7 5.0 5.6 5.0
Simple Multiple Colour Mono No Simple Multiple Colour Mono No Simple Multiple Colour Mono No
Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost
A B C D E A B C D E A B C D E
Base: Simple Traffic Light 332, Multiple Traffic Light 332, Colour Base: Simple Traffic Light 5358, Multiple Traffic Light 5342,
GDA 332, Monochrome GDA 332, No signposting 332: Q2a Colour GDA 5366, Monochrome GDA 5356, No signposting 5338:
Q2a
In Figure 71 it is clear that the signposting was claimed to have been used to a
similar extent with each of the signposts by each of C2, DE and ABC1 respondents,
with colour-coded GDA used significantly more than other concepts. The signposting
was however used slightly more with Multiple Traffic Lights among ABC1 groups than
C2 and DE groups.
Figure 71
C2 DE ABC1
80% 79%
78%
ABD ABD ABD
69%
A 70% AD 68%
A A 65% A
61% 62% 61%
60% 60%
E
39% 40% 39% 40%
38%
35%
31% 32%
30%
22% 21%
20%
Simple Multiple Colour Mono Simple Multiple Colour Mono Simple Multiple Colour Mono
Traffic Traffic GDA GDA Traffic Traffic GDA GDA Traffic Traffic GDA GDA
A B C D A B C D A B C D
Base: C2: Simple Traffic Light 1127, Multiple Traffic Light 1124, Colour GDA 1131, Monochrome GDA 1121; Used
DE: Simple Traffic Light 1338, Multiple Traffic Light 1334, Colour GDA 1337, Monochrome GDA 1136;
ABC1: Simple Traffic Light 2889, Multiple Traffic Light 2883, Colour GDA 2899, Monochrome GDA 2899; Q2a
Didnt use
13.2.1 Understanding
Secondly, the level of correct responses among C2 and DE respondents was lower
than that of ABC1 respondents for all signposting concepts.
81%
79% 79%
70%
67% 68%
64% 63%
61%
39%
36% 37%
33% 32%
30%
21% 21%
19%
16% 16% 16%
14%
12%
9%
Simple Multiple Colour Mono No Simple Multiple Colour Mono No Simple Multiple Colour Mono No
Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost
A B C D E A B C D E A B C D E
Correct Base: Simple Traffic Light 2264, Multiple Traffic Light 2260, Colour
response GDA 2274, Monochrome GDA 2266, No signposting 2256; Q1
Incorrect
response
In Figure 73 it can be seen that the level of correct responses among C2 and DE
respondents claiming to have used signposting was lower than that of ABC1
respondents for Colour-coded GDA and Monochrome GDA.
As among the total sample (shown in Figure 22), Colour-coded GDA performed best
among DE and ABC1 socio-economic groups, producing a significantly higher level
of correct response than the other three signposts. However, unlike among the total
sample, Colour-coded GDA did not score significantly better than either Multiple
Traffic Lights or Monochrome GDA among the C2 group.
61% 61%
58%
42%
39% 39%
16% 17%
13% 12% 13%
11% 11%
9%
6%
Simple Multiple Colour Mono Simple Multiple Colour Mono Simple Multiple Colour Mono
Traffic Traffic GDA GDA Traffic Traffic GDA GDA Traffic Traffic GDA GDA
A B C D A B C D A B C D
Correct
Base: C2: Simple Traffic Light 1127, Multiple Traffic Light 1124, Colour GDA 1131, Monochrome GDA 1121; response
DE: Simple Traffic Light 1338, Multiple Traffic Light 1334, Colour GDA 1337, Monochrome GDA 1136; Incorrect
ABC1: Simple Traffic Light 2889, Multiple Traffic Light 2883, Colour GDA 2899, Monochrome GDA 2899; Q3a response
Figure 74 shows that the general pattern of speed of response is similar across C2,
DE and ABC1 groups. However there is no difference in the speed of response
when using Colour-coded GDA and Multiple Traffic Lights among the C2 group,
whereas among the DE group and the Total sample, Colour-coded GDA elicited
significantly faster responses.
C2 DE ABC1
C C C C
D
10.4 10.6
9.7 9.9
9.3 9.2
Simple Multiple Colour Mono No Simple Multiple Colour Mono No Simple Multiple Colour Mono No
Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost Traffic Traffic GDA GDA Signpost
A B C D E A B C D E A B C D E
Base: C2: Simple Traffic Light 1127, Multiple Traffic Light 1124, Colour GDA 1131, Monochrome GDA 1121, No signposting 1118;
DE: Simple Traffic Light 1338, Multiple Traffic Light 1334, Colour GDA 1337, Monochrome GDA 1136, No signposting 1335;
ABC1: Simple Traffic Light 2889, Multiple Traffic Light 2883, Colour GDA 2899, Monochrome GDA 2899, No signposting 2881; Q2a
It can be seen from Figure 75 that the signposting was claimed to have been used to
a similar extent with each of the signposts by each of C2, DE and ABC1
respondents. The signposting was however used slightly less with Multiple Traffic
Lights among C2 groups than among ABC1 and DE groups.
Figure 75
Whether Signposting Referred To When Answering Any of Nutrient
Content Questions
ABC1 COMPARISON
C2 DE ABD
AB AB ABD
81%
79% 79%
81% AB 82%
80% AB
76% 76% 76%
A A
A 68% 68%
64%
36%
32% 32%
Simple Multiple Colour Mono Simple Multiple Colour Mono Simple Multiple Colour Mono
Traffic Traffic GDA GDA Traffic Traffic GDA GDA Traffic Traffic GDA GDA
A B C D A B C D A B C D
Used
Base: C2: Simple Traffic Light 220, Multiple Traffic Light 596, Colour GDA 669, Monochrome GDA 602
DE: Simple Traffic Light 293, Multiple Traffic Light 875, Colour GDA 1030, Monochrome GDA 890
Didnt use
ABC1: Simple Traffic Light 605, Multiple Traffic Light 1877, Colour GDA 2014, Monochrome GDA 1765; Q3c
Figure 76
31% 37%
64% 57%
Colour- Colour-
coded coded
GDA Base: C2: 561, DE: 667; GDA
Q5
Overall across both performance evaluations, the Multiple Traffic Lights and Colour-
coded GDA options performed strongest of the signposting concepts tested. All
signposting concepts performed better than when no signposting was available. A
three point scale has been used to show the relative merits of Multiple Traffic Lights
and Colour-coded GDA measured against the objectives of a government
signposting scheme as shown in the table below.
** ***
2. To allow a consumer to quickly and correctly
identify whether a product is a healthier option
or one high in fat, salt or sugar.
*** **
3. To allow the consumer to make comparisons
between products and to quickly identify
which is the one lower in fat, salt or sugar and
which is the higher. ** ***
4. To be easy and clear to understand.
*** **
5. To be applicable to as a wide a range of
consumers as possible (age, socio-
demographic group, ethnic minorities etc)
*** **
Objective 1. To help consumers make informed choices and to construct a
balanced diet.
From 2,676 respondents interviewed, 96% said they felt that front of pack signposting
would be useful in helping them make healthier food choices.
When consumer preference for signposting formats was explored, 95% of people
preferred a colour-coded individual nutrient based scheme. More than twice as many
respondents preferred the Colour-coded GDA option (65%), than preferred Multiple
Traffic Lights (30%).
When using Colour-coded GDA, consumers also responded more quickly (average
5.1 sec), compared with Multiple Traffic Lights (average 5.5 sec). This difference can
perhaps be explained by the fact that in the case of the Multiple Traffic Lights option,
respondents referred to the nutrition panel in a greater proportion13 of evaluations as
well as the signposting concept to differentiate between the products nutrient content
when comparing two products. When consumers claimed they did not use the
nutritional panel they were less likely to be able to differentiate between products
nutrient content using the Multiple Traffic Lights signpost alone, especially when
comparing products with the same colour coding for the same nutrients.
Those who preferred Colour-coded GDA, said they did so because they liked the
colour-coding (46%) and the detailed numerical information (30%). A smaller
proportion of people preferring this option said that Colour-coded GDA was easy or
clear to read (16%) or to use and understand (13%).
The performance data suggests that while people said they liked having access to
numerical information in the Colour-coded GDA signpost, they were not always able
to use it correctly and were more likely to use the colour coding information as the
basis on which to determine the healthiness of the product.
11
Difference is statistically significant at the 95% confidence level
12
Difference is statistically significant at the 95% confidence level
13
Difference is statistically significant at the 95% confidence level
When the signposting was claimed to have been used in the Two-Product
Comparison Evaluations, among C2 respondents, there was no significant difference
in the correctness level elicited by Colour-coded GDA compared to Multiple Traffic
Lights, while among DE respondents (and C2 and DE groups combined), ABC1
respondents and the Total Sample, Colour-coded GDA did perform significantly
better than Multiple Traffic Lights and Monochrome GDA.
With respect to the main ethnic minority groups, Multiple Traffic Lights performed
best in the Individual Product Evaluation, both in terms of the level of correct
responses, and the speed with which the information was interpreted. Among those
who claimed to use signposting to compare two products, Colour-coded GDA
performed significantly better than Multiple Traffic Lights (92% and 83% correct
responses respectively). There was no significant difference in the time taken to
respond.
In summary, both Multiple Traffic Lights and Colour-coded GDA perform well and
have obvious strengths. The Multiple Traffic Lights option largely performed better
than the Colour-coded GDA option in helping a wide range of consumers, including
the elderly and those from lower socio-economic groups, to quickly and correctly
assess the saturated fat, fat, salt and sugar levels in a range of foods. The Colour-
coded GDA option performed best in helping people compare the nutritional content
of two products. It is notable that among people who said that they used signposting
to respond, while the difference in the level of correct responses between Colour-
coded GDA and Multiple Traffic Lights is 6 percentage points when comparing two
products, Multiple Traffic Lights performed better during the individual product
comparisons by a margin of 21 percentage points.
On balance the Multiple Traffic Lights signpost was found to most closely meet the
research objectives and the Agencys stated objectives for a government signposting
system, and was found by the majority of consumers to be easy to use and
understand.
The shaded boxes indicate the two nutrients which were asked about for each
product in the evaluations.
For the purposes of this exercise, the Agencys guidance on A lot and A little was
used as the basis of the descriptors high (H), medium (M) and low (L). A nutrient
was described as low if it met the criteria for A little, as high if it met the criteria for
A lot and medium in all other cases.
The shaded boxes indicate the two nutrients which were asked about for each
product in the evaluations.
For the purposes of this exercise, the Agencys guidance on A lot and A little was
used as the basis of the descriptors high (H), medium (M) and low (L). A nutrient
was described as low if it met the criteria for A little, as high if it met the criteria for
A lot and medium in all other cases.
DO NOT ALLOW RESPONDENT TO SEE CONCEPT MATERIAL PRIOR TO THE START OF EACH
INTERVIEW THESE SHOULD BE PLACED ON A CHAIR NEXT TO YOU AND OUT OF VIEW.
some additional information on the front of the pack. POINT TO SIGNPOSTING ON FRONT
OF PACK. This information is designed to help you decide whether to choose this product (21-22) 1
as part of a nutritionally balanced diet. Please take a moment to look at it.
REMEMBER TO RECORD TIME RESPONDENT TAKES AT Q1 AT Q2A
Q1. Using the information on the pack, would you say this product is high, medium or low in
INTERVIEWER: ROTATE ORDER OF ASKING SUGAR SALT
(23) (24)
High content 1 1
Medium Content 2 2
Low content 3 3
Cant tell 4 4
Dont know 5 5
Observation only: (25) (26)
Nutritional panel on back of pack referred to: Yes 1 1
No 2 2
Q2a. INT. RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION
SUGAR SALT
(27) (28)
Immediately (within 2 seconds) 1 1
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Longer than 20 seconds 5 5
Q2b. What information, if any, were you using when answering the questions I just asked you
about (insert Nutrient) content? RECORD AS PER PRE-CODES. OTHERWISE
RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT TO POINT
TO CONCEPT BOARD, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
SUGAR SALT
(29-48) (49-68)
Signposting (on front of pack) 1 1
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
Other (SPECIFY)............................................................. 20
Q3a. Using the information on the packs, which, if any, of these products would you say has a
higher (insert Nutrient) content or would you say they are both the same?
INTERVIEWER: ROTATE ORDER OF ASKING
INTERVIEWER RECORD WHETHER NUTRITION PANEL ON BACK REFERRED TO AND FOR WHICH
PRODUCT(S):
SUGAR SALT
Product with higher content:
A21 has higher content 1 1 (69) (70)
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to:
A21 Yes 6 6
No 7 7
A32 Yes 8 8
No 9 9
Q3b. INTERVIEWER RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION:
SUGAR SALT
Immediately (within 2 seconds) 1 1 (71) (72)
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q3c. What information, if any, were you using when answering the questions I just asked you about
which product is higher in (insert Nutrient)? RECORD AS PER PRE-CODES.
OTHERWISE RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT
TO POINT TO CONCEPT BOARDS, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
SUGAR SALT
Signposting (on front of pack) 1 1 (73) (93)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
........................................................................................................................................ 20
Medium Content 2 2
Low content 3 3
Cant tell 4 4
Dont know 5 5
Observation only: (120) (121)
No 2 2
Q2a. INT. RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION
FAT SATURATED FAT
(122) (123)
Immediately (within 2 seconds) 1 1
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q2b. What information, if any, were you using when answering the questions I just asked you about
(insert Nutrient) content? RECORD AS PER PRE-CODES. OTHERWISE RECORD
UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT TO POINT TO CONCEPT
BOARD, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED.
FAT SATURATED FAT
Signposting (on front of pack) 1 1 (124)(144)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
REMOVE BOARD G16
Q3a. Using the information on the packs, which, if any, of these products would you say has a
higher (insert Nutrient) content or would you say they are both the same?
INTERVIEWER: ROTATE ORDER OF ASKING
INTERVIEWER RECORD WHETHER NUTRITION PANEL ON BACK REFERRED TO AND FOR WHICH
PRODUCT(S):
FAT SATURATED FAT
Product with higher content:
G27has higher content 1 1 (164) (165)
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to:
G27 Yes 6 6
No 7 7
G38 Yes 8 8
No 9 9
Q3b. INTERVIEWER RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION:
FAT SATURATED FAT
Immediately (within 2 seconds) 1 1 (166) (167)
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q3c. What information, if any, were you using when answering the questions I just asked you about
which product is higher in (insert Nutrient)? RECORD AS PER PRE-CODES.
OTHERWISE RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT
TO POINT TO CONCEPT BOARDS, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED.
FAT SATURATED FAT
Signposting (on front of pack) 1 1 (168) (188)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
No 2 2
Q2a. INT. RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION
SATURATED FAT SALT
(217) (218)
Immediately (within 2 seconds) 1 1
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q2b. What information, if any, were you using when answering the questions I just asked you about
(insert Nutrient) content? RECORD AS PER PRE-CODES. OTHERWISE RECORD
UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT TO POINT TO CONCEPT
BOARD, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED.
SATURATED FAT SALT
Signposting (on front of pack) 1 1 (219) (239)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
*IF SIGNPOSTING MENTIONED AT Q2B GO TO Q2C ON LAST PAGE, ELSE REMOVE BOARD M51
Q3a. Using the information on the packs, which, if any, of these products would you say has a
higher (insert Nutrient) content or would you say they are both the same?
INTERVIEWER: ROTATE ORDER OF ASKING
INTERVIEWER RECORD WHETHER NUTRITION PANEL ON BACK REFERRED TO AND FOR WHICH
PRODUCT(S):
SATURATED FAT SALT
Product with higher content:
M62 has higher content 1 1 (259) (260)
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to
M62 Yes 6 6
No 7 7
M73 Yes 8 8
No 9 9
Q3b. INTERVIEWER RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION:
SATURATED FAT SALT
Immediately (within 2 seconds) 1 1 (261) (262)
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q3c. What information, if any, were you using when answering the questions I just asked you about
which product is higher in (insert Nutrient)? RECORD AS PER PRE-CODES.
OTHERWISE RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT
TO POINT TO CONCEPT BOARDS, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
SATURATED FAT SALT
Signposting (on front of pack) 1 1 (263) (283)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
*IF SIGNPOSTING MENTIONED AT Q3C, GO TO Q3D ON BACK PAGE, ELSE REMOVE BOARDS M62
and M73
Q1. Using the information on the pack, would you say this product is high, medium or low in
INTERVIEWER: ROTATE ORDER OF ASKING
FAT SATURATED FAT
(308) (309)
High content 1 1
Medium Content 2 2
Low content 3 3
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to:
(310) (311)
Yes 1 1
No 2 2
Q2a. INT. RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION
FAT SATURATED FAT
(312) (313)
Immediately (within 2 seconds) 1 1
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q2b. What information, if any, were you using when answering the questions I just asked you
about (insert Nutrient) content? RECORD AS PER PRE-CODES. OTHERWISE
RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT TO POINT
TO CONCEPT BOARD, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED.
FAT SATURATED FAT
Signposting (on front of pack) 1 1 (314) (334)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
Q3a. Using the information on the packs, which, if any, of these products would you say has a
higher (insert Nutrient) content or would you say they are both the same?
INTERVIEWER: ROTATE ORDER OF ASKING
INTERVIEWER RECORD WHETHER NUTRITION PANEL ON BACK REFERRED TO AND FOR WHICH
PRODUCT(S):
FAT SATURATED FAT
Product with higher content:
T68 has higher content 1 1 (354) (355)
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to
T68 Yes 6 6
No 7 7
T79 Yes 8 8
No 9 9
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q3c. What information, if any, were you using when answering the questions I just asked you
about which product is higher in (insert Nutrient)? RECORD AS PER PRE-CODES.
OTHERWISE RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT
TO POINT TO CONCEPT BOARDS, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
FAT SATURATED FAT
Signposting (on front of pack) 1 1 (358) (378)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
pack. Please imagine you are in a supermarket looking at it. Please take a moment to look at
it. (401-402) 5
REMEMBER TO RECORD TIME RESPONDENT TAKES AT Q1 AT Q2A.
Q1. Using the information on the pack, would you say this product is high, medium or low in
INTERVIEWER: ROTATE ORDER OF ASKING
SATURATED FAT SALT
(403) (404)
High content 1 1
Medium Content 2 2
Low content 3 3
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to:
(405) (406)
Yes 1 1
No 2 2
Q2a. INT. RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS INFORMATION
SATURATED FAT SALT
(407) (408)
Immediately (within 2 seconds) 1 1
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q2b. What information, if any, were you using when answering the questions I just asked you
about (insert Nutrient) content? RECORD AS PER PRE-CODES. OTHERWISE
RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT TO POINT
TO CONCEPT BOARD, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
SATURATED FAT SALT
BLANK CODE 1 1 (409) (429)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
Q3a. which, if any, of these products would you say has a higher (insert Nutrient)
content or would you say they are both the same?
INTERVIEWER: ROTATE ORDER OF ASKING
INTERVIEWER RECORD WHETHER NUTRITION PANEL ON BACK REFERRED TO AND FOR WHICH
PRODUCT(S):
SATURATED FAT SALT
Product with higher content:
Z96 has higher content 1 1 (449) (450)
Cant tell 4 4
Dont know 5 5
Observation only: Nutritional panel on the back referred to
Z96 Yes 6 6
No 7 7
Z107 Yes 8 8
No 9 9
Q3b. INTERVIEWER RECORD APPROXIMATE ESTIMATION OF TIME TAKEN TO OBTAIN THIS
INFORMATION:
SATURATED FAT SALT
Immediately (within 2 seconds) 1 1 (451) (452)
Within 5 seconds 2 2
Within 10 seconds 3 3
Within 20 seconds 4 4
Q3c. What information, if any, were you using when answering the questions I just asked you
about which product is higher in (insert Nutrient)? RECORD AS PER PRE-CODES.
OTHERWISE RECORD UNDER OTHERS. IF FURTHER CLARIFICATION REQUIRED ASK RESPONDENT
TO POINT TO CONCEPT BOARDS, DO NOT PROMPT AS PER PRE-CODES. MULTI CODING ALLOWED
SATURATED FAT SALT
BLANK CODE 1 1 (453) (473)
(POINT TO DOUBLE CHECK THIS IS WHAT IS MEANT)
Nutritional panel (on back of pack) 2 2
................................................................................................................. 20
Q6a. Why do you like [POINT TO RELEVANT CONCEPT] the most? PROBE FULLY (498-557)
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
Q6b. Why do you like [POINT TO RELEVANT CONCEPT] the least? PROBE FULLY (558-617)
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
.....................................................................................................................................................
Quite useful 2
No 2 Q9c
On signs in stores 3
On the internet 4
SHOWCARD Q9c
Q9c. On which of the following food categories, if any, would you like to see this additional
information to be shown? PROBE: Any others? MULTICODE POSSIBLE
Fresh meat and fish 1 Fresh fruit and vegetables 15 (686-720)
Pizzas 5 Yoghurts 20
SHOWCARD Q10
Q10. Which of these products, if any, do you buy nowadays?
Breakfast cereal 1 (721-729)
Ready meals 2
Pizzas 5
Chicken Burgers 6
Cake 7
Crisps 8
Cereal bars 9
QD1. Id now like to ask you how important different nutrients are to you when you are in a shop
considering whether a product is healthier or less healthy.
Firstly, thinking about [INSERT NUTRIENT], how important would you say this is when you
are considering whether a product is healthier or less healthy?
SHOW CARD QD1
INTERVIEWER: ROTATE ORDER OF ASKING
TICK & Very Quite Not very Not at all (Dont
ROTATE START Important important important important know)
Fat 1 2 3 4 5 (730)
Salt 1 2 3 4 5 (732)
Sugar 1 2 3 4 5 (733)
Calories 1 2 3 4 5 (734)
QD2. Can I just check, how often do you read the nutritional panel on food packaging? By this I
mean the table on the back or the side of the packaging that tells you how much fat, sugar
and protein etc. is contained in the product.
SHOW CARD QD2
(735)
Never 1
Rarely 2
Occasionally 3
Usually 4
Always 5
QD2a. INTERVIEWER : Code whether respondent asked what GDA meant at any time during the
(736)
interview
Yes 1
No 2
Vegan (do not eat meat, fish, dairy products, or any product derived (741-742)
______________
from animals)
Following a strict plan to lose weight ______________ (743-744)
Saturated fat 2
Salt 3
Sugar 4
Calories 5
Two 2 2
Three 3 3
Four or more 4 4
QD6. And how many children aged live in your household? INTERVIEWER ASK FOR EACH AGE
GROUP
Children Children Children
aged 0-5 aged 6-10 aged 11-15
(763) (764) (765)
One 1 1 1
Two 2 2 2
Three 3 3 3
Four or more 4 4 4
None 5 5 5
Other (SPECIFY)............................................................. 20
........................................................................................................................................ 20
Other (SPECIFY)............................................................. 20
........................................................................................................................................ 20
(869)