Download as pdf or txt
Download as pdf or txt
You are on page 1of 70

Business and Social Sciences, Aarhus University

Department of Business Administration


June 2012

Understanding of food labels in Denmark

Master Thesis
MSc. Marketing

Author
Mathias Høg

Advisor
Klaus G. Grunert
ABSTRACTS

Purpose – To investigate how well Danish consumers understand food labels and which
determinants that explains the variation in understanding.

Design/Methodology/Approach – The research in this paper is conducted through both a


qualitative method (in-depth interviews) as well as a quantitative method (a survey
questionnaire) based on a frame work that describes how consumers understand. The findings
in the two methods led an insight into how well established the different food labels are in the
mind of Danish consumers.

Findings – This research showed that better known labels had a higher degree of
understanding, however for most of the labels in this research there were some confusion into
what it actually correct. Furthermore the research supported earlier findings about different
determinants for understanding of food labels however new findings were discovered as well.
The research also showed that consumers overall prefer labels on perishable food compared to
more long-life food and some of the most important information is expiration date.

Research limitations – The research is only done on Danish consumers and consumers had
to take part in the daily shopping furthermore respondents in the survey had to either consume
fish or chocolate. Furthermore it was impossible to include all labels on the Danish market
and certain labels in the survey were demanded by Danish Veterinary and Food
Administration.

Practical implications – It helps the Danish Veterinary and Food Administration to get a
deeper understanding of Danish consumer knowledge of the different food labels.

Originality/value – This thesis makes an empirical investigation into Danish consumers


understanding of food labels. Furthermore it looks at different determinates for understanding
and preferences for certain food labels.
Keywords – Animal welfare, Conjoint analysis, Convenience sampling, Danish consumers,
Demographic variables, Environment, Food labels, Determinants, In-depth interviews, Health,
, Knowledge, Logo based labels, Objective understanding, Preferences, Regression analysis,
Snowballing sample, social responsibility, Subjective understanding, Text based labels
Table of contents
1. Introduction .......................................................................................................................... 8

1.1 Problem statement and research questions ....................................................................... 9

1.2 Definitions ........................................................................................................................ 9

1.3 Limitations ..................................................................................................................... 10

1.4 Research approach.......................................................................................................... 10

2. Literature review ................................................................................................................ 11

3. Methodology ....................................................................................................................... 16

3.1 Theoretical foundation ................................................................................................... 17

3.2 In-depth interviews in theory ......................................................................................... 18

3.2 Conjoint analysis ............................................................................................................ 19

3.3 Sampling method............................................................................................................ 20

4. In-depth interviews ............................................................................................................ 21

4.1 Design of In-depth interviews ........................................................................................ 21

4.1.1 Question guide......................................................................................................... 22

4.1.2 Top-of- mind ........................................................................................................... 22

4.1.3 Other information on the products .......................................................................... 22

4.1.4 Demographic questions ........................................................................................... 22

4.2 Data collection................................................................................................................ 23

4.3 Results of in-depth interviews ........................................................................................ 23

5. Survey questionnaire.......................................................................................................... 26

5.1 Design of survey questionnaire ...................................................................................... 26

5.1.1 Design...................................................................................................................... 26

5.1.2 Pilot test ................................................................................................................... 30

5.2 Data collection through survey questionnaire ................................................................ 30

5.2.1 The sample .............................................................................................................. 30


5.2.1.1 Demographic characteristics of the sample ...................................................... 31

5.3 Results of survey questionnaire...................................................................................... 33

5.3.1 Data cleaning ........................................................................................................... 33

5.3.1.1 Merging dataset ................................................................................................ 33

5.3.1.2 Missing data ..................................................................................................... 33

5.3.1.3 Outliers ............................................................................................................. 34

5.3.1.4 Recoded data .................................................................................................... 34

5.3.2 Understanding ......................................................................................................... 34

5.3.2. 1 Subjective understanding ................................................................................ 34

5.3.2.2 Objective understanding ................................................................................... 35

5.3.3 Determinants of understanding ............................................................................... 50

5.3.4 Results of Conjoint analysis .................................................................................... 54

7. Reflection and Conclusion ................................................................................................. 59

8. References ........................................................................................................................... 65
List of tables
Table 1 – Attribute and levels for Conjoint analysis…………………………………………28
Table 2 – Demographic characteristics of the sample………………………………………..31
Table 3 – Subjective understanding…………………………………………………………..35
Table 4 – Objective understanding…………………………………………………………...36
Table 5 – Understanding of _GDA…………………………………………………………...37
Table 6 - Understanding of _MSC……………………………………………………………38
Table 7 - Understanding of _KEY……………………………………………………………39
Table 8 - Understanding of _FAIR…………………………………………………………...40
Table 9 - Understanding of _ORG……………………………………………………………41
Table 10 – Understanding of _ EUORG……………………………………………………...42
Table 11 - Understanding of _GTS…………………………………………………………...43
Table 12 - Understanding of _BGB…………………………………………………………..44
Table 13 - Understanding of _BOB…………………………………………………………..45
Table 14 - Understanding of _SA…………………………………………………………….46
Table 15 - Understanding of _MHT………………………………………………………….47
Table 16 - Understanding of _BF…………………………………………………………….48
Table 17 - Understanding of _ING…………………………………………………………...49
Table 18 - Understanding of _ADDI…………………………………………………………50
Table 19 – Relative attribute importance……………………………………………………..58

Lists of figures
Figure 1 – Conceptual framework……………………………………………………………18

Appendices, datasets and output on print and CD


Appendices – On CD
Appendix 1 – Example of product with different food labels
Appendix 2 – Overview of interest
Appendix 3 - Objective understanding overall statistics
Appendix 4 - Objective understanding - Statement level
Appendix 5 - Subjective understanding
Appendix 6 – Regressions for all 14 items
Appendix 7 – In-depth interviews.

Appendices – On CD but printed as well after references list


Appendix A - In-depth interviews question guide
Appendix B – In-depth interviews – Product example
Appendix C – Overview over In-depth interview results
Appendix D – Example for Conjoint
Appendix E - Picture of Item used in understanding
Appendix F-S – Regression analysis

Datasets – Only on CD
Dataset conjoint plan for Chocolate.sav
Dataset conjoint results for chocolate.sav
Dataset conjoint plan for Fish.sav
Dataset conjoint results for Fish.sav
Dataset Survey questionnaire.sav

Outputs – only on CD
Output 1 – Chi-square test for the two surveys
Output 2 – Conjoint analysis – Fish
Output 3 – Conjoint analysis – Chocolate
Output 4 – Regression analysis
1. Introduction
Within recent years we have seen an increasing number of food labels on the front of
packing like i.e. the fairtrade label compared to the more standardized information on
the back like the ingredient list.
Since Danish consumers were introduced to the red organic label (Ø-mærket) in 1989
the number has increased especially after the introduction of the green keyhole in 2009.
Even though the number of labels has increased in recent years the most well known
label today is still the red organic label, in which 97% of Danish consumers have seen
before according to a report by Fairspeak (2010).

The variety of labels has spread as well and consumers now faces labels that refers to
organic production, recycling labels in which the food are in boxes made out of
recycling material or food chains have made their own labels like COOPs Änglamark
(coop.dk, 2011) or with the introduction of labels on the market that addresses
sustainability; labels like the Fairtrade label, in which consumers that buy products with
this label indirectly supports farmers and workers in third world countries or the MSC
Label that stands for fishing under sustainable conditions or. Furthermore European
quality labels have been introduced on the market as well however they are more used
in Southern Europe (foedevarestyrelsen.dk).The number of different labels that supports
either this or that just keeps growing and if you do a count on the number of labels on
the Danish market for food products you will find more than 20 kinds of labels
(forbrug.dk).

In fact as a consumer you can go down to a store and take down a product from the
shelf i.e. a bag of hazelnuts which have a total of three different labels on the front (see
appendix 1 on CD). However it must remain unsaid if this increased number of food
labels on the Danish market led the Danish Veterinary and Food Administration to start
an investigation with the help from MAPP at Aarhus University. However MAPP was
asked to perform a research on this subject into how well Danish consumers know the
different food labels. What kind of knowledge does a Danish consumer have about the
different labels and do they have any preferences towards certain labels.

Page 8 of 70
So this thesis is based on this collaboration between MAPP and the Danish Veterinary
and Food Administration and seeks out to answer how much and what consumers
actually know about the different food labels on the market and what kind of
preferences they have towards the different labels.

1.1 Problem statement and research questions


There is a variety of different food labels on the Danish market today and this thesis sets
out to analyze the Danish consumers’ knowledge towards the different food labels. How
recognizable are the different food labels for the consumers, how much do a Danish
consumer know about the different food labels, like what is right and wrong when it
comes to the Danish red organic label. Furthermore what might determine the different
levels of understanding and do consumers have any preferences towards certain labels,
like what labels are more attractive than others.

How well established is the knowledge about the information on food in the Danish
consumers mind and what factors might explain the different degree of
understanding?

To answer the research question several investigation question would be answered


beforehand.

• How many of the different logo and text based labels does a Danish consumer
know?
• How much does a Danish consumer know about each of the different labels,
both logo and text based?
• What kind of information is most well known when it comes to the different
labels?
• What might determine the different levels of understanding?

The aim is to answer all questions in this thesis.

1.2 Definitions
Food label is a rather broad definition and when the term is used in this paper it is either
refers to a logo based item like the green keyhole (in Danish; nøglehullet), the red
organic-label (in Danish; øko-mærket), fair-trade and similar labels which you can find

Page 9 of 70
on the groceries today or text based like best before date, ingredient list or additives.
However if the talk is more specific about either a item like the green keyhole the term
logo based will be used while text based will be used in a situation where the focus is
towards a label like best before date.
The term participant without further explanation refers to the people involved in the
individual interviews while the term respondent refers people involved in the survey
questionnaire.
Concerning the 14 items in the thesis is named as such; _MHT (Mindst holdbar til), _SA
(Sidste anvendelesdato), _BF (Bedst før), _ING (Ingredienslisten), _ADDI
(Tilsætningsstoffer), _KEY (Nøglehulsmærket), _GDA (GDA), _BOB (Beskyttede opri
ndelsesbetegnelser mærket), _ORG (Det danske Økologimærke), _MSC (MSCmærket),
_FAIR (Fairtrademærket), _BGB (Beskyttede geografiske betegnelser mærket), _GTS
(Garanterede traditionelle specialiteter mærket) and _EUORG (Det europæiske
økologimærke).

1.3 Limitations
The thesis has some limitations when it comes to information on the package with
respect to both logo based items and as well as text based items since the thesis is made
in collaboration with the Danish Veterinary and Food Administration who had certain
demands for specific food labels that had to be included . So this thesis is based on 14
different item, nine logo based labels and five text based labels.
Besides that the thesis is restricted to people who are in charge of the shopping for the
household and either consumes fish or chocolate.

1.4 Research approach


The thesis will be based on different literature, empirical material and models. Since
this thesis is about consumers knowledge towards the different food items. A look into
the cognitive dimension of consumer understanding will be used to try to understand
how the different information is obtained in the consumers mind.
A mixture of a framework by Grunert & Mills 2007 which focused on how consumers
responded to nutrition labels and a model from 1961 by Ladvidge and Steiner will be
used in this thesis to develop a framework in which consumers brand awareness,

Page 10 of 70
consumers perception and factors that might influence the brand awareness like
consumer demographics, interest in health, environment and so on will be explained.
To assess the problem with the consumers’ awareness and understanding of the
different information items on food products I will focus on two elements: in-depth
interviews with ten people to get a deeper understanding of the area and use this
information later on for a survey questionnaire. The questionnaire will be used for in
depth analysis and to get a more generalized result of Danish consumers understanding.
This investigation will form an insight into how well established each food label is in
the mind of the consumer and what factors that might determine difference in level of
understanding for consumers.

2. Literature review
Nutrition information is available on almost every grocery today and as a consumer you
can see both voluntary nutrition information like the green keyhole as well as
mandatory information like the ingredient list in Denmark.
However as a qualitative study performed in Europe by Directorate General for Health
and Consumer Protection showed nutrition information is not equally relevant across all
kind of products. Since participants in this study claimed that it is less relevant on fresh
groceries like fruit and vegetables since they are consumed immediately. However
products that are not as transparent as fresh products like ready-meals were of a higher
interest for the participants.

Research on nutrition information on food labels have been thoroughly researched


within the last decade. However most of the studies performed on use and
understanding of nutrition information on food labels have been done in the UK, since
the UK have been the front runner for promoting nutrition labeling on food with special
regards to FOP signposting Cowburn & Stockley (2005); Grunert et al. (2009).

When analyzing determinates of understanding of food labels extensive research have


showed demographic variables such as age, gender and education to have an effect as
shown in studies by Nayga (1996) and Drichoutis, Lazaridis, & Nayga (2005). However
as pointed out in the Grunert et al. (2010) in a study of use and understanding of
nutrition information among consumers in the UK demographic determinates might be

Page 11 of 70
affected by other facts such as interest in health or nutrition knowledge.
Cowburn and Stockley (2005) also revealed that regarding age and its influence on
understanding especially older people had a hard time understanding the different terms
on labels. Furthermore higher level of reading was seen with consumers that had
interest in diet or heath. The effect of age was also revealed in a study by Hawks (2004)
in which she pointed out that older people and lesser educated people are least likely to
understand food labels. Furthermore to back up these finding a study in 2004 by FSA,
found that older age people and the less educated respondents had a lower level of
understanding. However there was a study that showed that older respondents were
more interested due to increased health concerns, however only when health concerns
were present Louriro et al. (2006). A study by Grunert et al. (2009) showed that older
people had less nutrition knowledge however the demographic variables in the study
was partly mediated by interest in health and nutrition knowledge. So studies have
showed that age can have both a negative and but also a positive effect on
understanding. However a study by Nayga (2000) showed that age had no significant
difference.

Several studies have been performed on education as well in a study by Follmer et al.
(1991) showed that lesser educated consumers understand less compared to better
educated consumers. While a study performed in Supermarkets in Greece in 2005 by
Dricoutis et al. showed that education had a positive influence as well. However a study
by Nayga (2000) showed no significant difference on education. A study by
Dallongeville et al. (2001) showed that higher educated people performed better in a
quiz about nutrition and a study by Nayga et al. (1998) showed that consumers with a
bachelor degree or higher are more likely to use nutrition labels than consumers without
any college degree.

In another study by Nayga (2000) it was pointed out that lower income consumers had a
lower level of nutrition knowledge. Another study by Dallongeville et al. (2001) had a
positive effect as well, higher income increased nutrition knowledge. While a similar
study by Hakeem et al. support these findings, in this study the respondents with higher
income had higher nutrition knowledge compared to lower income respondents.
Another study that supports these findings is a study by Nayga et al. (1998) in which
respondents with lower income was less likely to use labels to compare brands.

Page 12 of 70
Furthermore gender has also showed to have an effect on nutrition knowledge a study
by Hendri et al. (2007) showed that woman score higher on nutritional knowledge than
men. However a study performed on College students showed no significant difference
among males and females for nutrition knowledge Berzegari et al. 2011. Other studies
have showed that demographic variables as house size can have a negative effect on use
and understanding of labels Drichoutis et al. (2005).

A study performed in 2009 by Grunert et al. looked on the use and understanding of
GDA and TL labels in six European countries. The result of this study showed that the
consumers had little difficult in understanding FOP nutrition and the GDA label was
best understood by consumers in the UK, Sweden and Germany. With respect to the
GDA a study by Irish FSA (2009) pointed out that many consumers misunderstood the
% GDA as being the percentage of the nutrition in the pack, while the study by Grunert
et al. showed that 39.6% of respondents thought it were per serving while 29.1% meant
it were per 100g.

A study for ANZFA (Australia New Zealand Food Authority) that tried to shed light on
consumer’s awareness, knowledge and understanding of food labels and behavior
towards food labeling. This study concluded that participants were divided on
preferences for either per portion or 100g to be visible. Furthermore the study showed
that there was a widespread understanding of the ingredient list in which all of the
participants could locate the ingredient list and roughly 50% knew that the ingredients
were arranged in a specific order from most to least. The same study showed that origin
of product was very important.

In study by London Economics in association with ADAS and Ecologic (2008) looking
on the two labels PDO (_BOB), PGI (__BGB) showed that the majority 90% of these
labels were found on products originating from six countries Italy, France, which
accounted for more than 40%, while the rest from Spain, Portugal, Greece and
Germany. Furthermore they looked at how many of the respondents that were able to
recognize these labels in all European 27 member states and in Denmark only 3% were
able to recognize these labels. Furthermore the respondents were asked about their
understanding of these two labels by presenting them with nine statements. Overall for

Page 13 of 70
the 27 member states the statements which had the highest score were in fact the four
statements related to the two labels PDO and PGI, in which they scored between 33% to
51% however the score for Denmark alone would be lower since an exclusion of Greece
and Italy showed that respondents that were able to recognize and identify just one of
the four statements reduced the average of recognition and understanding to just 4%.

As mentioned earlier a lot of research has been done on the subject of understanding
and use of nutrition information; however research performed on Danish consumers had
been limited.
A yearly study on Danish consumers knowledge of the green keyhole by Ministry for
Veterinary and Food Administration showed in 2011 that 64% of the respondents were
able to identify lesser fat as one of the correct statements for the green key hole, while
62% knew it concerned lesser sugar, 55% knew it was about more fibers etc. Overall all
four correct statements for the keyhole experienced an increase compared to the same
study performed a year earlier.

In another research done by Fairspeak in May 2010 Danish consumers were asked if
they knew 15 different food labels and if so, if they were able to tell something about it,
and the study showed that the label consumers understood the most were the recycle
label followed by the red organic label (_ORG) with 82.9%, while labels like _GDA
only were understood by 31.9% however it has to be mentioned that the interpretation
of the answer were kind of vague since a some sort of correct answer would be seen as
correct. However the study showed that the even though that vague answers were
interpreted as correct less than 10% of the respondents were able to understand labels
like the _EUORG, _BOB and _BGB labels.
While the green keyhole (_KEY) were only understood by 22.8% of the respondents,
even though you can not directly compare it with the study performed yearly by the
Ministry for Veterinary and Food Administration, it still show a big difference in
understanding for the respondents involved in the two studies.

Turning the attention away from understanding of food labels and look at what
consumer prefer on food products, a study by Bernues et al. (2003) on what kind of
information labeling that consumers preferred for meat products in Europe showed that
nutrition information were rated lower than information of origin and best before date.

Page 14 of 70
While another study looking on food products in general in Ireland by FSA of Ireland
(2003) showed that the top five items preferred by Irish consumers were best before
date, additive content, fat content, production date and origin of product.

As mentioned earlier 14 different labels were included in this study, in which nine of
the labels are voluntary and the last five are mandatory.
The nine voluntary labels contains two items referring to organic products; the Danish
organic label (_ORG) and the European (_EUORG) equivalent, products with these
labels are produced under strict requirements for organic production, in which certain
demands needs to be met during the growth process.
Two other items in the study focus on nutrition in which it should help the consumer to
make a better choice when shopping, so they avoid the unhealthy products and these
items are the green keyhole (_KEY), which is a Nordic nutrition label, and the European
GDA (_GDA) label which have been used in several studies (Grunert et al. 2009).
Three items in the study relates to quality and they all share the same aim, promote and
protect names of quality, agricultural products and foodstuff in Europe i.e. Olive Oil
from specific region in Italy, these items are PDO Protected designation of origin (in
Danish; _BOB), PGI Protected geographical Indication (In Danish; _BGB) and the last
one is TSG Traditional specialty guaranteed (In Danish; _GTS).
The last two voluntary items in the study are items related to sustainability; one is
related to sustainability for fish, the MSC label (_MSC) that guarantees consumers that
if they buy products with this label they support areas with sustainable fishing. The
other one is Fairtrade (_FAIR) which aim is to help producers in developing countries,
to make better trading condition and promote sustainability.

For the mandatory labels, three of them related to expiration date _MHT, _SA and _BF.
The differences between the three items are that _SA is used on products that perish fast
i.e. meat and it is therefore stricter in nature than the other two items and consumers are
advised to throw out items that exceed expiration date. The two items _MHT and _BF
are usually seen on products that can be stored in extended periods and they share the
exact same meaning and consumers are not advised to throw it out even though it
expire, but use their common sense to see if the product looks alright.
The last two items in the survey relates to the ingredient list, one of the items is the
ingredient list (_ING) while the other one is additives in food products (_ADDI). The

Page 15 of 70
ingredient list shows the ingredient in the product in a specific order from least to most
while additives refers to the content of additives in a product.

3. Methodology
To achieve an in-depth and as well as a more generalizable knowledge towards the
Danish consumers perception and understanding of the logo and text based labels the
system approach will form the base of the methodological approach for this thesis even
though the analytical approach will be included relatively to fit under the system
approach.
Back in the days the system approach was seen as a gap between the different fields of
technical, biological and social fields. The system approach tries to solve more complex
problems with problem oriented investigation and at the same time it solves some of the
problems with the two other approaches which in character is either explanatics or
hermeneutics i.e. the analytical approach (explanatics) would trade of in-depth
knowledge for a more generalizable result while the opposite would happen for the
actor approach (hermeneutics) which will be more in-depth of the situation under
investigation without being able to generalize the result.
In other words the system approach can generate knowledge for a certain situation to a
generalized result while it explores the interrelations within the system or system
components Arbnor & Bjerke, 1997, Heldbjerg 2006.

And it is therefore the reason for using the system approach as point of origin to create
an understanding of the individuals’ perceptions and understanding of the different
labels in this thesis, such as best before date and fair-trade.
The choice of method in this thesis for generating in-depth knowledge is in-depth
interviews as it is a good method for creating in-depth knowledge of a specific area. To
be able to get a more generalizable result than achieved through a pure actor approach a
more explanatory approach will be used as a method for data collection and analysis but
still in adaption with the system approach.

The results from the interviews will be used as prior understanding and inspiration for
the survey questionnaire furthermore if the interviews are unable to give insight into all
items in the thesis, input from the literature review and inputs from the Danish

Page 16 of 70
Veterinary and Food Administration will be included. The questionnaire is also the
applied method for data collection and statistical analysis will take place afterwards. By
doing it this way the results will contribute to knowledge that is more generalizable but
still have a sort of deeper understanding, due to the inputs from the in-depth interviews,
and thereby creating a better overall result Arbnor & Bjerke, 1997, Heldbjerg 2006.

3.1 Theoretical foundation


To investigate how well Danish consumer understand the different food labels a
framework was developed to asses consumers understanding of the 14 food items in the
study. The framework was developed with inspiration from a framework by Grunert &
Willis (2007) which aim was to understand how consumers understood nutrition
information provided on food labels. The framework was inspired by the two streams of
analyzing effects of nutrition information for consumers; consumer decision making and
attitude formation and change.

The original model had two extra components search and use, which are excluded from
this survey since it is out of the boundaries of this thesis.
In relation to the original framework exposure is only included as a prerequisite for
consumers to perceive the information, the exposure will be perceived either conscious
or subconscious, which leads to understanding.
Understanding can be divided into two categories subjective and objective
understanding. The distinction between the two types of understanding is that when it
comes to subjective understanding it is when a consumer “think” he understood the
message, while objective understanding refers to how much the consumer understood of
the intended message by the sender. Furthermore understanding and perception is seen
as cognitive dimensions in a model by Lavidge & Steiner (1961) in which they came up
with a model to describe a purchase situation.
An effect of information processing and perception is liking however in this thesis it is
called preferences in which a consumer might like the label, for varies reasons, it might
be easy to understand or simple because they like the logo. Liking is not necessarily
linked to understanding since but can still have an impact on the label, since liking of an
label can have a positive effect on the evaluation of the product carrying the label even
though It might not be understood correctly, in other words peripheral information
processing Eagly, AH. & Chaiken, S. 1993. Furthermore Liking and preferences was

Page 17 of 70
seen as the next step in the model by Lavigde & Steiner (1961) and it described the
affective dimensions of a purchase. Furthermore the last stage in their model is related
to the last step in the model by Grunert & Willis (2007) the use component, which are
excluded in this study as mentioned earlier.
However as seen with the original framework by Grunert & Willis (2007) certain
factors influence perception, understanding and preferences with regards to food labels;
however factors influencing preferences are not included in this survey, which were let
out in the model by Lavidge & Steiner.
The factors included in this study are interest in health, animal welfare, social
responsibility and the environment. Additional consumer demographics are usually
mentioned as determinates of nutrition knowledge as well, see literature review,
however it is more common that they correlate with actual determinates as interest in
nutrition and knowledge about nutrition which was seen in the study by Grunert et al.
(2009).

Figure 1 Conceptual framework

Interests
Demographic variables

Perception
Preferences
Box of Fish
Box of Chocolate
Understanding
Subjective understanding
Objective understanding

3.2 In-depth interviews in theory


The reason for using in-depth interviews as opposed to focus group interviews is that
the goal of the qualitative approach was to get pure dimensions of understanding and
perception of the different labels more than interactions between the individuals.
So even though a focus group would get a more lively discussion going and might
stimulate new ideas and thoughts from the participants the in-depth interviews were

Page 18 of 70
chosen to get an opportunity to dig deeper into the participant’s point of view, since the
purpose of the interviews are to get a deeper understanding.
Another upside with in-depth interviews are that there are more time for each
participant to hear their meaning compared to a focus group interviews since some of
the participants might hinder some of the other participants to speak up, because some
people are more open than others. Furthermore peer pressure is not an issue with in-
depth interviews since the interviewed is performed face to face. A thing to that have to
be mentioned is that in in-depth interviews the participants meaning are not influenced
by other participants which can be seen in focus group situations Boyce & Neale, 2006.
Looking at how participants are gathered for in-depth interviews and focus groups, there
is a difference when looking at the timeframe. For planning and arranging the
interviews it is less time consuming in in-depth situations since you are not depended on
a specific number of people, only one at the time and the location of the interviews is
more flexible to a certain aspect, since it is easier to perform the in-depth interviews in
the home of the participant or at their workplace compared to the more complicated way
in focus groups. But if we however look at conducting the actual interview it is more
time-efficient to perform focus groups interviews since you get a lot of information
within a short-time span Morgan, 1997 & Mack et al. 2005.
So when choosing the appropriate form of interview it is something that has to be
considered. Concerning the right number of interviews a study by Guest et al. 2006 tried
to shed light on the matter, and was able to conclude that around 12 interviews the
amount of information is sufficient enough so you do not have to perform any more.

3.2 Conjoint analysis


According to Green and Srinivasan (1990) conjoint analysis is a method that is used to
uncover consumer’s preferences for a product or service by i.e. part-worths or
importance weights.
Each product has an amount of different attributes i.e. color, size, content and each
attribute usually have different levels i.e. red/blue color, 200g/400g.
Consumers have preferences for each level of the attribute but an optimal combination
for consumers for every attribute, which give the highest level of attractiveness does not
exist and this leads to a trade-off situation in which consumers implicitly evaluate the
attributes and levels, this process is either done conscious or subconscious to an optimal

Page 19 of 70
situation for the consumer, in this case the optimal information for either fish or
chocolate products when it comes to food labels.
Full profile is the preferred method of analysis in this thesis and as the name implies,
the product or service alternatives are presented as complete profiles, consisting of one
level or each of the attributes. Which mean that every attribute in the model are included
in every profile and consumers consider each of the profiles one at the time and either
rate them on some preferences measure or order them by ranking.
Different methods have been used to estimate the value placed on each of the attributes
and levels.
Within a simple design, like two or three attributes, it would be possible for respondents
to evaluate all possible combinations; this is called full factorial design. However this is
rarely possible so researchers has to use fractional factorial design, in which it is
possible to create a smaller design based on a number of attributes and levels which is
still large enough to estimate the utility of each of the attribute levels.

The overall statistic for a conjoint analysis is the following.

    





 = The overall utility for any given attribute (overall utility)

 The individual attribute (Part worth contribution (utility) associated with the j’th
level and the I’th attribute)
  Number of attributes (i.e. size, volume)
  Number of levels for attribute i.

3.3 Sampling method


Addressing the selection of participants there are different techniques with their own
advantages and disadvantages. Overall we talk about two kinds of techniques
probability sampling and non-probability sampling.

Page 20 of 70
Probability sample categorized as a technique in which the participants are chosen
completely random, in other words every individual should have an equal chance to be
picked, overall four kinds of techniques exist Simple random sampling, Stratified
random sampling, Cluster sampling and Systematic random sampling Rasor & Barr,
1998.

Non-probability sample is categorized as a technique in which the participants are not


chosen at random. Techniques within non-probability are purpose sampling, quota
sampling, convenience sampling and snowball sampling (Hart, 2007).
As the time and resources in this thesis is very limited, the sampling technique used for
the in-depth interviews are convenience sampling and snowballing.

Hart (2007) describes convenience sampling as a technique in which the researcher


obtain the sample from an easy accessible population, in this respect snowballing is a
technique closely related to this, in which the participants are asked to refer to other.
While sampling for the questionnaire was done by a market research company YouGov
A/S.

4. In-depth interviews

4.1 Design of In-depth interviews


As based earlier the reason for choosing In-depth interviews was to get a deeper
understanding of how much consumers know about different food labels before using a
questionnaire to generalize the result. Furthermore it was decided to gather 10
participants for interviews and expectation about the subject being an area of low
involvement since the area of FMCG is usually low involvement a reason for this is that
the product is not important for the consumer Silayoi & Speece 2004 so further
interviews would not give any new information and as mentioned a study showed that a
sufficient number of interviews are 12. Moreover it was not possible to give participants
any compensation for their time, since it was not within the resources of the thesis and it
was constrained by a time frame.
As mentioned in the prior paragraph participants for the in-depth interviews were
gathered through a mixture of convenience and snowballing which meant that the
researcher asked friends and family to refer to people in their social circle to be used in
Page 21 of 70
these in-depth interviews, furthermore these participants referred others to participate in
the interviews. The reason for this was to avoid in predefined relationship, in that way
the participant and the interviewer would not have any prior understanding of each
other, and thereby avoiding a common understanding of expressions.
In the interview a semi structured approach were used, meaning that the researcher were
guiding the interview to a certain degree so the interview would stay on track, but still
allowing some discussion of trail. Furthermore the interview had some predefined
topics and tasks the participant had to discuss and complete before the end of the
interview.

4.1.1 Question guide


To begin with the involved participants were presented with a small introduction to the
subject before starting the actual interview, and they were told to say what comes to
mind during the interview, the participants were also informed that their answers would
be anonymous in the project. See appendix A for interview guide

4.1.2 Top-of- mind


To begin with each participant were told to tell when they last time noticed any food
labels on groceries. By letting the participants open op first, any influence by the
researcher would thereby be avoided. After having identified the prior knowledge of
each participant, they were presented for food labels they haven’t mentioned yet, to see
if they could cognize it and had any knowledge about it.

4.1.3 Other information on the products


After having talked about varies food labels on the products, the interview were led in a
direction towards text based items and how well they were able to recognize and
elaborate on these.
The presentation of both logo and text based items were done by presenting the
participants with a product that had the item in question, see appendix B for an example.

4.1.4 Demographic questions


In the end of the interview and as a last exercise each participant were asked to fill out a
small questionnaire with different demographics questions to get an understanding of
the participants’ background.

Page 22 of 70
4.2 Data collection
As mentioned earlier the interview did not have any prior knowledge of the involved
participants, so the ten interviews were conducted in different places spread across
Jutland, two from Aarhus, one from Aalborg, seven from different cities in Thy.
The discussions were taped on a Dictaphone. The interviews took between five to ten
minutes and were transcribed afterwards in keywords, see appendix C for an overview
of the participants and their answers furthermore as expected new information were not
gathered when the last interviews were conducted since it is an area of low involvement.

4.3 Results of in-depth interviews


Below is a short description of the involved participants and their answers, which will
be used later in the survey questionnaire to come up with statements referring to the
understanding of the different food labels, both logo and text items respectively.
All the participants were between the ages of 25 to 82. All ten interviews were done in
the home of the participant or at their workplace.
For seven of the participant they were either responsible for the grocery shopping or
divided with their partner. Education wise the participants were spread from primary
school to university and some of the participants were with job, unemployed or retired.
Six of the participants had children in the age from 4 to 27, so some with children still
living at home. Income wise the participants were also spread on the income scale from
0-100.000 to +500.000.

Regarding the different labels all ten were able to give some explanation towards the
_ORG, what it meant and most of the participants was able to elaborate further about
this logo.

“Hvordan dyrene er behandlet, retningsliner for sprøjtemidler og medicin forbrug,


dyrkningsmetoder, gødningsprodukter, hvor meget der må bruges af gødning”
Participant 2, Male, 51 years.

Overall most of the participants had a clear idea about what it meant. A similar pattern
were seen when focus turned into _KEY in which most of the participants had an idea
on what it meant. Especially lesser fat was mentioned multiple times, but overall
participants knew more about _ORG.

Page 23 of 70
“Fedt indhold” Participant 1, Male, 54 years

“Noget med fedtfattighed”Participant 3, Female, 48 years.

But when it came to questions about _GDA and _BOB the participants were unsure of
the correct meaning, in which some of the participants were not able to say something
about it, since they did not recognize it but one of the participants were able to
recognize it a bit.

”... er lavet i besteme områder af et land, hvor det høre til” Participant 8, Male, 58
years.

Looking at _MSC the participants were able to figure out that it had to be something
with fish, since the logo has a fish on it and some participants like.

“noget med man ikke overfisker, fiskekvoter overholdes” Participant 8, Male, 58 years.

Turning attention to the last logo based element in the interview _FAIR, nine out of the
ten participants were able to give an explanation to what groceries with the fair-trade
logo meant. Over half of the participants meant that one of the things it stood for were
better working conditions and paycheck for the workers.

”Dem der arbejder, får en ordentlig løn” Participant 5, Female, 33 years.

Turning to the text based elements and talking about the three different usages of
expiration date, talking about _MHT most of the participants were able to understand
what it meant, even though some of the participants would throw it out if the date has
been exceeded while others would still use it even though the date was exceeded.

”jeg går aldrig over tiden, det kommer ud..” Participant 4, Female, 63 years.

”dåsetomater vil jeg ikke være så nøje med, kød og lign vil jeg ikke servere efter
overskredet dato”Participant 9, Male, 28 years.

Page 24 of 70
Turning the attention to the second item _SA showed similar answers to _MHT some of
the participants meant it was the same, while others meant that it was a more strictly and
you had to be more careful than reading _MHT.

“På same made som Mindst holdbar til” Participant 5, Female, 33 years.
“Den er lidt mere streng, fordi den siger jo det er den sidste dag du må bruge tingene”
Participant 10, Female, 25 years.

Looking at the last of the three items referring to expiration date _BF the participants
were more divided into the exact meaning since someone took it lighter than _MHT
while a participant meant that all three items had the same meaning meanwhile others
meant that it was not as clear as the other two items regarding the meaning of it, it was
not as concrete.

“den tager jeg lidt mere lempelig på fremfor mindst holdbar til” Participant 3, Female,
48 years.
”Det er det samme igen” Participant 5, Female, 33 years.
” Lidt difus, den siger ikke rigtigt noget bestemt, om hvor langt det kan holde sig og
anvende det” Participant 8, Male, 58 years.

The last items in the interview focused on the ingredient list and additives in groceries.
All of the participants were able to give an explanation to _ING that it showed the
content of the grocery, but not everyone knew if there was a meaning behind the order
of the ingredients.

“Det der er i, rækkefølge, mener ikke der er nogen bestemt rækkefølge, blot e-numre
står sidst” Participant 2, Male, 51 years.

” Hvad den indeholder, rækkefølge, det der er mest af står først” Participant 5, Female,
33 years.

Regarding the last item _ADDI most of the participants were able to recognize additives
in groceries.

Page 25 of 70
”e-numre” Participant 9, Male, 28 years.

Overall it was clear that more well-known logos as _ORG, _KEY and _FAIR were better
known for the involved participants and the overall understanding were fairly equal
however the participants were more confused to the actual meaning of _GDA, _MSC
and _BOB.
Concerning the more text based items most of the participants had some sort of idea
about different variation of expiration date _MHT, _SA and_BF and items referring to
the ingredient list _ING and _ADDI however the actual understanding of these elements
were not as equal as with _ORG, _KEY and _FAIR since the participants meaning of
these items were more divided.

5. Survey questionnaire
As a result of the literature review and the findings in the in-depth interviews a survey
questionnaire were made. First of all since the survey is about a Danish consumer’s
knowledge about the different food labels, the survey questionnaire is in Danish and
therefore is all references to statements in the survey are in Danish since some of the
questions aren’t easily translatable.

5.1 Design of survey questionnaire


In order for the respondents to give qualified answers to the questionnaire regarding text
and logo based labels on the packing, they would naturally have to be involved in the
shopping for the household and either eat fish or chocolate, since we want to check if
the choice of product would make a difference in preferences for the respondents
towards the different items in the survey. This meant that respondents were either
showed pictures with fish or chocolate, while the rest of questionnaire remained the
same.

5.1.1 Design
Overall the questionnaire contained four different sections, understanding of the
different food labels, background information, preferences of food labels and interest in
certain subjects.

Page 26 of 70
As the first task in the questionnaire the respondents were asked several demographic
questions since demographic variables have showed to determinant variation in
understanding of food labels as demonstrated in the literature review. The demographic
variables included in the questionnaire were the following sex, age, allergy, amount of
people in the household, number of people living in the household below the age of 12,
income, education, job, height and weight. Furthermore a do not wish to answer option
was included in some of the questions since some of the questions might be perceived
as sensitive.

The next task for the respondents were to answer the conjoint analysis, which is the
preferences part of the study and as mentioned earlier preferences are not necessarily
depended on understanding.
In a conjoint analyses it is important to measure only attributes that matter and are
realistic for a consumer, however it is possible to include levels that consumers are not
used to see, whether it is a higher price or bigger volume etc.
So in order to fulfill the requirement of realistic alternatives two separate conjoint
analyses were made, since not all food labels exist on one kind of grocery and as
mentioned earlier to see if there is a different in preferences for food labels among
choices of product.

So the design of the conjoint analyses was set to be a 3x2x2x2x2x2x2x2x2 orthogonal


main effects plan (OMEP) for the Fish survey and a 3x2x2x2x2x2x2x2 orthogonal main
effect plan (OMEP ) for the chocolate survey. However since the number of attributes
were so high a fractional-factorial design was made with 20 profiles see outputs from
the two conjoint analysis on the CD, under output 2 for fish and output 3 for chocolate
which included 4 holdout cases to check for validity.

Respondents were asked 20 times to rate each profile on a scale from 1 (Not important
at all) to 7 (Really important). The reason for choosing rating as the method for
evaluating the different profiles as opposed to ranking is that the two conjoint analysis
contains 20 profiles each and therefore would ranking not be useful since the rank order
approach is only usable when few profiles are evaluated. Because it would be
impossible for respondents to say which alternative they prefer over another with 20
different profiles Green & Srinivasan, 1978. Furthermore respondents are able to

Page 27 of 70
express indifference among alternatives by giving them identical rating. For this reason,
the rating method offers more benefits to the research Boyele et al. 2001.

Table 1 - Attributes and levels


Fish Chocolate
Attribute Levels Information level Attribute Levels Information level
1 Expiration date 3 No label, _MHT, _SA Expiration date 3 No label, _MHT, _SA
2 _ING 2 No label, label _ING 2 No label, label
Origin of Origin of
3 product 2 No label, label product 2 No label, label
4 _ADDI 2 No label, label _ADDI 2 No label, label
5 _KEY 2 No label, label
6 _GDA 2 No label, label _GDA 2 No label, label
7 _BOB 2 No label, label _BOB 2 No label, label
8 _ORG 2 No label, label _ORG 2 No label, label
9 _MSC 2 No label, label
10 _FAIR 2 No label, label

The design of the stimuli for the two conjoint analysis was made from a picture of either
a box of chocolate or a box of fish, labels with the above mentioned attributes. This
visualization of an actual product would help respondents to achieve the right mindset
for the choice-stimuli.
An example of one of the stimuli’s can be found in appendix D.

The third task in the questionnaire was Understanding with regard to all 14 items in the
survey and there was distinguished between subjective and objective understanding and
as mentioned earlier understanding can be distinguished between objective
understanding and subjective understanding.
Subjective understanding refers to what a respondent think he knows about any given
subject, and to see if the respondents ranked their understanding higher than it actually
was, to avoid any influence from the next part of the survey, the respondents were asked
to do the subject task first, in which the respondents were asked to rate how well they
meant they understood the following item 14 items by showing them a picture of the
logo in question i.e. _KEY on a scale from 1 (Understand nothing at all) to 7
(Understand everything). After this task was completed the respondents were asked
about their objective understanding.

Page 28 of 70
Objective understanding of the 14 items were measured by presenting the respondent
with the same picture as in subjective understanding however for each picture different
statements related to this item were present i.e. for _KEY the respondents were asked if
the following statements maksimumgrænse for indhold af fedt, maksimumgrænse for
indhold af mættet fedt, maksimumgrænse for indhold af fiber og fuldkorn,
maksimumgrænse for indhold af sucker, maksimumgrænse for indhold af salt, økologisk
production, ingen tilsætningsstoffer and minimumgrænse for indholdet af vitaminer og
mineraler were either true, false or they did not know when it came to this label. The
picture used for _KEY can be seen in appendix E, furthermore it has to be mentioned
that all pictures used for all 14 items were of a size so it would be possible to read the
text presented on the logos.

As mentioned earlier the different statements in this section of the survey were inspired
mainly by the in-depth interviews, however as mentioned earlier as some of the labels
were lesser known for the participants involved in the in-depth interviews, so the
statements are also based on the literature review as well as input from the Danish
Veterinary and Food Administration. In the end every statement was also approved by
them.
Furthermore a do not know option for each statement were added since respondents
might be in doubt about the rightful answer, they will either look for way to avoid
answering the question Acock 2005, or simply mark it random, in this case as true or
false, and thereby not the true value of their knowledge towards the specific item.
Furthermore previous research had acknowledged a “don’t know” answer as an
indicator for low involvement Ray 1973 in Zaichowsky 1985.

As mentioned in the literature review interest in certain subjects might determine the
variation in understanding of food labels so as a last task in the questionnaire the
respondent were asked about their interest in health, social responsibility, environment
and animal welfare and they were all measured on a scale from 1(Completely disagree)
to 7 (Completely Agree).
The health questions were based on a study by Chrysochou et al. (2010), animal welfare
questions were based on statements from Krystallis et al. (2009), and questions referring
to social responsibility were made by inspiration from Bezencon & Blili (2008).

Page 29 of 70
However the questions about the environment were based on questions taken from the
NEP-scale by Dunlap et al. (2000) and Krystallis et al. (2009).

As a side mark it has to be mentioned that during the period item _EUORG , _GTS and
_BGB were added to the understanding part of the questionnaire however as mentioned
earlier in the interviews not all item were equally recognizable, so it would probably not
have led to more information from the participants while origin of product was dropped
by the Danish Veterinary and Food Administration so the information gathered in the
in-depth interviews about origin of product were not used further.

5.1.2 Pilot test


For each of the individual parts of the survey a pilot test were performed, to test
readability and checking for errors in the system before sending out the final
questionnaire. These evaluations resulted in some minor changes for a couple of
question to minimize confusion and the addition of numbers of the answering scales to
help respondents choose more easily since the survey is quite long (+30 min) and the
interest questions were checked for internal consistency and all groups of interest had
scores above 0.6 which is a good fit of internal consistency between the questions
Nunnally & Bernstein 1994.

5.2 Data collection through survey questionnaire


The questionnaire was sent out by the help of YouGov as mentioned earlier. The data
was collected using the online survey tool, Qualtrics, from January 26th to February 6th.
The total sample contained 646 respondents

5.2.1 The sample


See the next page

Page 30 of 70
5.2.1.1 Demographic characteristics of the sample
Table 2 Socio-demographic characteristics of the sample
N %
Total n 646
Gender
Male 294 45,5
Female 352 54,5
Age
18-35 205 31,7
36-50 223 34,5
51-65 218 33,7
Education**
Primary school 50 7,7
High School 99 15,3
Technical Education 206 31,9
University Degree or higher 287 44,4
Position
Working 388 60,1
Selfemployed 25 3,9
Student 91 14,1
Pensionist 70 10,8
Other 29 4,5
Unemployed 43 6,7
BMI*
0-24,99 307 48,8
25-29,99 199 31,6
30+ 123 19,6
Allergy**
Yes 68 10,5
No 563 87,2
Income**
0-399.999 218 33,7
400.000-799.999 243 37,6
800.000+ 104 16,1
Household size
1 147 22,8
2 288 44,6
3+ 211 32,7
Children under 12
0 510 78,9
1 74 11,5
2+ 62 9,6
*Some of the respondents did not answer questions about weight and height
**Respondents who answered "don't know" or “won’t say” are not
included

Page 31 of 70
45.5% were males and 18 to 65 years old. 4.1%-points less than the Danish population
(Statistics Denmark) and 54.5% were females 18 to 65 years old. The sample population
is fairly consistent with the population of Denmark 31.7% were 18 to 35 years old
1.4%-points lower than the Danish population. 34.5% were age 36 to 50 years old
0.2%-points less than the Danish population while 33.7% were 51 to 65 years old 2.3%-
points more than the population of Denmark. The average age of the sample was 42.9
years.

22.8% of the respondents lived alone 15.7%-points below the Danish population, while
77.2% lived with others. 32.7% live with at least two others 4.2%-points more than
Danish population.
In the sample 21.1% had at least one child below the age of 12 in the household while
the remaining 78.9% were without children below the age of 12.

Regarding education, the sample is more educated than the Danish population, in the
sample 7.7% had attended primary school as their highest completed education 21.7%-
points lower than then Danish population. 15.3% had marked high school as their
highest education in the sample while the Danish population was 6.3%-points lower.
Concerning a technical education 31.9% had this as the highest education 4%-points
lower than the Danish population. University degree or higher 44.4% of the sample had
this as the highest completed education 22.6%-points higher than the Danish population.

The average BMI in the sample was 26.42 slightly overweight compared to the BMI-
scale, since overweight starts at 25.00. It has to be mentioned that the calculations
regarding BMI had a sample size of 629 since some people did not want to answer
either weight/height or both. The sample divided into the different groups on the BMI-
scale was as the following. 48.8% were in the normal weight category, 31.6% were
overweight while the remaining 19.6% were obese. Compared to a rapport by the
national board of health, the numbers aren’t that far apart since they reported that 47%
were overweight 4,2%-points less than in this sample.
In the sample 10.5% of the people had some sort of food allergy.
In the sample 33.7% had an income lower than 399.999DKK, 37.6% had an income in
the region 400.000-799.999DKK while 16.1% had an income above 800.000DKK.
12.5% neglected to answer this question.

Page 32 of 70
5.3 Results of survey questionnaire

5.3.1 Data cleaning


Prior to any analyzing the data was cleaned.

5.3.1.1 Merging dataset


Before any data cleaning could proceed a X2-test was performed on the demographic
variables GENDER, AGE, ALLERGY, H_SIZE, CHILD, INCOME, EDU, JOB,
HEIGHT and WEIGHT to check for any significant difference between the two datasets
(The chocolate survey and Fish survey). And no significant difference were found on
the 0.05 level, so the two datasets were merged, see output 1 on CD.

5.3.1.2 Missing data


Throughout the survey there was a lot of missing data to begin with, the total dataset
contained 759 cases in which 317 were obtained in the chocolate survey while 442 were
obtained in the fish survey. Out of the 759 cases 113 cases were not finished; most of
them gave up halfway through the survey. Out of these cases which did not finish 60 of
them belonged to the fish survey while the rest 53 belonged to the chocolate survey; all
113 were deleted since the dataset was so large. The deleted cases included one lazy
respondent (same response throughout the whole survey). The overall goal was still
obtained (250 respondents for each survey) with 264 for chocolate and 382 for fish.
Looking at other variables for missing data it showed HEIGHT had four missing values
(0.6%) but as thought upon earlier it might be a sensitive subject and thereby are some
people avoiding answering it, the same could be said for the variable WEIGHT in which
there was a total of 17 missing values (2.6%). Both under 5% so it wouldn’t be
discussed further just being accepted since it that important to the analysis and it won’t
have any big influence on the results, because of the large sample size (N=646)
Tabachinck & Fidell. Even though a Little’s MCAR test was run to check if the missing
values were missing by random or not, and it showed the values to be missing at
completely random (Chi-Square=, 100. DF = 1, Sig. =,751).
The reason for low amount of missing values is due to the way the online questionnaire
was constructed, since respondents were forced to fill out every question before they
could move on, so it was mainly due to the length of the survey in which 113
respondents did not finish the survey and even though they were forced to answer only
one respondent had marked the same throughout the survey.

Page 33 of 70
5.3.1.3 Outliers
First of all the dataset was checked for univariate outliers, z-values were produced for
all variables in the dataset. Scores with values outside the range of 3.29 (p <.001, two-
tailed test) were seen as potential outliers but upon checking it out, it was decided that it
was legit cases, just extremes and the number of outliers were really few under 5%.

5.3.1.4 Recoded data


Some variables had to be recoded before analyzing. Three of the demographic variables
had to be recoded into comprehensible data since they were given a different value than
the original value, so the following variables were transformed into their real value by
adding different values for each variable the three values were HEIGHT, WEIGHT and
AGE
While four of the interest variables ANIMAL_6, ANIMAL_7, ANIMAL_8, ENVIRO_1
and ENVIRO_3 had to be reversed as well because of the way the four questions were
asked. All cronbach’s alpha (α) was above 0.60 for the different interest groups which
are a good fit of internal consistency between the questions Nunnally & Bernstein 1994.
For an overview of the different values for each of the five different interest scales look
at appendix 2 on cd.

5.3.2 Understanding

5.3.2. 1 Subjective understanding


As mentioned earlier the respondents were asked about their subjective understanding
of the 14 items in the survey, and the three items in the questionnaire referring to
expiration date _MHT, _SA and _BF scored high on the scale, in which a total of
between 67.7% to 76.9% of the respondents meant they knew perfectly, while the three
items _BOB, _GTS and _BGB scored the lowest of all items, since less than 5% of the
respondent meant they knew all about these three items.
On average respondents valued four items _SA, _MHT, _BF and _ORG higher than 6 on
the scale.

Page 34 of 70
Table 3 - Subjective understanding
% - of respondents Average scores for
with answered 7 respondents
_GDA 43.7%5 5.895
_ING 26.0%7 5.377
_SA 75.7%2 6.642
_MSC 9.3%10 3.2810
_MHT 76.9%1 6.651
_KEY 23.4%8 4.299
_FAIR 39.9%6 5.556
_ADDI 10.2%9 4.398
_BF 67.6%3 6.433
_GTS 2.6%14 2.2412
_ORG 58.2%4 6.234
_BOB 3.6%12 2.3111
_EUORG 3.7%11 0.6413
_BGB 3.1%13 2.3111

5.3.2.2 Objective understanding


Out of the 14 different items (text and logo based) the respondents were asked about,
the one that most respondents guessed correct were _GDA. In which a number of 337
respondents were able to answer correctly to all six statements, which mean only around
52% of the respondents were able to give the “perfect” answer. The element who the
respondents were least familiar with were _BGB where not a single respondent could
give a correct answer to all six statements, closely followed by _EUORG and _BOB
with four and six respondents guessing correct, less than one percent of the respondents
in the survey.
If we, however, look on the average of correct answer for respondents for each of the 14
items the graph looks a bit different compared to the graph above which show number
of perfectly correct answers. _GDA still got the highest score with 5.01 and are
followed by four relatively close items _ING, _SA, _MHT and _BF with 4.29, 4.16, 3.95
and 3.82 respectively.
While especially _ORG strikes out, remarkably good compared to the graph above from
an 11th place to a 7th place, while the opposite happened for _MSC from a 4th place to a
10th place.

Page 35 of 70
Table 4 - Objective understanding
% - of respondents Average scores for
with perfect scores respondents
_GDA 52.2%1 5.011
2
_ING 36.7% 4.292
_SA 22.9%3 4.163
_MSC 17.2%4 2.2810
_MHT 13.0%5 3.954
6
_KEY 8.7% 2.828
_FAIR 8.4%7 3.686
_ADDI 7.7%8 2.709
_BF 7.3%9 3.825
10
_GTS 6.5% 1.2211
_ORG 1.2%11 3.017
_BOB 0.9%12 1.0513
_EUORG 0.6%13 0.6414
14
_BGB 0.00% 1.2012

To get a deeper understanding of the individual items the next paragraph will focus on
the individual statements for each of the 14 items in the survey. The following tables
shows how many of the respondents in percentage who answered each of the three
categories, a correct answer is marked with bold, the paragraph starts with the nine
items referring to FOP labels followed by the five items related to the backside of the
package

FOP Labels
_GDA
Looking at the different FOP labels in the survey, the item that was most recognizable
by the 646 respondents were _GDA. Which a total of 52.5% had guessed perfectly
correctly, furthermore it was the item with the highest mean with a score of 5.01.
Looking at the different statements for _GDA in table 4 the statement respondents were
most familiar with were statement five mærket garanterer, at fødevaren er sund in
which a total of 89.94% of the respondents had marked it as false, closely followed by
two other false statements, statement six mærket garanterer, at fødevaren ikke
indeholder tilsætningsstoffer and statement four mærket garanterer, at fødevaren har et
lavt indhold af fedt with scores of 88.85% and 87.00% respectively.
In fact the two statements the respondents were almost least familiar with were the two
true statements, statement one mærket viser fødevarens næringsindhold i en portion and

Page 36 of 70
statement two mærket viser fødevarens næringsindhold i procent af det daglige
anbefalede indtag for en voksen person which had scores of 75.08% and 80.35%
respectively. It was only the false statement three mærket viser fødevarens
næringsindhold pr. 100 g that scored equally as low with a score of 79.57%.
Overall most of the respondents had an idea about what _GDA meant since it was below
9% of the respondents that had marked the “don’t know” category for all six statements.

Table 5 - _GDA (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
75,08% 18,42% 6,50%
Mærket viser fødevarens næringsindhold i en portion. (1)
Mærket viser fødevarens næringsindhold i procent af det
80,34% 12,69% 6,97%
daglige anbefalede indtag for en voksen person. (2)
14,09% 79,57% 6,35%
Mærket viser fødevarens næringsindhold pr. 100 g. (3)
5,57% 87,00% 7,43%
Mærket garanterer, at fødevaren har et lavt indhold af fedt. (4)
2,79% 89,94% 7,28%
Mærket garanterer, at fødevaren er sund. (5)
Mærket garanterer, at fødevaren ikke indeholder
2,79% 88,85% 8,36%
tilsætningsstoffer. (6)

_MSC
Looking at _MSC in which a total of 17.2% of the respondents had marked perfectly,
had an average of 2.28 correct statements among the involved respondents.
Looking at the individual statements for _MSC , the statements with the highest percent
of correct are the two true statements, statement one der tages i forbindelse med
fiskeriet hensyn til fiskebestanden and statement two der tages i forbindelse med
fiskeriet hensyn til havets økosystemer with scores of 48.1% and 44.4% respectively.

Looking at the four false statements, starting with the highest scoring statement, 38.08%
had guessed statement five fødevaren tæller med i det anbefalede indtag af fisk as false,
while 37.15% had correctly marked statement six fødevaren indeholder ingen
tilsætningsstoffer as false. While the last two statements three and four were only
guessed correctly as false by roughly a third of the respondents.

Furthermore if we look at how many of the respondents who were not able to give an
answer to either statement about _MSC it shows that around 50% had marked the
“don’t” know” category for each statement.

Page 37 of 70
Table 6 - _MSC (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
Der tages i forbindelse med fiskeriet hensyn til fiskebestanden.
48,14% 2,01% 49,85%
(1)
Der tages i forbindelse med fiskeriet hensyn til havets
44,43% 5,26% 50,31%
økosystemer. (2)
Fisken er fanget i et område med ingen eller meget lav
10,37% 28,95% 60,68%
forurening. (3)
12,07% 31,27% 56,66%
Fisken er af høj kvalitet. (4)
9,60% 38,08% 52,32%
Fødevaren tæller med i det anbefalede indtag af fisk. (5)
4,18% 37,15% 58,67%
Fødevaren indeholder ingen tilsætningsstoffer. (6)

_KEY
The next item _KEY only had 8.7% of the respondents guessing it correct. While on
average respondents were able to answer 2.82 statements.
A quick overview of the eight statements in table 6 involving _KEY it is clear that most
of the respondents were clueless when it came to what was true or false since roughly
50% of the respondents had marked the “don’t know” category.

If we however look at how many of the respondents that were able to give a correct
answer, it shows that the statement, in which most respondents were able to answer
correctly were statement six økologisk production in which a total of 44.58% of the
respondents had marked as false, the second highest score were statement one
maksimumgrænse for indhold af fedt were guessed correct by 37.31%, while the third
highest statement, with a score of 37.15% were statement number four
maksimumgrænse for indhold af sukker.

The fourth highest scoring statement, were a false statement in which 35.60% of the
respondents knew that statement seven ingen tilsætningsstoffer was false, the fifth
highest score belonged to statement three minimumgrænse for indhold af fiber og
fuldkorn in which a total of 33.75% of the respondents had marked it as true. For the
last three statements around 30% had guessed them correctly however statement five
maksimumgrænse for indholdet af salt had the highest number of respondents marking
it wrongly with 14.09% marking it as false even though it was true.

Page 38 of 70
Table 7 - _KEY (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
37,31% 10,99% 51,70%
Maksimumgrænse for indhold af fedt. (1)
31,89% 12,23% 55,88%
Maksimumgrænse for indhold af mættet fedt. (2)
33,75% 13,93% 52,32%
Minimumgrænse for indhold af fiber og fuldkorn. (3)
37,15% 9,91% 52,94%
Maksimumgrænse for indhold af sukker. (4)
31,58% 14,09% 54,33%
Maksimumgrænse for indhold af salt. (5)
6,35% 44,58% 49,07%
Økologisk produktion. (6)
7,59% 35,60% 56,81%
Ingen tilsætningsstoffer. (7)
9,91% 30,19% 59,91%
Minimumgrænse for indholdet af vitaminer og mineraler. (8)

_FAIR
The item _KEY was closely followed by _FAIR with percentage of respondents who
understood it perfectly with 8.4% and on average respondents guessed 3.68 correctly.
Looking at the individual statements for _FAIR most of the respondents were confident
about statement one bønder og arbejdere får en rimelig løn and statement two bønder
og arbejdere har rimelige arbejdsvilkår theses two statements that both were true were
guessed correctly by 84.06% and 85.60% respectively while the last true statement,
statement three der tages hensyn til miljøet under produktionen af fødevaren were only
guessed correct by 39.01% of the respondents in fact 32.20% thought it was false.

Looking at the statements that were false in relation to _FAIR 54.64% of the
respondents were able to identify statement four fødevaren er af høj kvalitet as false
even though 15.94% thought it was true. Statement five overskuddet fra salget af
fødevaren går til U-lande were marked correctly by 45.98% of the respondents,
however 23.07% thought it was true. For the last statement about _FAIR statement
number six fødevaren er økologisk produceret 58.51% knew that this was false while
8.51% of the respondents thought it was true.
Apart from the two high scoring statements one and two, roughly 30% had marked the
“don’t know” category for the remaining four statements.

Page 39 of 70
Table 8 - _FAIR (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
84,06% 3,10% 12,85%
Bønder og arbejdere får en rimelig løn. (1)
85,60% 2,32% 12,07%
Bønder og arbejdere har rimelige arbejdsvilkår. (2)
39,01% 32,20% 28,79%
Der tages hensyn til miljøet under produktionen af fødevaren. (3)
15,94% 54,64% 29,41%
Fødevaren er af høj kvalitet. (4)
23,07% 45,98% 30,96%
Overskuddet fra salget af fødevaren går til U-lande. (5)
8,51% 58,51% 32,97%
Fødevaren er økologisk produceret. (6)

_ORG & _EUORG


If we turn our attention to the two items referring to organic products _ORG and
_EUORG they both scored low with only 1.2% who understood _ORG perfectly and
0.6% for _EUORG.
However the difference were larger for the average scores of the two items, since
respondents on average guessed 3.01 correctly for _ORG, while respondents on average
guessed less than one statement correct for _EUORG with a score of 0.64.

Looking at the individual statements for _ORG the two statements which were true were
marked correct by 42.11% for statement one fødevaren er produceret med minimal brug
af særligt tilladte pesticider, while 43.96% had marked statement two fødevaren er
produceret med øget fokus på dyrevelfærd correct. However as seen in table x roughly
around 40% had wrongly marked them as false.

Looking at the four false statement for _ORG, the statement which had the highest score
by the respondents among all six statements were statement four fødevaren er af ekstra
høj kvalitet were a total of 65.33% of the respondents were able to identify it as false.
However there were still 18.11% of the respondents who thought it was true.
Statement fivefødevaren indeholder ikke tilsætningsstoffer divided the respondents
equally between respondents who thought it was true and respondents who thought it
was false with 41.18%. Concerning statement six fødevaren er produceret i Danmark
33.59% had wrongly marked it as true while 45.98% knew it was false. The last false
statement for _ORG were statement three fødevaren er produceret helt uden pesticider
in which a total of 62.54% of the respondents had wrongly marked as true and only
21.67% of the respondents knew it was false.

Page 40 of 70
Table 9 - _ORG (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW

Fødevaren er produceret med minimal brug af særligt


42,11% 39,63% 18,27%
tilladte pesticider. (1)
43,96% 40,09% 15,94%
Fødevaren er produceret med øget fokus på dyrevelfærd. (2)
62,54% 21,67% 15,79%
Fødevaren er produceret helt uden pesticider. (3)
18,11% 65,33% 16,56%
Fødevaren er af ekstra høj kvalitet. (4)
41,18% 41,18% 17,65%
Fødevaren indeholder ikke tilsætningsstoffer. (5)
33,59% 45,98% 20,43%
Fødevaren er produceret i Danmark. (6)

Turning out attention to the other item for organic products the European equivalent to
the Danish _ORG which had less than one correct answer on average _EUORG most of
the respondents had marked the “don’t know” category above 80% for all six
statements. However 15.94% of the respondents were able to mark statement six der
tages hensyn til miljøet under produktionen af fødevaren as true. Concerning the other
two true statements for _EUORG only 8.98% were able to mark statement three
fødevaren er økologisk produceret as correct, while 5.11% had marked statement two
Fødevaren er produceret med øget fokus på dyrevelfærd correct.

Looking at the false statements for _EUORG the statement in which most respondents
were able to answer correctly was statement one fødevaren har et lavt indhold af fedt,
sukker og salt where a total of 12.85% of the respondents had marked it as false. While
10.84% of the respondents were able to mark statement four fødevarens kvalitet er
relateret til et geografisk afgrænset område correctly as false, roughly the same amount
of respondents were able to mark statement five fødevaren indeholder ikke
tilsætningsstoffer as false.

Page 41 of 70
Table 10 - _EUORG (% of respondents who have marked each category, Bold is the
correct answer)
DON'T
TRUE FALSE
KNOW
2,48% 12,85% 84,67%
Fødevaren har et lavt indhold af fedt, sukker og salt. (1)
5,11% 7,89% 87,00%
Fødevaren er produceret med øget fokus på dyrevelfærd. (2)
8,98% 9,29% 81,73%
Fødevaren er økologisk produceret. (3)
4,95% 10,84% 84,21%
Fødevarens kvalitet er relateret til et geografisk afgrænset område. (4)
4,49% 10,68% 84,83%
Fødevaren indeholder ikke tilsætningsstoffer. (5)
15,94% 3,25% 80,80%
Der tages hensyn til miljøet under produktionen af fødevaren. (6)

_GTS, _BOB and _BGB


Concerning the last three items _GTS, _BOB and _BGB the outcome of the survey
showed that out of these three items, the item respondents understood the best was
_GTS in which a total of 6.5% were able to give a correct answer for the six different
statements. For the other two items _BOB and _BGB it was less than 1% of the
respondents who were able to give a correct answer 0.9% and 0% respectively.
Looking at the average score of the three items, they scored 1.22 for _GTS, 1.05 for
_BOB and 1.20 for _BGB. The only item who scored lower was _EUORG.

Looking at them individually starting with _GTS around 70% of the respondents
marked the “don’t know” category.
Looking at the two true statements for _GTS only 13.31% of the respodents were able to
identify statement six fødevaren er ikke fremstillet i et afgrænset geografisk område,
men er traditionel og forskellig fra lignende fødevarer as correct while a quarter of the
respodents (25.23%) were able to identify statement five Fødevaren er fremstillet af
traditionelle råvarer efter traditionelle produktionsmetoder eller har en traditionel
sammensætning as true as well. Looking at the false statements, statement one, two and
three were roughly guessed correctly by 18% of the respondents, while 30.03% of the
respondents were able to guess statement four fødevaren er økologisk produceret as
false. Overall most of the respondents did not have an idea of the right answer since
roughly 70% had marked the “don’t know” category for each statement.

Page 42 of 70
Table 11 - _GTS (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
Fødevaren både stammer fra og er fremstillet og forarbejdet i et
10,37% 17,03% 72,60%
'bestemt geografisk område. (1)
Fødevaren stammer fra og er enten fremstillet eller forarbejdet i
9,91% 18,11% 71,98%
et bestemt geografisk område. (2)
Kvaliteten af fødevaren er relateret til et geografisk afgrænset
9,60% 18,73% 71,67%
område. (3)
1,08% 30,03% 68,89%
Fødevaren er økologisk produceret. (4)
Fødevaren er fremstillet af traditionelle råvarer efter traditionelle
25,23% 4,64% 70,12%
produktionsmetoder eller har en traditionel sammensætning. (5)
Fødevaren er ikke fremstillet i et afgrænset geografisk område, men
13,31% 12,54% 74,15%
er traditionel og forskellig fra lignende fødevarer. (6)

Looking at _BGB, which none of the respondents were able to guess perfectly around
70% of respondents had marked the “don’t know” category for each statement just as
with _GTS.
Looking at the individual statements for _BGB starting with the two true statements
roughly the same number of people were able to guess them correctly with 20.74% and
18.89% for statement two fødevaren stammer fra og er enten fremstillet eller
forarbejdet i et bestemt geografisk område and statement three kvaliteten af fødevaren
er relateret til et bestemt geografisk område respectively.
Turning the attention to the four false statements the spread of correctness was larger
with score of 5.57% for statement one fødevaren både stammer fra, er fremstillet og
forarbejdet i et bestemt geografisk område to 30.34% for statement four fødevaren er
økologisk producer.

Page 43 of 70
Table 12 - _BGB (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
Fødevaren både stammer fra, er fremstillet og forarbejdet i et
26,47% 5,57% 67,96%
bestemt geografisk område. (1)
Fødevaren stammer fra og er enten fremstillet eller forarbejdet
20,74% 9,75% 69,50%
i et bestemt geografisk område. (2)
Kvaliteten af fødevaren er relateret til et bestemt geografisk område.
18,89% 12,23% 68,89%
(3)
1,86% 30,34% 67,80%
Fødevaren er økologisk produceret. (4)
Fødevaren er fremstillet af traditionelle råvarer efter traditionelle
7,89% 19,97% 72,14%
produktionsmetoder eller har en traditionel sammensætning. (5)
Fødevaren er ikke fremstillet i et afgrænset geografisk område, men
3,72% 24,61% 71,67%
er traditionel og forskellig fra lignende fødevarer. (6)

For the last item _BOB it was again around 70% respondents like with _GTS and _BGB
who were unable to give a true or false answer to the six statements and had therefore
crossed off the “don’t know” category.
As seen with _GTS and _BGB a total of six statements were present with two true
statements and four false statements.
For the two true statements 19.50% were able to marked statement one fødevaren både
stammer fra og er fremstillet og forarbejdet i et bestemt geografisk område as correct
while 15.94% of the respodents were able to mark statement three kvaliteten af
fødevaren er relateret til et bestemt geografisk område correctly. As seen earlier with
_BGB the spread of correct score by the respondents is more divided when it comes to
the false statements in which the amount of respondents who are able to mark them
correctly as false are from 8.36% for statement two fødevaren stammer fra og er enten
fremstillet eller forarbejdet i et bestemt geografisk område to 27.55% for statement four
Fødevaren er økologisk produceret.

Page 44 of 70
Table 13 - _BOB (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
Fødevaren både stammer fra og er fremstillet og forarbejdet i et
19,50% 6,35% 74,15%
bestemt geografisk område. (1)
Fødevaren stammer fra og er enten fremstillet eller forarbejdet i et
17,80% 8,36% 73,84%
bestemt geografisk område. (2)
Kvaliteten af fødevaren er relateret til et bestemt geografisk område.
15,94% 11,15% 72,91%
(3)
1,08% 27,55% 71,36%
Fødevaren er økologisk produceret. (4)
Fødevaren er fremstillet af traditionelle råvarer efter traditionelle
10,68% 14,09% 75,23%
produktionsmetoder eller har en traditionel sammensætning. (5)
Fødevaren er ikke fremstillet i et afgrænset geografisk område, men
5,73% 19,50% 74,77%
er traditionel og forskellig fra lignende fødevarer. (6)

The backside of the package


Turning the attention to information you would usually find on the backside of the
package items and start by looking at the three items referring to expiration date _SA,
_MHT and _BF.
Out of these three items _SA had the highest percentage of correct respondents with
22% and on average respondents had 4.16 correct answers.

Looking at the individual statements for _SA the statement with the highest percentage
was a true statement were 83.59% of the respondents were able to mark statement one
kvaliteten af fødevaren er i orden til og med denne dato as correct. The statement with
the second highest percentage was another true statement were 76.93% of the
respondents were able to mark statement two fødevaren skal være anvendt senest denne
dato as true. The last true statement which was guessed correct by 68.58% of the
respondents were statement five Fødevaren kan efter overskridelse af datoen udgøre en
sundhedsrisiko.
Turning the attention to the three false statements 68.89% of the respondents were able
to mark statement six uanset om fødevarens emballage åbnes, kan fødevaren altid
spises indtil denne dato as false.
For statement three det betyder det samme som "Mindst holdbar til” 60.99% of the
respodents were able to mark it as false while 57.28% had marked statement four
fødevaren kan spises efter den angivne dato, såfremt den lugter normalt og ser fin ud as

Page 45 of 70
false. However for both of these statements around 30% thought that these statements
were true.

Table 14 - _SA (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
83,59% 10,37% 6,04%
Kvaliteten af fødevaren er i orden til og med denne dato. (1)
76,93% 16,72% 6,35%
Fødevaren skal være anvendt senest denne dato. (2)
30,80% 60,99% 8,20%
Det betyder det samme som "Mindst holdbar til". (3)
Fødevaren kan spises efter den angivne dato, såfremt den lugter
29,72% 57,28% 13,00%
normalt og ser fin ud. (4)
Fødevaren kan efter overskridelse af datoen udgøre en
68,58% 16,41% 15,02%
sundhedsrisiko. (5)
Uanset om fødevarens emballage åbnes, kan fødevaren altid spises
17,18% 68,89% 13,93%
indtil denne dato. (6)

Looking at the second item referring to expiration date _MHT in which 13% of the
respondents guessed correct and with an average of 3.95 statements correct for each
respondent.
The two statements with the highest percentage were the two true statements. For
statement one Kvaliteten af fødevaren er i orden til og med denne dato a total
percentage of 82.35% of the re respondents were able to mark it as true, while 77.55%
of the respondents were able to mark statement four Fødevaren kan spises efter den
angivne dato, såfremt den lugter normalt og ser fin ud as correct.
For the false statements referring to _MHT a total of 68.11% of the respondents were
able to mark statement two Fødevaren skal være anvendt senest denne dato as false,
while 63.47% were able to mark statement three Det betyder det samme som "Sidste
anvendelsesdato" as false and 68.11% were able to mark statement six Uanset om
fødevarens emballage åbnes, kan fødevaren altid spises indtil denne dato as false.
However around a quarter of the respondents thought that statement two and three were
true while roughly 20% thought that statement six was true. Furthermore 47.21% of the
respondents thought in fact that statement five Fødevaren kan efter overskridelse af
datoen udgøre en sundhedsrisiko was true while only 35.76% of the respondents knew
it was false.

Page 46 of 70
Table 15 - _MHT (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
82,35% 12,23% 5,42%
Kvaliteten af fødevaren er i orden til og med denne dato. (1)
26,01% 67,96% 6,04%
Fødevaren skal være anvendt senest denne dato. (2)
26,63% 63,47% 9,91%
Det betyder det samme som "Sidste anvendelsesdato". (3)
Fødevaren kan spises efter den angivne dato, såfremt den
77,55% 13,93% 8,51%
lugter normalt og ser fin ud. (4)
Fødevaren kan efter overskridelse af datoen udgøre en 47,21% 35,76% 17,03%
sundhedsrisiko. (5)
Uanset om fødevarens emballage åbnes, kan fødevaren 19,97% 68,11% 11,92%
altid spises indtil denne dato. (6)

The last item referring to expiration date and which was guessed by the lowest
percentage of respondents were _BF in which only 7.3% were able to guess it perfectly
and respondents guessed on average 3.82 statements correct.
Looking at the individual statements for this item the statement with the highest
percentage was a true statement 79.88% of the respondents were able to identify
statement four Fødevaren kan spises efter den angivne dato, såfremt den lugter normalt
og ser fin ud as true however almost equally as many had marked another true statement
correct. Statement one Kvaliteten af fødevaren er i orden til og med denne dato was
marked correctly by 78.02% of the respondents.
However the last true statement were only guessed correctly by 42.11% of the
respondents while 46.13% thought that statement three Det betyder det samme som
"Mindst holdbar til" was false.
Looking at the false statements for _BF, statement two Fødevaren skal være anvendt
senest denne dato were marked correct as false by 75.08 of the respondents.
45.82% had correctly marked statement five fødevaren kan efter overskridelse af datoen
udgøre en sundhedsrisiko as false however 34.83% thought it was true.
Statement six Uanset om fødevarens emballage åbnes, kan fødevaren altid spises indtil
denne dato were marked correctly as false by 60.99% of the respondents while 20.74%
thought it was true.
Furthermore around 20% of the respondents were in fact in doubt about the correct
answer for the last two statements.

Page 47 of 70
Table 16 - _BF (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
78,02% 13,93% 8,05%
Kvaliteten af fødevaren er i orden til og med denne dato. (1)
17,03% 75,08% 7,89%
Fødevaren skal være anvendt senest denne dato. (2)
42,11% 46,13% 11,76%
Det betyder det samme som "Mindst holdbar til". (3)
Fødevaren kan spises efter den angivne dato, såfremt den lugter
79,88% 11,15% 8,98%
normalt og ser fin ud. (4)
Fødevaren kan efter overskridelse af datoen
34,83% 45,82% 19,35%
udgøre en sundhedsrisiko. (5)
Uanset om fødevarens emballage åbnes, kan fødevaren altid
20,74% 60,99% 18,27%
spises indtil denne dato. (6)

Switching focus from expiration date to the item referring to the ingredient list _ING in
which a total percentage of 36.7% of the respondents were able to guess perfectly, the
second highest percentage score among all 14 items only beaten by _GDA as mentioned
earlier. The average score for _ING was 4.29 correct statements for respondents.
Looking at the individual items, the statement with the highest percentage was
statement one Fødevarens indhold where a total of 85.76%of the respondents were able
to mark it as correct.
The next highest percentage score was statement four xx in which a total of 82.04% of
the respondents had marked correctly as false closely followed by statement five xx in
which 80.19% had correct.
However none of the three other statements scored as high as these three statements,
since only 64.55% of the respondents were able to mark statement six Det er frivilligt
hvilke ingredienser, der angives i ingrediensliste as false while only 58.82% of the
respondents got statement Fødevarens indhold i en rækkefølge efter vægt, hvor første
ingrediens er den, der er mest af correct as true. Only the false statement three
Fødevarens indhold angivet i en tilfældig rækkefølge had a lower correctness with only
57.74% of the respondents able to mark it correctly.
Furthermore the statements referring to the order of the ingredient list statement two and
three had the highest number of wrong markings with 19.50% and 22.91% respectively

Page 48 of 70
Table 17 - _ING (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
85,76% 6,50% 7,74%
Fødevarens indhold. (1)
Fødevarens indhold i en rækkefølge efter vægt, hvor første
58,82% 19,50% 21,67%
ingrediens er den, der er mest af. (2)
22,91% 57,74% 19,35%
Fødevarens indhold angivet i en tilfældig rækkefølge. (3)
Kun fødevarens indhold af 2,32% 82,04% 15,63%
allergene (allergifremkaldende) ingredienser. (4)
5,88% 80,19% 13,93%
Kun fødevarens indhold af tilsætningsstoffer. (5)
Det er frivilligt hvilke ingredienser, der angives i
8,51% 64,55% 26,93%
ingredienslisten. (6)

In addition to the ingredient list and as the last item in the questionnaire _ADDI were
investigated. Overall 7.7% of the respondents were able to give a perfect answer for this
item, while respondents on average had 2.7 statements correct.
Regarding the different statements for _ADDI the statements with the highest
percentages were the two true statements, statement one Tilsætningsstoffer kan være
angivet i ingredienslisten med deres navn f.eks. sorbinsyre which were marked correctly
by 71.05% and statement two Tilsætningsstoffer kan være angivet i ingredienslisten
med deres E-nummer f.eks. E-200 which were marked correctly by 79.72% .

If we however look at the three false statements only 35.45% of the respondents marked
statement three Tilsætningsstoffer er altid angivet med E-nummer i ingredienslisten,
f.eks E-200 as false while 40.71% marked it as true. Another statement in which had
more wrong marking than correct marking were statement five Indeholder fødevaren
tilsætningsstoffer, skal der på bagsiden stå: ”Indeholder tilsætningsstoffer” where
36.22% of the respondents had marked it wrongly as true while 30.34% had marked it
correctly as false.

Looking at the last false statement which had the highest number of percentage correct a
total of 53.25% of the respodents were able to identify statement four Tilsætningsstoffer
fremgår ikke nødvendigvis på fødevaren, da det ikke er et krav at oplyse as false.
Furthermore the three false statements which scored the lowest in percentage of the five
statements had the highest percentage of “don’t know” from 23.84% to 33.90%.

Page 49 of 70
Table 18 - _ADDI (% of respondents who have marked each category, Bold is the correct
answer)
DON'T
TRUE FALSE
KNOW
Tilsætningsstoffer kan være angivet i ingredienslisten med 6,97% 21,98%
71,05%
deres navn f.eks. sorbinsyre. (1)
Tilsætningsstoffer kan være angivet i ingredienslisten med deres 6,35% 13,93%
79,72%
E-nummer f.eks. E-200. (2)
Tilsætningsstoffer er altid angivet med E-nummer i
ingredienslisten,
40,71% 35,45% 23,84%
f.eks E-200. (3)
Tilsætningsstoffer fremgår ikke nødvendigvis på fødevaren, da det
12,85% 53,25% 33,90%
ikke er et krav at oplyse. (4)
Indeholder fødevaren tilsætningsstoffer, skal der på
36,22% 30,34% 33,44%
bagsiden stå: ”Indeholder tilsætningsstoffer”. (5)

5.3.3 Determinants of understanding


To test what factors in combination determines how much consumers understood of the
14 items in the study, it was investigated if the variation in number of correct statements
for each of the 14 items could be explained by other variables in the dataset. Two
groups of variables were used to investigate determinates as mentioned in the literature
review demographic variables have earlier showed to have an effect on use and
understanding of nutrition information and interest in health had also been used in
earlier studies as determinates for nutrition understanding, furthermore as mentioned
other interest variables were added to this study, these were animal welfare, social
responsibility and interest for environment.
For each of the 14 items a regression analysis were performed in two blocks, since
earlier studies by Grunert et al. 2009 showed that demographic variables might be
mediated by interest in Health. Furthermore a dummy variable were created for the
category JOB.
The output for every regression analysis can be found in the appendix F-S. Furthermore
the SPSS outputs can be found on the CD under appendix 6.

Results for the variable _GDA showed the highest R2 14.2% for the complete model. A
number of significant effects occurred. Income had a negative effect on understanding
of the label while education had a positive impact on understanding. Furthermore
working respondents and students had a positive impact on understanding of the _GDA
label.

Page 50 of 70
For the interest variables three of the health interest had a significant effect health
interest; Indulgent had a positive affect for understanding the label while negative
effects were shown with Health interest; Resigned and Health interest; Controlled. This
last variable that had a positive effect was environment in general that had a positive
effect on understanding the label.

Results for the label _MSC showed the highest R2 for the complete model with a score
of 10.5%. The test showed some significant effects as well, for gender male respondents
had a higher understanding than females for _MSC. Furthermore education had a
positive effect as well.
Looking at the interest variables, Health interest; Resigned had a negative effect on
understanding while Health interest; Indulgent had a positive effect on understanding.
And the environment had a positive impact on the understanding of _MSC.

Results for the label _KEY had the highest R2 in the full model with a score of 9.3%.
The test showed some significant effects. Age had a negative effect on understanding, it
was however the only demographic variable that had any influence on the understanding
of _KEY.
Looking at the different interest variables, two of them were significant, Health interest;
Resigned had a negative impact on understanding. Health interest; Indulgent had a
positive impact on understanding.

Results for the label _FAIR showed the highest explaining with the full model with an
R2 of 10.1%. There were some significant effects as well. For the demographic variable
only education level had an effect on understanding the _FAIR label.
While several of the different interest variables had a significant effect, three of the four
health interest variables had a significant effect. Health interest; Indulgent showed a
positive effect on understanding of _FAIR however for Health Interest; Resigned and
Health Interest; Controlled were there a negative effect on understanding the label.
Moreover respondents who showed an interest in the environment in general had
positive effect as well.

Looking at the first of the two organic labels, starting with the Danish _ORG showed
that the full model had the highest R2 with a score of 11.7% for this label some

Page 51 of 70
significant effects were shown. Gender had a significant effect; in which Male
respondents in the survey understand the label better than females in the survey.
Furthermore education showed a positive effect on understanding of the label.
For the interest variables several significant effects were obtained. Health interest;
Indulgent had a positive effect while Health interest; Controlled had a negative effect,
moreover Health interest; Resigned almost had a significant effect on the 0.10 level.
The last variable to have an effect on understanding of the organic label was animal
welfare, which had a positive effect on understanding of _ORG.
Turning the attention to the other organic label the European equivalent to the Danish
one showed the largest R2 when then full model was present with a score of 8.1%. Some
significant effects were shown. Age had a negative effect on understanding.
Furthermore two demographic variables had a significant effect on the 0.10 level.
Number of children in the household below 12, had a negative effect o n understanding
the European organic label. Job situation had a significant effect as well, in which
unemployed showed to have a better understanding of the label compared to working
respondents.
For the different interest variables only one of the variables were significant however it
was on the 0.10 level. In which social responsibility had a positive effect on
understanding the label.

Turning the attention to the item _GTS it showed that the full model had the highest R2
with a score of 4.8%. There were only one significant variable present in the result, even
on the 0.10 level. The only significant variable was Health Interest; Resigned which had
a negative impact on understanding the label.

The next item _BGB were best explained by the full model with a score of R2 on 5.4%.
The only significant demographic variable was household size, which have a negative
impact on the understanding of the label. Looking at the different interest variables two
of the variables were significant however one of them only on the 0.10 level.
The one on the 0.10 level were Health interest; Resigned, which had a negative
influence on understanding the _BGB label. While social responsibility showed a
positive effect on understanding.

Page 52 of 70
For _BOB the full model explained it best with a score of R2 of 5.7%. There were some
significant effects, Gender almost had a significant effect on the 0.05 level in which it
showed that male respondents were better in understanding the label than female
respondents, it was however the only demographic variable there were significant.
For the different interest variables two were significant on the 0.10 level. Health
interest; Indulgent were however significant on the 0.05 level and had a positive impact
on understanding furthermore the variable Health interest; Controlled had a positive
impact as well on understanding _BOB.

Turning the attention to the items that belongs to the backside of the package is showed
some significant effects as well.
Starting with the first of three items referring to the expiration date the item _SA which
was best explained by the full model with an R2 of 5.7% showed some significant
effects, for the demographic variables only one of the variables were significant
however it was on the 0.10 level.
In which gender was significant and it showed that female respondents had a better
understanding of _SA compared to male respondents in the survey.
Regarding the interest variables two variables were significant however only one on the
0.05 level. The two significant variables were Health interest; Resigned and Health
Interest; Controlled which both had a negative impact on understanding the _SA label.

Looking at the second variable referring to expiration date _MHT it showed that the full
model had the highest R2 with a score of 10.6%. Some significant affects were present,
however only one demographic variable were significant. The variable Education had a
positive effect on understanding of _MHT.
For the different interest variables several effects were shown to be significant. On the
0.10 level almost five of the variables were significant. Health interest; Resigned and
Health Interest; Controlled both had a negative influence on the understanding of _MHT
while Health Interest; Indulgent had a positive effect on understanding.
For the two variables that were almost significant on the 0.10 level an interest in the
environment and social responsibility led to better understanding of _MHT.

Concerning the last item referring to expiration date _BF it showed the highest R2 in the
full model with a score of 9.8%. Several variables had a significant effect. Looking at

Page 53 of 70
the demographic variables; only one of the variables was significant on the 0.10 level.
In which education showed to have a positive impact. For the different interest variables
three variables showed a significant effect. Health Interest; Indulgent had a positive
effect while the two variables Health Interest; Resigned and Health Interest; Controlled
had a negative influence on understanding _BF.

Switching focus to the two items related to the ingredient list _ING and _ADDI several
effects were present. For the item _ING the full model had the highest R2 with a score of
12.4%. It showed several significant effects. Foremost age had an negative impact on
understanding of the ingredient list. Furthermore a income had a negative impact on
understanding of _ING.
For the interest variables several significant effects took place, foremost Health Interest;
Resigned had a negative impact on understanding while Health Interest; Indulgent had a
positive effect on understanding the ingredient list. Furthermore an interest in the
environment in general showed to have a positive impact on understanding; however it
was only on the 0.10 level it was significant.

For the last item in the survey _ADDI several significant effects were present. The
biggest R2 were seen in the full model with a score of 11.8%. Only one of the
demographic variables was significant and it showed that income had a negative effect
on understanding of _ADDI.
For the interest variables some significant effects were shown. Health Interest; Resigned
had a negative influence on understanding, while Health Interest; Indulgent had a
positive influence on understanding.
The last significant showed that an interest in the environment had a positive effect on
understanding additives.

5.3.4 Results of Conjoint analysis


For the two conjoint analyses 20 respondents were not included as they had marked the
same value for each profile. The remaining 244 for chocolate and 362 for fish remained
in the sample for conjoint analysis, using PSS 19.
For an interpretation of the utility estimates for each attribute, a higher positive utility
estimate corresponds to a higher preference and a negative utility corresponds to a lower

Page 54 of 70
preference. However if a profile has an overall negative utility it does not mean that it
had an overall negative effect for the respondent and was unacceptable, just that it is an
inferior profile compared to the other profiles in the analysis.

Expiration date labels utility estimate


The estimation in SPSS for the discrete attribute expiration date label was for fish and
chocolate respectively (1).

0.436 !" #$ %!",$# '!(' 0256 !" #$ %!",$# '!('


    0.136 !" #$ %!",_+,- 2 3 #4#   0.093 !" #$ %!",_+,- 2
0.239 !" #$ %!",_01 0.163 !" #$ %!",_01

As expected present of information, _MHT and _SA, had a higher utility than none-
labeled products. However _SA was the preferred information for both surveys
compared to _MHT.

_ING label utility estimate


The estimation in SPSS for the discrete attribute _ING label was for fish and chocolate
respectively (2).
.5237# '!('  $" 2 .2347# '!('  $" 2
   6 3#4#  6
. 5238!('  $" . 2348!('  $"

For both surveys visibility of _ING was preferred especially for the fish profiles.

Origin of product utility estimate


The estimation in SPSS for the discrete attribute origin of product was for fish and
chocolate respectively (3).

.1437# '!('  $" 2 .047# '!('  $" 2


   6 3#4#  6
. 1438!('  $" . 048!('  $"

Visibility of origin of product was preferred for both surveys however it was least
preferred in the chocolate survey.

Page 55 of 70
_ADDI utility estimate
The estimation in SPSS for the discrete attribute _ADDI label was for fish and
chocolate respectively (4).

.1037# '!('  $" 2 .0497# '!('  $" 2


   6 3#4#  6
. 1038!('  $" . 0498!('  $"

And again visibility of the _ADDI was preferred for both surveys.

_KEY utility estimate


The estimation in SPSS for the discrete attribute _KEY was only presented in the fish
survey since it is not used on chocolate products (5).

.1207# '!('  $" 2


   6
. 1208!('  $"

Visibility of _KEY was preferred as well for respondents in the Fish survey. However it
was not present in the chocolate survey, since it is not possible to add this logo to
chocolate products

_GDA utility estimate


The estimation in SPSS for the discrete attribute _GDA was present in both surveys (6).

.3387# '!('  $" 2 .2627# '!('  $" 2


   6 3#4#  6
. 3388!('  $" . 2628!('  $"

_GDA was preferred visible for both surveys and in both surveys scored relatively high.

_BOB utility estimate


The estimation in SPSS for the discrete attribute _BOB was present in both surveys (7).

.1157# '!('  $" 2 .0587# '!('  $" 2


   6 3#4#  6
. 1158!('  $" . 0588!('  $"

Page 56 of 70
_BOB was preferred visible for both surveys and in both surveys scored relatively low.

_ORG utility estimate


The estimation in SPSS for the discrete attribute _ORG was present on both surveys as
well (8).

.1787# '!('  $" 2 .1347# '!('  $" 2


   6 3#4#  6
. 1788!('  $" . 1348!('  $"

The visibility of _ORG was preferred in both surveys however it was less important for
both surveys compared to _GDA

_ MSC utility estimate


The estimation in SPSS for the discrete attribute _MSC was only present on the fish
product since the label concerns fishing (9).

.1127# '!('  $" 2


   6
. 1128!('  $"

The visibility of _MSC was preferred as well however not as important as both _ORG
and _GDA and it were only used in the fish survey since it is only possible to use it on
fish products.

_FAIR utility estimate


The estimation in SPSS for the discrete attribute _FAIR was only presented in the
chocolate survey since it does not exist on fish products (10).

.1177# '!('  $" 2


3#4#  6
. 1178!('  $"

For this attribute visibility was the preferred choice as we have seen with all other
attributes.

Page 57 of 70
The additive model
Assumes that the different utilities can be directly added to make up the overall utility
of any given profile in either of the two conjoint analysis, by combining (1), (2), (3),
(4), (5), (6), (7), (8) and (9) for the fish survey and (1), (2), (3), (4), (6), (7), (8) and (10)
for the chocolate survey or construct profiles not present in the current analysis based on
the individual utilities.
However for the profile mentioned in appendix D the overall utility would be for one of
the chocolate profiles;

1$%  ; <.=>?@.=AB@.CB@<.CBD@.=?=@.C>E@.AB@.F G.HI

Attribute importance
Based on the aggregate range of utilities estimates concerning each attribute the table
below shows the relative attribute importance for the two surveys chocolate and fish.

Table 19 – relative attribute importance


Fish Chocolate

Expiration date _MHT 17,00% 17,94%


_SA
_ING 18,96% 13,51%
Origin of product 8,17% 8,31%
_ADDI 7,76% 8,44%
_KEY 8,22%
_GDA 14,13% 21,28%
_BOB 7,75% 8,43%
_ORG 10,06% 10,71%
_MSC 7,95%
_FAIR 10,59%

There were an overall difference between the two surveys, in the fish survey
respondents were most influenced by _ING in their choice of fish products, which was
the attribute with the highest relative importance, then followed by expiration date and
the least important was _BOB.

Page 58 of 70
In the chocolate survey most respondents were influenced by _GDA in their choice of
chocolate products, which had the highest relative importance and then followed by
expiration date as with the fish survey. While the least favorable attribute for the
chocolate survey was origin of product. Furthermore similar analysis can be performed
on the individual level but it is not covered here.

7. Reflection and Conclusion


This study has showed several interesting facts, both when it came to understanding of
the different labels in the survey but also preferences for the different items and
furthermore several different significant effects were obtained when looking at
determinants of understanding.

Regarding the understanding part of the survey respondents showed some interesting
results. Foremost when respondents were asked about their subject knowledge of the
different items in the study it showed that respondents rated the three items related to
expiration date highest, in fact roughly 70% of the respondents meant that they knew
these three items _SA, _MHT and _BF perfectly.
While respondents showed high uncertainty for the three quality items _GTS, _BOB and
_BGB, in which less than 5% meant they knew perfectly however similar results were
shown in the studies by both London Economics in association with ADAS and
Ecologic (2008) and Fairspeak (2010). Furthermore one of the items referring to organic
products scored high among the respondents as well since roughly 60% meant they
knew all about the _ORG.

However the following part that described their actual knowledge of the different items
in relation to the set of statements in this study showed some interesting discoveries.
The individual statements for the three items referring to the expiration date showed that
around a quarter of the respondents were of the idea that _SA and _MHT have the same
meaning even though it is not true. Furthermore roughly half of the respondents thought
that products with the item _MHT would endure a health risk if eaten after the last date
was exceeded even though it is not the case usually, while around 30% of the
respondents were of the idea that it would not harm to eat products if the day had
expired on products containing _SA however here is the case that products that

Page 59 of 70
exceeded expiration date can pose a risk if consumed. Furthermore around 50% of the
respondents were unaware off that _MHT and _BF have the exact same meaning and
moreover around a third of the respondents were of the impression that products with
_BF would endure a risk if eaten, though not as decisive as number of respondents with
_MHT . Furthermore even though most of the respondents were under the impression
that they knew all about these three items less than a quarter of the respondents were
able to give perfect answer to all statements, in fact when it came to _MHT around
every 10th respondents were only able to come up with a perfect score, these results
looks similar to the divided understanding which was present in the in-depth interviews.

Regarding the two items that related to organic products the Danish item _ORG showed
that respondents had a hard time understanding this label even though roughly 50%
meant they knew all about it. Four out of six questions divided the respondents more or
less equally. And the statement which scored highest among respondents was only
guessed correctly by around 65% of the respondents. Furthermore around 60% of the
respondents were under the impression that organic products are produced without any
pesticides, even though some are allowed, these results are quite different to the results
by Fairspeak (2010), however as mentioned in that study a vague answer was seen as
correct, which not is the case hear.
However not surprisingly was the European equivalent almost unrecognized by the
respondents this was also shown when looking in the individual statements in which
roughly 85% of the respondents had marked the don’t know category for each
statement.

Looking at the items referring to sustainability _FAIR and _MSC they showed some
interesting facts. Foremost around 10% of the respondents expected they knew all about
_MSC however around 17% was able to score perfectly furthermore around half of all
respondents knew that this item referred to sustainable fishing even though it has to be
mentioned that for every statement around half of the respondents were unaware of the
correct answer. For the other label _FAIR in which roughly 40% meant they knew
perfectly were only guessed correctly by less than 10% of the respondents. More than
80% of the respondents knew however that products with this item supported the
farmers and workers, however only around half of the respondents had ticked of the
correct answer for the other statements.

Page 60 of 70
For the items _GDA and _KEY referring to nutrition some interesting facts emerged.
First of all the _GDA label were around 40% of the respondents confident about when
asked about if they know this label however the study showed that in fact roughly 50%
of the respondents were actually able to give perfect scores once it came to it and it was
furthermore the highest scoring item of all 14 items in the study and all statements were
guessed correct by roughly 80% of the respondents. However it has to be mentioned
that if you were able to read you should be able to answer at least three of the six
statements correct, since the answer for three of the statements indirectly were in the
logo and it was as size where it should be possible.
Looking at the other item in this category _KEY in which a quarter of the respondents
meant they knew perfectly however only around 9% were able to come up with the
perfect score. When looked at the different statements for _KEY it became clear that
most respondents were unaware of the correct answer since all statements were
unanswered by roughly half of the respondents and each statement were only guessed
correctly by around a third of the respondents, these results differ from the yearly report
by Ministry for Veterinary and Food Administration on the green keyhole which
showed that around two third of the respondents were able to answer correctly.

Looking at the three items referring to quality _BOB, _BGB and _GTS in which less
than 5% of the respondents meant they knew perfectly. The results of the second part of
understanding showed that only _GTS had a higher score afterwards 6.5% of the
respondents were able to come up with the perfect score while the two other items
scored even lower than initially and the individual score for each statement revealed that
for each statement around 70% of the respondents were unaware of the correct response.
Furthermore it has to be mentioned that for these three items the full name is included in
the logo and not just the three abbreviations, so of you are smart you should be able to
answer some of the statements regarding these three logos by reading the text on the
logo, however the results were fairly consistent with the results from the in-depth
interviews as well as the study by London Economics in association with ADAS and
Ecologic that showed that Danish consumers to almost no knowledge of two of these
labels.

Page 61 of 70
For the last two items in the survey referring to the ingredient list _ADDI and _ING the
study showed that even though around a quarter of the respondents initially meant they
knew all about the ingredient list and every 10th respondents initially meant they knew
all about additives the study showed that around 35% were able to score perfectly with
regards to the ingredient list while only around 7% were able to guess additives
correctly.
Looking at the individual statements for _ING it showed that what respondents were
most unsure of was the order of the ingredient list and if it is voluntary what ingredients
you put on the list which around 60% was able to answer correctly, which was a little
higher than the study performed in Australia by ANZFA, in which 50% were able to
understand the order of the ingredient list.
For the other item _ADDI respondents were unsure if additives always were marked
with an e-number and if the producers has to write on the product that it contains
additives, around 40% had answered each of these two statements true even though it
was false.

Overall the high number of don’t know answers for several of the labels in the study
could be a further indication that it is an area of low involvement since previous
research had acknowledged a “don’t know” answer as an indicator for low involvement
Ray 1973 in Zaichowsky 1985 and according to Silayoi & Speece (2004) the area of
FMCG is usually low involvement.

As expected different effects were showed when different determinants were


investigated for the FOP labels which all are voluntary. Several significant were present
and as seen in the study by Grunert et al. (2009) the demographic variables were partly
mediated by the introduction of the different interest in step two.

Foremost age had a negative effect on three variables _KEY, _EUORG and _ING.
Which showed an increase in age resulted in weaker understanding of these labels and
thereby supported the findings from Cowburn & Stockley (2005), Hawks (2004), FSA
(2004), Grunert et al. (2009).
However findings on income contradicted the findings by Nayga (2000), Dallongeville
et al. (2001) and Nayga et al. (1998), since higher income had a negative effect on
understanding _GDA, _ADDI and _ING.

Page 62 of 70
While prior research on gender, in which Hendri et al. (2007) concluded that woman
had a higher understanding were both supported and refused in this study, since male
had a higher understanding of _MSC and _ORG while female showed a higher
understanding of _SA.

Several studies on education have showed that higher education led to better
understanding Follmer et al. (1991), Dricoutis et al. (2005), Dallongeville et al. (2001)
and Nayga et al. (1998). These results were also supported in this study in which higher
education led to a better understanding of _GDA, _MSC, _FAIR, _ORG and _MHT.

Several other findings were also present in this study number of preteens in the house
had a negative effect on understanding of _EUORG while house size had a negative
effect on understanding of _BGB which supports findings by Drichoutis et al. (2005)
that showed house size had a negative effect on understanding.

The last demographic variable that had an effect were working respondents and students
which had a better understanding of _GDA compared to unemployed respondents.

Furthermore a higher interest in the environment led to a higher understanding of


several labels _GDA, _MSC, _FAIR and _ADDI which especially for the two labels
referring to sustainability makes sense.
Moreover interest in animal welfare had an effect on understanding of _ORG which
again makes perfect sense. Furthermore a higher interest in social responsibility led to
an increase in understanding of _EUORG and _BGB.

The items which showed significant effects on multiple occasions were three items by
Chrysochou (2010) in which he presented different variations of healthy eating.
And the segment Health Interest; Indulgent had a positive effect on all items except
_EUORG, _GTS, _BGB and _SA. This segment describes consumers that are more open
minded to eating and experimenting food so people in this segment would usually be
more interested in food labels.
For the two other segments that was represented with significant effects were Health
interest; Resigned which was significant in all items except the two organic labels while

Page 63 of 70
Health interest; Controlled were present in _GDA, _FAIR, _ORG, _SA, _MHT and _BF.
Both of these segments had a negative effect on understanding. Health interest;
Resigned segment describes consumers that that are not actively seeking health
information while the other segment Health interest; Controlled describes consumers
that are more controlled and have a more positive look on health eating, however this
information on health eating should be of a scientific character (Chrysochou 2010).
So it is obvious way at least the segment Resigned have a negative effect on
understanding however it is more vague to why Controlled have a negative effect on
understanding.

Looking at preferences for the different food labels, it was not unsurprisingly preferred
that labels were visible compared to no label at all. Furthermore it showed that the
ingredient list were more important on the fish product than the box of chocolate, which
you would expect since the fish product was minced fish and it is impossible to see
what it contains moreover a expiration date was important as well, which are not that
surprisingly since we are taking about perishable food.
And the of the two nutrition labels on the fish product _GDA was preferred over _KEY.
While the most important type of information for the box of chocolate was the _GDA
label and expiration date as well.
Furthermore a quality label like _BOB was not important at all for both type of
products, which agrees well with the understanding of _BOB label in the study.
And overall all kind of information were more important for the fish product than the
box of chocolate, which are not that surprisingly since it is two different kind of
products.

Page 64 of 70
8. References
Journals
Acock, A. C. (2005). Working With Missing Values. Journal of Marriage and Family,
Vol 67. pp. 1012 - 1028.
Bernues A, Olaizola A, Corcoran K. (2003). Labelling information demanded by
European consumers and relationships with purchasing motives, quality and safety of
meat. Meat Science 65 pp. 1095–1106.
Barzegari, A., Ebrahimi, M., Azizi, M. & Ranjbar, K. (2011). A Study of Nutrition
Knowledge, Attitudes and Food Habits of College Students. World Applied Sciences
Journal 15 (7): pp.1012-1017
Bezençon, V. & Blili, S. (2010). Ethical products and consumer involvement: what's
new?, European Journal of Marketing, Vol. 44 Iss: 9/10, pp.1305 - 1321
Boyle, K.J., Holmes, T. P., Teisl, M. F. & Roe, B. (2001). A Comparison of Conjoint
Analysis Response Formats. American Journal of Agricultural Economics 83(2). pp.
441-454
Boyce, C & Neale, P. (2006). Conducting in-depth interviews: A guide for Designing
and conducting in-depth interviews for Evaluation input.
Chrysochou et al. (2010). Social discourses of healthy eating: A market segmentation
approach. Appetite, 55(2), pp. 288-297.
Cowburn G, Stockley L (2005). Consumer understanding and use of nutrition labeling:
a systematic review. Public Health Nutrition 8:pp. 21–28
Dallongeville, J. MMarecaux, N. Cottel, D. Bingham, A. & Amouvel, P. (2001).
Association between nutrition knowledge and nutritional intake in middle-aged men
from Northern France. Public Health Nutrition. 2001, Feb 4(1), pp. 27-33.
Directorate General for Health and Consumer Protection. (2005). The European
consumers’ attitudes regarding product labelling-qualitativestudy in 28 European
countries. Optem, Versailles. QL, EU+ Norway and Switzerland.
Drichoutis, A. C., Lazaridis, P., & Nayga, R. M. (2005). Nutrition knowledge and
consumer use of nutritional food labels. European Review of Agricultural Economics,
32, pp. 93–118.
Dunlap, R.E., Van Liere, K.D., Mertig, A.C., Jones. R.E.(2000). Measuring
endorsement of the new ecological paradigm: A revised NEP scale. J. of Soc. Issues
56(3), pp. 425-442.

Page 65 of 70
European Heart Network (2003). A systematic review of the research on consumer
understanding of nutrition labeling. European Heart Network, Brussels

FSA Ireland (2009). A research study into consumers’ attitudes to food labeling.
Food Standards Agency (2004b) Consumer attitudes to food standards wave 4. Food
Standards Authority, London. QT, UK
Food Safety Authority of Ireland . (2003). Consumer attitudes to food safety in Ireland.
Food Safety Authority of Ireland, Dublin.

Fullmer, S., Geiger, C. J., & Parent, C. R. (1991). Consumers’ knowledge,


understanding, and attitudes toward health claims on food labels. Journal of the
American Dietetic Association, 91, pp. 166–171
Gaschler, R. et al (2009). Change detection for new food labels. Food Quality and
Preference 21 (2010) 140-147.

Green, P. E. & Srnivasan, V. (1978). Conjoint analysis in consumer research: issues and
outlook. Journal of Consumer Research, 1978, vol. 5, issue 2, pp. 103-23

Green, P. E & Srinivasan, V. (1990). Conjoint Analysis in Marketing: New


Development With Implications for Research and Practice. Journal of Marketing

Grunert, K. G., & Wills, J.M. (2007). A review of European research on consumer
response to nutrition information on food labels. Journal of Public Health, 15, 385–399.
Grunert, K. G., Wills, J. M. & Fernandez-Celemin, L. (2009). Use and understanding of
nutrition information on food labels in six European countries. J Public Health.
Grunert, K. G, Wills, J. M., Celemin, L. F. (2010). Nutrition knowledge, and use and
understanding of nutrition information on food labels among consumers in the UK.
Appetite 55. pp. 177–189

Guest, G., Brunce, A. & Johnson, L. (2006). How many interviews are enough? Field
Methods Vol. 18 no. 1 pp. 59-82.

Hendri, G. A, Coveney, J. & Cox, D. (2007). Exploring nutrition knowledge and the
demographic variation in knowledge levels in an Australian community sample. Publich
Health Nutrition 11(12), pp. 1365-1371.
Hakeem, R. Asar, F. & Shaikh, A. H. Socioeconomic differences in Housewife’s ability
to take Nutritional Care. Department of Food and Nutrition, RLAK Department of
Food and Nutrition, RLAK Government College of Home Economics and Department
of Botany, Federal Urdu University of Arts, Science and Technology, Karachi.

Page 66 of 70
Hart, M. (2007). Sampling. International Journal of Childbirth Education, Vol 22, No 2
Hawkes, C. (2004). Nutrition labels and health claims: the global regulatory
environment, World Health Organisation.
Krishnan, H.S., & Chakravati, D. (1999). Memory measures for pretesting
advertisements: An integrative framework and a diagnostic template. Journal of
Consumer Psychology, 8, pp. 1-37

Krystallis, A., et al. (2009). Attitudes of European citizens towards pig production
systems, Livest. Sci.
Lavidge R.J., Steiner G.A. (1961). A model for predictive measurements of advertising
effectiveness. Journal of Marketing 25(4). pp. 59–62

Loureiro M.L., Gracia A, Nayga R.M. Jr. (2006). Do consumers value nutritional
labels? Eur. Rev. Agricultural Economics 33. pp. 249–268
Mack, N., Woodsong, C., MacQueen, K., Guest, G., & Namey, E. (2005). Qualitative
Research Methods: A Data Collector’s Field Guide.
McGuire, W. J. (1985). Attitudes and attitude change. In Lindzey, G., & Aronson, E.
(Eds.), The handbook of social psychology. vol. 2. New York: Random House. pp.233–
346

Morgan, D. L. (1996). Focus Groups. Annual Review of Sociology, Vol. 22, pp. 129-
152.
Nayga, R. M. Jr. (1996). Determinants of consumers’ use of nutritional information on
food packages. Journal of Agricultural and Applied Economics, 28, pp. 303–312.
Nayga, R. M. Jr.. (2000). Nutrition Knowledge, Gender, and Food Label Use. Journal
of Consumer affairs
Nayga, R. M. Jr., Lipinski, D. & Savur, N. (1998). Consumers' Use of Nutritional
Labels While Food Shopping and At Home. Journal of consumer affairs, Vol, 32, Issue
1. pp. 106-120
Rasor, RE, & Barr, J. (1998). Survey sampling of community college students: For
better or for worse. American River Coll.
Schaeffer, Nora Cate & Presser, Stanley. (2003). THE SCIENCE OF ASKING
QUESTIONS. In: Annual Review of Sociology, Vol. 29, Iss. 1, pp. 65-88.

Page 67 of 70
Schwarz, N. & Oyserman, D. (2001). Asking questions about behaviour: Cognition,
communication, and questionnaire construction. American Journal of Evaluation, 22,
pp. 127-160.

Silayoi, P. & Speece, M. (2004). Packaging and purchase decisions. British food
Journal, Volume 106, No. 8 pp. 607-628.

Zaichowsky, JL, 1985, ‘Measuring the Involvement Construct’, Journal of Consumer


Research, Vol. 12, No 3, pp. 341-352.

Books
Aczel, A. D. & Sounderpandian, J. (2006). Complete Business Statistics. Sixth Edition.
McGraw-Hill.
Arbnor, I. & Bjerke, B. (1997). Methodology for Creating Business Knowledge. Sage
Publications.
Blumberg, B., Cooper, D. R. & Schindler, P. S. (2005). Business research methods.
McGraw-Hill.
Eagly, A. H. & Chaiken, S. (1993). The psychology of attitudes. Harcourt Brace College
Publishers, Forth Worth TX.
Heldbjerg, G. (2006). Grøftegravning i metodisk perspektiv. Samfundslitteratur
Kvale, S. (1996). Interviews: An introduction to Qualitative Research Interviewing.
Thousand Oaks, CA: Sage.
Nunnally, J. & Bernstein, I. (1994). Psychometric Theory 3rd Edition. New York:
McGraw-Hill.
Peter, J. P., Olson. J. C. & Grunert, K. G. (1999). Consumer behavior and marketing
strategy european edition. McGraw-Hill.
Sundhedsstyrrelsen (2010). Den nationale sundhedsprofil. Sundhedsstyrelsen
Schiffman, L. G., Kanuk, L. L. & Hansen, H. (2008). Consumer behaviour: An
European outlook. Upper Saddle River, NJ: Pearson.
SPSS inc. (2007). SPSS Statistics Base 17.0 User's Guide. Chicago IL.
Tabachnick, B. G. & Fidell, L. S. (2007). Using Multivariate Statistics. Pearson
International.

Publications

Page 68 of 70
C. K. Marie, J. Pernille, Yougov - Public, April 2011, Kendskab til nøglehulsmærket og
nøglehulskampagner præ- og posttest 2011 og udvikling fra 2009-2011, Ministeriet for
Fødevarer, Landbrug og Fiskeri, viewed 7 October 2011,

Evaluation of the CAP policy on protected designations of origin (PDO) and protected
geographical indications (PGI) by y London Economics in association with ADAS and
Ecologic http://ec.europa.eu/agriculture/eval/reports/pdopgi/report_en.pdf
ANZFA (2001).FOOD LABELLING ISSUES – CONSUMER QUALITATIVE
RESEARCH.

Web
Alt om kost. n.d. Viewed on 15th May 2012
http://www.altomkost.dk/Forside.htm
Danmarks statistik. n.d. Viewed on 7 April 2012.
http://www.dst.dk
European Food Information Council. n.d. Viewed on 15th May 2012.
http://www.eufic.org
Europa-Kommissionen. (2011). Smag for Europa - 15/02/2011. Viewd on 15th May
2012
http://ec.europa.eu/news/agriculture/110215_da.htm
Fairtrade. n.d. Viewed on 15th May 2012
http://fairtrade-maerket.dk/
Fairspeak. n.d. Kendskab til mærkningsordninger. Viewed on 15th May 2012.
http://fairspeak.org/Publikationer/resultater/Kendskab_til_maerkningsordninger_FairSp
eak2010.pdf
Forbrug. Mærkningsguide. Viewed on 15th May 2012
http://www.forbrug.dk/Test-og-raad-foer-du-koeber/Baeredygtigt-
forbrug/Maerkningsguide
Forbrugerrådets magasin. n.d. Viewed on 15th May 2012.
http:///www.taenk.dk
Fødevarestyrelsen. n.d. Viewed on 15th May 2012.
http://www.foedevarestyrelsen.dk
Fødevarestyrelsen. (2011). Fødevarer med beskyttede betegnelser. Viewed on 10th
April 2012

Page 69 of 70
http://www.foedevarestyrelsen.dk/Foedevarer/Foedevarekvalitet/Beskyttede%20bet.%2
0og%20traditionelle%20specialiteter/Sider/forside.aspx
GDA - Kend din mad. n.d. Viewed on 15th May 2012.
http://gdainfo.dk/home.aspx
Landbrug og fødevare. n.d. Viewed on 15th May 2012.
http://www.lf.dk
Marine Stewardship Council. n.d. Viewed on 15th May 2012
http://www.msc.org/home-page-da?i18nredirect=true
Nøglehullet. n.d. Viewed on 15th May 2012
http://www.noeglehullet.dk
Samvirke. n.d. Viewed on 15th May 2012.
http://www.samvirke.dk
Økologisk landbrug. n.d. Mærkning. Viewed on 15th May 2012.
http://ec.europa.eu/agriculture/organic/eu-policy/logo_da

Page 70 of 70

You might also like