Table of contents:
Abstract ................................................................................................................................................ 1
Chapter 1. Introduction ........................................................................................................................ 1
1.2. Problem statement ......................................................................................................................... 6
1.3. Methodology ................................................................................................................................. 7
1.4: Delimitation .................................................................................................................................. 8
Chapter 2. Definitions .......................................................................................................................... 9
2.1. Model Description....................................................................................................................... 13
2.2. Model Description, Apple ........................................................................................................... 14
Model description, electric car market:....................................................................................... 16
2.3. Green advertising: ....................................................................................................................... 19
2.4. The Ansoff Matrix and the target group in the electric car market............................................. 29
2.5. Product life cycle (PLC) and MEC-theory for the electric car market ....................................... 30
Chapter 3. The Experiment, Theory................................................................................................... 32
3.1.1. The Experiment, Theory; Needs as a prerequisite for motivation ........................................... 33
3.1.2. The Experiment, Theory; Elaboration Likelihood Model (ELM): .......................................... 34
3.1.3. The Experiment, Theory; Attitudes: ........................................................................................ 37
3.1.4. Underlying theory for variable 1, Q2 and Q3 in the questionnaire: ........................................ 39
Chapter 3.2. The Experiment, Method ............................................................................................... 41
3.2.1. Designing methods for collecting the data ............................................................................... 42
3.2.2. Collecting the data: .................................................................................................................. 46
3.2.3. Coding and arranging the data: ................................................................................................ 46
Chapter 3.3. The Experiment, Results ............................................................................................... 47
3.3.1. Reliability and validity: ............................................................................................................ 47
3.3.2. Assumptions for the profile analysis:....................................................................................... 49
3.3.3; Results of the experiment: ....................................................................................................... 52
3.3.4: Assumptions for the independent-samples t-test, Tesla Roadster & Apple ad: ....................... 56
3.3.5: Hypotheses and independent-samples t-test results for the Tesla Roadster & Apple ads ....... 57
3.3.6. Independent-samples t-test for variable 1, group: ‘females’ vs. ‘males’ and
‘Scandinavians’ vs. ‘Other nationality/foreigners’ .................................................................. 60
3.3.7. Independent-samples t-test for variable SUMQ2, Q2 items, groups: ‘females’ vs. ‘males’
and ‘Scandinavians’ versus ‘Other nationality’ ....................................................................... 61
3.3.8. Independent-samples t-test for SUMQ3, Q3 items, groups: ‘females’ vs. ‘males’ and
‘Other nationality’ vs. ‘Scandinavians’: ................................................................................... 64
Chapter 4.1; The experiment, summary of results: ............................................................................ 65
Chapter 4.2; Conclusion:.................................................................................................................... 68
The thesis will primarily focus on sustainable communication in respect to the Apple Company and
the electric car market. Sustainable communication also includes green advertising. The goal of the
thesis will therefore be to analyze how green technological innovations like the new, sustainable
family of laptops from Apple and electric cars can be promoted in the most efficient way.
This report will mainly describe green marketing for the electric car market, as this has turned out to
be more problematic than promoting products for Apple.
An experiment is made in the end of the thesis. Results indicate that the attitudes of people who are
highly involved in environmental matters are more favourable when argument strength is enhanced.
This result is confirming the Elaboration Likelihood Model (ELM).
Respondents are also compared by sex and nationality in respect to items representing
environmental responsibility and pro-environmental behaviour. Females show overall greater
responsibility for the environment, and they also engage in more eco-friendly behaviour than males.
Comparing respondents grouped by nationality show less overall discrimination on the items.
The introduction of this paper will start explaining why sustainability is important to take into
consideration. Also a short overview of human numbers will be given to show the negative impact
humans have on the planet.
Chapter 1. Introduction
Organizations can no longer be treated primarily as rational, neutral, technological systems of
production, as is done in traditional organization studies. They must also be seen as systems of
destruction. They systematically destroy environmental value. This destruction cannot be dismissed
away as an ‘externality’ of production. Organizations must become accountable for these
externalities, which are a central and systematic feature of organized economic activity.
(McDonagh, Pierre, 1998, p. 594).
The paragraph sheds light on how the organization can be seen as a natural part of our environment.
This can also be seen as interactions between economic and ecological systems (see appendix 1,
figure 1). Here the environment is the whole natural environment (planet earth). The economy is
located within the environment, and exchanges energy and matter with it. Humans extract various
useful things from the environment, for example oil, iron ore, timber etc. Humans put back waste
which necessarily arise when making their living, e.g. sulphur dioxide and carbon dioxide from
burning oil for example. The economy and the environment can be seen as interdependent. Over the
last three decades, the interactions between these two entities have increased rapidly, and the levels
of extraction from and the insertions into the environment do affect the way the environment works
(Common, Michael & Stagl, Sigrid, 2005).
Organizations must take an active step, when we talk about environmental destruction. The
companies have to establish the legitimacy of link between the environment and any products or
services they promote (McDonagh, Pierre, 1998). Organisations have to produce products that are
able to minimize the harmful impact they have on the environment, not only when the products are
used by the customers, but also in the production process. Furthermore, companies have to come up
with solutions of how to control the impact the products have at the end of the products’ life cycles
– for example by recycling material which is harmful for the environment.
The debate on climate change and the release of greenhouse gases is huge. Over the last 50 years,
human activities have released sufficient quantities of carbon dioxide and other greenhouse gases.
These gases, which are released particularly through the burning of fossil fuels, trap additional heat
in the lower atmosphere and affect the global climate. Over the last 100 years, the global
temperature has increased by 0,75oC and when we look over the last 25 years, the rate of global
warming has accelerated, at over 0,18 oC per decade. There are many consequences and they are
severe for our planet, for examples are sea levels rising, glaciers are melting, habitats around the
world are changing, and extreme weather events are becoming more intense and frequent. Global
warming may also bring some benefits, for example fewer winter deaths in temperate climates and
increased food production in some areas, but the overall health effects of the changing climate are
likely to be overwhelmingly negative. The fundamental requirements for health are affected, e.g.:
clean air, safe drinking water and sufficient food (http://www.who.int).
In order to give an overview of the way humans affect the environment, the following paragraph
will describe the history of the population of humans and also the amount of energy which was used.
The Human Energy Equivalent (HEE) is a unit of measurement which is the amount of somatic
energy required by a human individual. Somatic energy refers to the energy an animal acquires in
its food and which is converted into work, growth and heat. When the human animal learned how to
control fire in the hunter-gatherer phase, it began the exploitation of extrasomatic energy – in other
words it began to exert more power than what was available from its own muscles (Common,
Michael & Stagl, Sigrid, 2005). The overview of human numbers and energy use will be carried out
by looking at different phases ranging from the transition to agriculture to the industrial revolution.
In appendix 1, figure 2 human history is divided into three broad phases. The longest phase is the
hunter-gatherer phase, which lasted from the beginning of the existence of humans until around
12.000 BP (before the present, where the present is taken to be 2000 AD). It has been estimated that
there at this point where four million people living in the world. Human numbers have been
growing very slow during this phase.
Agriculture started around 12.000 BP, and human numbers began to accelerate. By 2.500 BP, that
is BC 500, the human population has been roughly estimated to be around a hundred million. The
growth rate increased in this period. With agriculture the food production increased; in the hunter-
gatherer system food production could support 0,2 – 2 humans per square kilometre, whereas the
agricultural system could support 25 to 1000 humans per square kilometre (Common, Michael &
Stagl, Sigrid, 2005). In the agricultural phase, which lasted approximately 12.000 years, the
technology of energy evolved, and the human animal was deploying some 3-4 HEE, so that in
addition to his muscle power he was using extrasomatic energy at the rate of 2-3 HEE. The per
capita use of energy in the agricultural phase was doubled compared to the hunter-gatherer phase,
and population size increased with a factor of 200. Urbanisation was now possible and the first
towns appeared around 9000 BP.
By the time of the industrial revolution 1800 AD, that is 200 BP, the global population reached 0.9
billion. A characteristic of the industrial revolution is the use of coal to drive the machines in the
factories and to transport raw materials to manufacturers from the factories. The straight line
between BC 500 and 200 BP (see appendix 1, figure 2) does not reflect a smooth increase, as there
where periods where human numbers declined because of catastrophic events. Today the human
population has been estimated to be six billion people, and the growth from one to six billion people
during the last 200 years corresponds to a growth rate of 0.9 per cent per year. From 12.000 BP to
200 BP the total annual growth rate was 0.04 per cent, so human numbers have clearly increased
more rapidly after entering the industrial revolution phase.
By the year 1900, humans used 14 extrasomatic HEE – this is the same as saying that the average
human had command of 14 human slaves. By the end of the 19th century, the average human used
about 19 extrasomatic HEE, which corresponds to 19 human slaves. In 1997 extrasomatic energy
use in USA was 93 HEE, while in Bangladesh it was 4 (Common, Michael & Stagl, Sigrid).
Clearly the above overview shows us that humans have grown rapidly, because the living
conditions were enhanced. The increased welfare comes with a negative impact on our environment,
because of the extraction of non-renewable resources. Non renewable resources are resources which
are not replaced, or which are replaced only very slowly by natural processes. Primary of non-
renewable energy resources are the fossil fuels – oil, natural, gas, and coal. Fossil fuels are
continually produced by the decay of plant and animal matter. The rate of production is extremely
slow – very much slower than the rate at which we use them. Non-renewable energy resources that
we use and which not are replaced in a reasonable amount of time (our lifetime, our children’s
lifetime) are considered ‘used up’ and are not available to us again (www.cpast.org).
Governments around the world are trying to implement regulations which firms have to follow, to
force them to act more environmental friendly. Companies who strive to be sustainable in the
production process have not only the possibility to enhance the corporate image of the firm, but also
to get first-mover advantages. They can get these advantages by developing green technological
innovations, which shape the future market place. This will be described more extensively later in
One thing companies can do to minimize their negative impact on the environment is to develop
innovations that can replace the non-renewable resources, for example by using solar energy. They
should also be able to substitute products which are harmful for the environment with products
which are less harmful. One challenge in this respect is to change people’s ways of living - that is
changing their habits. To be successful in accomplishing this task, they have to communicate their
green technological innovations, and this is exactly what this thesis will be about.
Underneath a figure will show the reader, how the thesis is organized.
Figure 1, Structure:
1.2 Chapter 2 Chapter 3
Problem Definitions The experiment, Theory Chapter 4.1
1.3 2.2 Chapter 3.2
Methodology Model description The experiment, Method
1.4 2.3 Chapter 3.3
Delimitation Green advertising The experiment, Results
After the introduction, section 1.2 will be the problem statement. This will give the reader an insight
of what will be analyzed in the thesis, and why it is considered important. Section 1.3 describes the
method used to answer the problem statement, and section 1.4 what is not the focus of the thesis.
Chapter 2 defines the concept of sustainability and describes an example of a sustainable
technological innovation. It is the communication of such innovations which is analyzed in the
A model has been found, which describes companies’ relationship with other stakeholders than
customers (for example investors). A description of this model is given, but it is not the main part of
the thesis, because it is the objective to analyse the communication/marketing efforts primarily
made by firms from the automobile industry (green advertising). Still it is important, in the first
place, to explain why firms have to establish good relationships with the public as a whole. To get a
deeper insight into the model, examples are discussed for the Apple Company as well as the electric
car market. Apple has also been chosen because of considerations concerning the green ads from
The communication options available for firms are described in section 2.3. This section is broad;
as it describes techniques used to promote the electric cars. Only the electric car market is described
here, because promotion of this has turned out to be difficult. This is not the case for Apple. Apple
is a company which is environmental aware, but this is very essentially related to the production
process. People do not buy products from Apple, because they are persuaded by a green message,
but more because of the design and the quality of the products. The thesis does not analyse
marketing in general; instead the focus is on green marketing. Making an analysis of the electric car
market will in this respect give opportunities to make recommendations of what future electric car
producers should focus on in the marketing process.
In section 2.3 there will be a description of the electric car market according to the product life
cycle (PLC). Furthermore, there will be a description of the target group of electric cars.
An analysis is made of how potential buyers of green products are influenced by the quality of the
arguments. In the experiment chapters (starting from chapter 3.1) the theory will serve as the basis
of such an experiment. The method of the experiment will be described; how the questionnaire is
developed, what the hypotheses are etc. The final part of the chapter contains the results, and these
are compared to what could be expected from the theory. The results are based on the ads of two
green products: an ad about sustainable laptops by Apple and the Tesla Roadster electric car ad.
An analysis is also made to assess the opinions of the respondents about environmental matters. It is
examined whether there are any differences between females versus males and Scandinavians
versus people who have another nationality in their environmental behaviour, and in the
responsibility they have for the environment.
The thesis is wrapped up by chapter 4, where the results are summarized. A conclusion follows this
chapter and suggests what firms have to be aware of in their efforts to communicate their green
1.2. Problem statement
One thing is how to make a green technological innovation, but the marketing efforts used to
promote the innovations is especially vital, because it can be difficult to communicate to a green
market, which is not yet fully developed. Further it is a challenge to persuade consumers to buy a
product like an electric car, because many car buyers do not see the advantages and are not really
willing to engage in eco-friendly behaviour, because this eco-behaviour goes against their former
way of living – i.e. their habits. Therefore, it will not only be the politicians responsibility to give
green incentives to the companies, but also the firm’s responsibilities as they decide what is
available to the consumers on the market. The companies have also the opportunity to change the
habits of the consumers, if they accomplish the task of making efficient green products, which at the
same time are communicated well.
In the marketing process, the producers have to keep in mind that the target groups can react quite
differently to an ad. This can be explained by looking for differences between high and low
involvement groups. Many studies have shown that strong arguments in advertisements do change
the attitude of some consumers, i.e. consumers who are highly involved in the product.
Technological green innovation is an area which today is heavily debated as the following phrase
from a recent interview by Interbrand’s global practice leader Tom Zara implies: ‘corporations are
now looked to act in ways that reduce the sins of the past. They are expected and held accountable
to innovate to make the Earth a better place’ (cited in www.interbrand.com).
The overall research question and important sub-questions for the thesis are shown here:
Research question: How do companies like Apple and the electric car companies communicate
their green technological innovations to the public, and how are groups (high/low involvement)
reacting to the content of the communication/ads?
- Why is green technology important, and why is it necessary to change the habits of
- How do companies communicate their green technology?
- Which is the target group for electric cars and how does it adapt to green technological
- How do companies in the electric car market communicate their green technology? (the
- Do high and low involvement groups react differently to communication when weak and
strong arguments are involved?
- Are there any differences in people’s pro-environmental behaviour when people are
grouped by sex or nationality?
- Do people feel equally responsible for the environment, when they are compared by sex
To answer the problem statement, theories from the field of marketing communication and green
communication will be used. Various examples will be given in respect to the car manufacturers
and the Apple Company.
It may seem surprising that Apple has been included in an analysis about the communications of
green innovations. But this company is interesting as it recently launched the greenest family of
notebooks ever made (www.apple.com). Apple has been criticized by some environmental
organizations for not being a leader in removing toxic chemicals from its new products and not for
recycling their products in an efficient way. But investigating Apple’s current practices further
shows that Apple now is ahead of most of its competitors in these areas. The fact that environmental
organizations have criticised Apple before might be a signal, that Apple have not communicated its
environmental concerns clearly enough to the public. An analysis will be made of what steps Apple
takes to minimize its ecological footprint, and what advantages can be obtained by pursuing a green
way of doing business, e.g. why it is important to have a good relationship with investors, suppliers
etc. The model is described later in the thesis.
Gasoline cars has one of the largest impacts on the environment as they contribute to global
warming, acid rain, resource depletion, and congestion (Joireman, Jeffrey A., 2004). Electric cars
are interesting in respect to this because, if they in the future are introduced successfully, they can
serve as substitutes for traditional, gasoline cars and thereby minimize the exhaustion of green
house gases significantly. The communication efforts of the car manufacturers will be analysed by
using the communication mix, also called the 4 p’s (Product, Price, Place, Promotion).
The thesis describes the arguments and the design of the electric car ads, which are made to
persuade potential buyers. With the design is meant an analysis of the pictorial contents of the ads,
as these elements can influence the customers attitudes besides the attributes given by the products.
An experiment will be made in order to analyse how consumers perceive the products after viewing
ad’s with both strong and weak arguments. What is expected in this regard can be read in the theory
and method section of the experiment (3.1 and 3.2). The questionnaire also gives us the opportunity
to look for differences in the respondent’s eco behaviour and responsibility as mentioned in the
There are three methodological approaches in business research/consulting/investigation today;
these are (Arbnor, Ingeman, 1997): the analytical approach, the systems approach and the actors
approach. The analytical approach which is the one used in this paper has its origin in classical
analytical philosophy. Existing theory is available in advance which make the verification or
falsification of stated hypotheses possible. Verified or falsified hypotheses clarify objective facts
that constitute subsets of objective reality.
The analytical approach gives also the opportunity to reproduce causal relations. This means that it
is possible to explaining some effect by finding the prior or current cause. This is done when
explaining the causal relationship which is hypothesized to exist between high and low involvement
groups and the quality/strength of arguments in the advertisements.
The results in this approach are pure cause-effect relations, logical models, and representative cases.
The end-result of the thesis will therefore have a generalizable character in order to be a prerequisite
of continuing research.
The overall goal of the approach/method is to
To determine a problem
To forecast/to guide
The problem has already been determined. The problem can not be described until it has been
determined and it cannot be explained until it has been described (Arbnor, Ingeman, 1997). A good
forecast is based on an explanation (and consequently a description as well), and a guide is useless
if it does not refer to the future – i.e. if it is not based on a forecast. The goal of the thesis is
especially also to give a guide about what firms in the car-industry have to take into consideration
in the future marketing process.
The research is limited to cases from Apple and the electric car market. Because this paper is
analysing the communication options for green firms, it has been necessary to include something
more than ‘just’ the automobile industry for green cars to give a broader insight into how ‘green
communication’ works. Apple will be described to show how a firm in the market of electronics can
achieve benefits by communicating their green initiatives to stakeholders. The electric car market is
included as well, because electric car producers have invented interesting technological innovations
by the introduction hybrid and electric cars. Hybrid cars, which can be seen as both a gasoline and
an electric car is not ignored, as it serves as consumers’ transition from the gasoline car to the
We will look at the electric car market in the whole world. In this way it will be apparent if some
countries can learn from others. Also cooperation exists across countries, and this can not be
In respect to the experiment, described later in the thesis, the main focus will be on classifying
respondents into high and low involvement groups – this is done by grouping them in relation to
how involved they are according to environmental concerns. Attitudes for the products will be
measured, but the essential focus will be on how argument strength affects the respondents through
communication. For this reason it has been chosen not only to show an electric car ad, but also an
Apple ad to fully hedge against the possibility that not all of the respondents have a desire to buy a
car. Most of the respondents have a computer and for many it will be easier to evaluate their liking
and attitudes of ads with subtexts that have more relevance to them. Only two cases have been
chosen to include in the experiment; Tesla Roadster and Apple. In this way the questionnaire was
not exhausted in considerations of the time each respondent had to use.
A qualitative interview has been made to get a deeper insight into the electric car market. The
interviewee is Jens Christian Lodberg Høj, who is the project manager at Insero E-Mobility – a firm
which responsibility is to bring actors together to develop solutions in the electric car market. The
interview can be seen as a supplement to the analysis, which has been necessary to make in order to
give solutions in respect to the problem statement. A detailed description of the interviewee and the
interview itself can be found in the transcript after the appendices.
Chapter 2. Definitions
As mentioned in the structural part before, it will be necessary to define what sustainability is and
what is meant with a green technological innovation.
Sustainability is a very broad term. We have to get an understanding of why sustainability is
important, not only for the environment but also because it will have a clear impact on how
companies produce and survive in the future market place. Further, it will be defined what is meant
with sustainable communication, which is the overall theme of the thesis.
Sustainability can be defined as ‘a development that meets the needs of the present without
compromising the ability of future generations to meet their own needs’ (Jaeger, William K., 2005).
Sustainable communication is important because there is a need to clearly communicate the
ecological issues of sustainable development to wider audience than is presently achieved in society
thus raising the importance of the communicative act (McDonagh, Pierre, 1998). The
communication of sustainable innovation is also important because the society is moving away
from a situation of hyper consumption 1 to a preferred situation of sustainable consumption.
Companies in the communications industry have to establish the legitimacy of link between the
environment and any products or services they promote (McDonagh, Pierre, 1998).
The idea of eco-innovation is fairly recent. The concept of eco-innovation can be defined as new
products and processes which provide customer and business value but significantly decrease
environmental impacts (James, Peter, 1997). Many companies believe that the more environment-
friendly they become, the more the effort will erode their competiveness (Nidumolu, Ram et al.,
2009). Companies believe that it will add costs to get sustainable and that it will not deliver
immediate financial benefits. But this point of view is false. Ram Nidumolu and M.R. Rangaswami
studied 30 large corporations for some time and their results suggested that green technological
innovations yield both bottom-line and top-line results. Costs related to sustainability, which many
believe would rise, were actually lowered because the companies end up reducing the inputs they
use in producing the products.
means that there is no connection between the thing consumed and the consumption act itself i.e. it is consumption for
its own sake
Innovations can be classified into (Tidd, Joe & Bessant, John, 2009):
Product innovation – changes in the product/services than the organization offers.
Process innovation – changes in the ways in which the products are created and delivered – e.g. a
change in manufacturing methods and equipment used to produce a particular product.
Position innovation – changes in the contexts in which the product services are introduced – e.g.
re- position the product in the minds of the consumer.
Paradigm innovation – changes in the underlying mental models which frame what the
organisation does, e.g. change an industry like Itunes changed the music industry (customers not
buying CD’s anymore), other examples can be: the shift to low-cost airlines, provision of online
insurance and other financial services and repositioning of drinks to designer products – requires
also process innovation and product innovation.
The firms which will be described here can gain a lot by creating certain associations in the minds
of the customers, which means that ‘position innovation’ important in respect to the thesis.
Paradigm innovation should not be ignored either. If firms are successful in communicating green
technological innovations, they have the chance to affect the habits of their customers and thereby
change the whole market place – changed to be more ecological responsible and energy efficient.
Joe Tidd and John Bessant are describing the innovation concept even further by identifying the 4
zones of innovation:
Zone 1: incremental innovation: improvement of the products or processes and the use of
knowledge, which the company got from core competences.
Zone 2: modular/application innovation: a change in one element but the architecture remains the
same. Here new knowledge is used within an established framework – for example the use of new
materials to airframe components.
Zone 3: Discontinues/radical innovation: neither the success or the ways in which to achieve the
success/end state are known – the rules of the game/the way to undertake the processes are known –
something completely new is developed, and new entrants will be attracted.
Zone 4: Technical/architectural innovation: we can here use existing knowledge and recombine it
or combine new and old knowledge – e.g.: we can have knowledge about making a product and as
soon we obtain/get insight into a new technology, this old and new knowledge can be combined to
create an improved product.
The central zones for the green innovations in this thesis are all four zones mentioned above.
Apple’s innovations are more incremental than they are radical. Zone 2 is relevant as well, because
Apple has learned to make laptops with toxic-free material (www.apple.com). The automobile
industry falls into zone 3 and 4. Many of the car manufacturers may have existing competences
which can help them to produce completely new electric cars. Some of the components like for
example the battery of the car can be seen as a radical product innovation – later incremental
innovations are used to improve the battery, which makes the car able to drive longer distances on
each charge. The end state i.e. whether the products will be a success or a failure in the future is not
known for certain. The technical/architectural innovation (zone 4) is also relevant when looking at
the car industry; car manufacturers have an existing knowledge about manufacturing cars, but new
knowledge has to be combined with old knowledge in order to make the electric car.
Before an analysis is made of green marketing and what advantages the innovations can give
society, it is necessary to describe why it is important for firms to use green communication - not
only to persuade customers to buy their green products through advertising, but also why it is
important in relation to other stakeholders than consumers. Both Apple and the automobile industry
will be used as an example in connection with this.
2.1. Model Description
In an article by Pierre McDonagh the author asks whether businesses in a Western industrial society
can be transformed into ecologically sustainable organizations (www.ingentaconnect.com). He
further describes how these organizations communicate with the world in which they operate. He
does not only look at the relationship between the company and the customer, but instead at all the
people involved as the organisations interact and communicate with. In appendix 2, figure 1 it is
shown what other stakeholders than customers the organization may have. In the box to the left in
the figure, we see the environmental marketing communications. Cause-related marketing, green
advertising and being ecological responsible can enhance the corporate image of a firm. Corporate
environmental reporting is also important, and here eco-disclosure and eco-access are important
terms. By ecological disclosure organisations are being assessed by publics, because of the
information that is available about the company’s production processes etc. The firms are judged by
what they are freely prepared to reveal about their activities. Companies who have disclosed the
truth about what they do are more likely to be trusted than those organizations where the truth has
been found out (McDonagh, Pierre, 1998). The public may feel that there is much more to be
hidden by the firm, when the truth has been found out.
In the figure ecological dialogue is also an important principle of the Sustainable Communication
process. Organisations have to engage in an ongoing dialogue with their publics to help them
understand the issues, and become ecological meaningful if the organisations are to be trusted. The
value of firms can be increased if publics are convinced of the ecological worth of the firms as
publics can air their views directly to members of the organisation and see how the organisation
reacts. In this situation the companies will draw their audiences to the decision making process.
Dialogue and trust is important for firms and only when these criteria are fulfilled, the company can
promote itself using ecological issues. Pierre McDonagh stresses further that promotion of e.g. the
company’s products or services can facilitate their own and the customer’s behaviour. If a green
communication gap is present, the organization can manage it through the process of Sustainable
Communication. Sustainable Communication should not be misconstrued as a quick fix, but should
be more radical, so all the members of the organization are aware of a shared, green vision. If the
company cannot develop a clear vision for its members, the promotion of the activities of the firm
may be blurry in the eye of important actors like the government, investors, legislators and
consumers. If the company can not live up to what it promises its stakeholders, the company can not
be trusted which is crucial for the corporate image of the firm and thereby also the success of the
2.2. Model Description, Apple
Now that the importance of Sustainable Communication has been explained, it is time to see how
Apple communicates and what they communicate in respect to environmental aspects.
Most of Apples green communication can be found on their webpage. Apple seems to be a very
environmental aware organization. They stress that it is important to look at the impact their
products have on the planet, and therefore they make comprehensive measures of the company’s
ecological footprint. For the past three years, Apple has used a comprehensive life cycle analysis to
determine where their green gas emissions come from. They add up the emissions generated from
transportation, manufacturing, use, and recycling of their products as well as the emissions
generated by their facilities. About 98 percent of Apple’s carbon footprint (see appendix 3, figure 1)
is directly related to their products, the last 2 percent is related to their facilities
Apple is also minimizing the impact of the growth (see appendix 3, figure 2). One way to reduce the
impact on the environment is to improve the environmental performance of the products. Apple
uses less material, shipping with smaller packaging, free of toxic substances used by others and is
being as energy efficient and recyclable as possible. Apple’s growth outpaces the rest of the
industry and therefore it is important that they minimize the impact they have on the environment.
In 2008 their revenue grew 57 percent and they are the only company in their industry that is able to
claim that every product not only meets but actually exceeds the strict energy guidelines of the
ENERGY STAR specification. Here it becomes apparent that Apple is concerned with the
guidelines the government and legislators set up to be followed by firms to reduce their ecological
Apple is also concerned about the manufacturing processes. The company has a clear, shared vision
and guidelines for its designers and engineers in the development of smaller, thinner, and lighter
products (http://www.apple.com/environment/). When the products become more powerful, they
are using less material to produce and generate fewer carbon emissions. A product like today’s
21.5-inch iMac is more powerful than the first-generation 15-inch iMac. It is designed with 50
percent less material and generates up to 50 fewer emissions. In only one generation the iPad
became 33 percent thinner and up to 15 percent lighter, producing 5 percent fewer carbon emissions.
In appendix 3, figure 3 it is possible to see the reduction of carbon emissions for some of Apples
In appendix 2, figure 1 there is a link between the firm and its suppliers. In order for firms to be
ecological friendly in the whole supply chain, all members have to engage in participative
ecological sustainability. Apple is committed to ensuring that working conditions in their supply
chain are environmentally responsible. They have sited their ‘Supplier Code of Conduct’ as well as
their supplier audit reports on their supplier responsibility website. In the Supplier Code of
Conduct’, Apple is demanding from their suppliers that they disclose information about their
business activities, structure, financial situation, and performance in accordance with the applicable
laws and regulations and prevailing industry practices (apple.com/supplierresponsibility/). Further,
Apple mentions its environmental considerations as an integral part of the business practices.
Suppliers shall commit to reducing the environmental impact of their designs, manufacturing
processes, and waste emissions. Apple has set clear specifications which its suppliers must follow
and also rules and regulations prohibiting or restricting the use or handling of specific substances.
Suppliers are forced to identify and manage substances that pose a hazard if released to the
environment. Suppliers also have to ensure safe handling, movement, storage, recycling, reuse, and
In the model from appendix 2, figure 1, ecologically responsible packaging is mentioned to promote
the corporate image of the firm. Apple explains on their webpage that they are employing teams of
design and engineering experts who develop product packaging that is slim and light but still
protective for the product. The efficient packaging design not only reduces the material and waste,
it also helps reduce the emissions produced during transportation. An example is the packaging for
the new iPhone 4, which now is 42 percent smaller than for the original iPhone 3 shipped in 2007.
This results in 80 percent more iPhone 4 boxes to fit on each shipping pallet where more pallets fit
on each boat and plane, and fewer boats and planes are used which results in fewer carbon
Another important point in the environmental marketing communication box is ‘socially
responsible consumption’. This is a point Apple is concerned with as a significant portion of
greenhouse gas emissions are produced when the customer plug in the products and start using them.
Therefore is Apple designing both the hardware and the operating system, which makes sure that
they work together to conserve power (http://www.apple.com/environment/). An example is the
Mac mini which uses even less power than a single 13-watt CFL light bulb, making it the most
energy-efficient desktop computer in the world. In appendix 4, figure 1 the figure shows how
energy efficient some of Apples products are compared to 60-watt light bulb2.
Apple is also using corporate environmental reporting. For example they have ‘Product
environmental reports’. As already explained, Apple is concerned with the way their products are
manufactured, used, and recycled because the products represents the largest percentage of Apple’s
total greenhouse emissions. In the environmental reports by Apple, publics have the opportunity to
find detail information about categories like (http://www.apple.com/environment/reports/):
climate change: how much does each phase of the products life cycle contribute to total
greenhouse gas emissions
Restricted Substances: What toxic substances that are not in the product
Energy efficiency: How efficient the product is while on, off and in sleep mode
Material efficiency: What exactly is in the product
Recycling: What will happen at the end of the product’s life that contribute to its
Model description, electric car market:
Although the chapter primarily has been concerned with Apple, it is necessary to conclude the
chapter with a short analysis of the electric car market as well. From line 130, p.7 in the transcript,
Jens Christian Lodberg Høj (JC) acknowledges that there exists some kind of triangulation in the
electric car market. Stakeholders like the customers will have as much as possible for the lowest
price, the companies want to earn as much as possible and the government has a long-term
economic and environmental friendly perspective. Many countries rely on the Kyoto agreement,
which is important to consider, if the goals of minimization of greenhouse gases are to be realised.
The government has to promote environmental friendly technology, and the companies have to
consider if they want to pursuit this technology in order to develop Corporate Social Responsibility.
JC says that incentives which are not directly related to the costs of an electric car have been
implemented (incentives which are related to the price of electric cars will be dealt with in the next
section). People can park for free, if they own an electric car (see transcript, line 30, p. 3). This is
The carbon emissions are calculated while the system is idle and has completed loading its operating system. For
products with displays, the display is set to maximum brightness and it is assumed that the emissions are generated from
a mix of power grids in the U.S.
possible in for example Copenhagen and Odense in Denmark, but it is actually still prohibited. So
in this area there is something which does not fit between local plans and the law set up by the
government, he explains. In Oslo, Norway electric car owners are allowed to drive in the bus lane
and it is also possible to cross bridges for free with an electric vehicle.
The companies in the market can also influence the decisions by the government. This is done by
lobbyism. In the transcript from line 332, p. 12 JC explains that an alliance has been made called
‘Dansk El-bil Alliance’ (in English ‘Danish Electric-car Alliance’). This alliance is made to force
the government to make certain regulations and incentives which enhances the firms’ possibilities
to make attractive offerings for potential buyers. If the government takes an active step, advantages
will be gained for the whole Danish economy, because stakeholders, companies and the government
together can make the firsts steps to be leaders in the area of e-mobility. A report made by Deloitte
and the Danish Electric-car Alliance shows that 20.000 unoccupied jobs could be available by the
year 2030. This can be compared to the Danish wind-mill industry today.
From line 279, p. 11 in the transcript, JC says that he does not think that the automobile industry is
deviating from other industries, when talking about the communication between the parties in the
model. Stakeholders include infrastructure firms (e.g. Better Place or Choose EV) that are
establishing charger-stands for charging the batteries in the cars in different areas across Denmark.
Big players in the energy-sector, e.g. ‘Dong Energy’ should not be forgotten. He further stresses,
that the model is very broad. It should be seen consisting of free layers, which are overlapping.
Firms like Microsoft, Apple, Sisco, ITC and automobile firms like General Motors, Bosch and VW-
group are all companies coming from different sectors, and they influence each other, because of
the dialogue which exists between them and their stakeholders. In this way agreements are made,
which the organisations have to stick to and from which they finally can get advantages.
There is a strong interest in hybrid and electric vehicles among politicians and the public because
with respect to this there is danger of confusion and too-high expectations. Accurate information
and international collaborations are more important than ever. An agency has been named the
‘International Energy Agency’ (IEA). It has made an agreement called ‘Implementing Agreement
for co-operation on Hybrid and Electric Vehicle Technologies and Programmes’ (IA-HEV) with
the goal to provide information to all parties involved. This shows that the communication process
is very important before success can be obtained in the market for electric and hybrid cars (IEA
hybrid and electric vehicles annual report, March 2010). Further, the membership and collaboration
in the Implementing Agreement provides access to a valuable forum for all countries interested in
electric and hybrid vehicles. An IA-HEV membership is also showing a sign about how serious a
country is about putting together a well-informed strategy to tackle the field of hybrid and electric
IEA’s programme drives the communication process between all parties in each member country,
by what is called ‘Annexes’ in the IEA language (IEA hybrid and electric vehicles annual report,
March, 2010). The annexes are analogous to working groups, where at least two or more countries
have to find an agreement about a work plan. The work can be shared by tasks (task-shared) or by
costs (cost-shared). The task-shared approach is the most common. Examples of annexes/tasks that
have been undertaken by member-countries are:
Information exchange: here the working group establishes a regular information exchange
on hybrid, electric and fuel cell developments and promotion measures in the IA-HEV
Electrochemical systems: the focus is on technical details concerning experiences with
specific test protocols or abuse testing. Themes for discussion include aging tests for high-
energy batteries and the world supply of lithium. Further topics are batteries under extreme
conditions, testing PHEV (plug-in hybrids and electric vehicles) batteries, first responder
issue, recycling and reclamation of battery materials, government incentives for production
of batteries and improved collaboration among battery researchers.
Fuel cells for vehicles: Market deployment for electric vehicles, lessons learned: an
analysis is made each year, which analyses the reasons for success and failures at
introducing electric can hybrid vehicles on the market. Understanding what worked and
what did not work is important as car manufacturers and users have several options when
choosing a clean propulsion technology (e.g.: alternative fuels like natural gas, fuel cell
In the last 20 years electric vehicles and EV manufacturers come and go. The individual
stories show identifiable patterns which can serve to educate manufacturers, policy makers
and others to avoid repetition of mistakes. Some of the results include: 90 % of EV
charging takes place at home, introducing fast charging points in Tokyo resulted in more
extensive use of EVs, although the fast charging points were not used much, the lack of a
standard defined interface between vehicle and grid has been a problem for both utilities
and EV manufactures, past efforts underestimated the importance of establishing a network
of sales points with trained staff that also showed competence in maintenance and service
Working groups in planning: here information can be found for the working group’s future
tasks, these include: life-cycle analysis and e-waste, battery electric vehicle system
integration, electrification and road transport, lightweight design, test procedures, e-motors,
controllers and chargers.
The IEA is engaging in dissemination activities, were the main target public for the work of the IA-
HEV are governments and public institutions of the member countries. Confidential reports and the
results of the working groups are the most important things in the Implementing Agreement. A
wide range of information is available even for non-member countries like annual reports and IA-
HEV clean vehicle awards.
In the model we see that feedback from the parties is important to strengthen the communication
process. In the IEA annual report, it is written that IEA forecasts have been criticized by experts
from member countries of the IEA as too conservative. Therefore it is important to have a tighter
feedback process between experts in the member countries and at the IEA headquarters.
In the same report, it is also mentioned that it is crucial that multiple stakeholders work together to
meet all the challenges posed by the threat of climate change in an interconnected world. This
includes restructuring the automotive industry and restraining the use of fossil fuels and other
resources (IEA hybrid and electric vehicles annual report, March 2010).
The description of how the IEA works shows that a large international network has been established.
The tasks of the working group involve a lot more than what has been described here. It has been
the intention to show how much can be accomplished if countries, who really are willing to move
forward in the area, can gain in knowledge and expertise from each other in respect to sustainable
In the next chapter, the electric car market will be analysed in connection with green advertising.
Green advertising is important, because the ultimate goal is of course to persuade consumers to buy
the products. The goal is to make consumers engage in more sustainable behaviour, that is to
change their habits – and this is not an easy task.
2.3. Green advertising:
This chapter is primarily analysing the marketing opportunities and practices for firms in the
electric car market. Apple will not be drawn into the discussion, because their marketing activities
are not concerned with green advertising as such. Apple is focusing more on the functional benefits
of their products and the image customers can get by choosing their products. It seems that Apple
wants to be seen as eco-friendly, but this is more in the production process, which already has been
described in the previous chapter.
The chapter will not only describe the available marketing tools which the automobile firms have at
their disposal, but also how the advertisements are designed, and what arguments are used.
The electric car market will be analysed with ‘The Ansoff Matrix’, the product life cycle (PLC) and
with the Means-end-chain theory (MEC) as well. Also the target group of electric cars will be
Green advertising is far from being straightforward. Green advertising can be considered as any ad
that meets one or more of the following criteria (Hartmann, Patrick; ‘Green advertising revisited’,
1) Explicitly or implicitly addresses the relationship between a product/service and the biophysical
environment. 2) Promotes a green lifestyle with or without highlighting a product/service. 3)
Presents a corporate image of environmental responsibility.
Green advertising is here understood as the promotion of a product or brand that has environmental
claims. It is advertising which addresses the relationship between the product and the biophysical
The above-mentioned marketing tools are more precisely known as instruments of the ‘marketing
mix’. Marketing means the process of planning and executing the conception, pricing, promotion
and distribution of ideas, good and services to create and exchange value, and satisfy individual and
organizational objectives (Pelsmacker, Patrick De, 2007). When firms have defined the goals and
objectives and which target group and market position that has to be defended, they have to decide
on which tools to use of the marketing plan. The marketing mix is a set of instruments that are
divided into four categories, called the 4 P’s of the marketing mix (Product, Price, Place,
Table 1: Instruments of the marketing mix
Product Price Place Promotion
Benefits List price Channels Advertising
Features Discounts Logistics Public relations
Options Credit terms Inventory Sponsorship
Quality Payment periods Transport Sales promotions
Design Incentives Assortments Direct marketing
Branding Locations Point-of-purchase
Services Personal selling
Source: Pelsmacker, Patrick De, 2007
In the interview with Jens Christian Lodberg Høj (JC), I ask him about which arguments are used in
the ads of electric cars (see transcript line 343, p.13). What can be promoted is attributes which are
similar to what we see when we are exposed to ads for gasoline cars. This means consumers can be
told about the benefits, the features, quality, design, the brand and services connected to the car.
One example of this is what can be seen in the ads of the Tesla Roadster later in the thesis, but we
can also use the Nissan Leaf as an example. The Nissan Leaf is UK’s first mass-produced car
(http://www.guardian.co.uk). It is claimed that the car is 100 % electric, it has a high response
80kW AC synchronous electric motor, the range is 100 miles before the battery has to be recharged,
it can drive up to 90 mph and there is room for 5 passengers (http://www.nissanusa.com). All these
factors are the features of the car. Furthermore, information can be found on the website about the
interior of the car; it has front and rear heated seats, a heated 3-spoke steering wheel, front centre
armrest with storage place etc.
Comfort and convenience is also an important dimension, as it is written that the car has an
intelligent key and push button start, Bluetooth hands-free phone system, Automatic Temperature
Control (ATC), remote keyless entry, cruise control with steering wheel-mounted controls, rear
window defroster and a lot more.
Details about the audio-system in the car and safety can also be found on Nissans website.
Clearly information about benefits, features, options, quality and design of the products is provided.
This is the case when looking across all electric vehicles. The marketing of the cars are mainly
concerned about how long distances the cars can drive, and that no CO2 emissions are released.
The packaging of the product is not relevant here. This is more related to Apple’s products.
Numerous research papers have dealt with how important packaging is, because it is the package
which has to catches the eye of the consumer in a shopping situation. The package of the product
can also serve as a convenience factor for transportation as explained in the previous chapter.
The car producers can gain advantages by having a strong brand (like Nissan or Renault for
example). However, it does not seem that it is a necessity, but it can be a huge advantage. In the
interview on page 8, line 178, JC compares the car market with the market for smart-phones. The
essence of what he says is that smaller firms, which have not gained competences by producing
gasoline cars, are able to make some very good electric cars as well. Nokia was the biggest mobile-
telephone producer some years ago, but now Apple and HTC are having the hugest share of the
market. Small electric car producers are now having the chance to build up a strong brand by
introducing electric cars in a good quality. If they are lucky, they can gain first-mover advantages,
because the electric car market does not seem to be fully developed at the moment. Quite
surprisingly, Mercedes and BMW has not yet made any electric cars which are worth talking about,
except the Smart electrics by Mercedes and the Mini-e’s by BMW (see from line 168 on page 8 in
Taking the product tool from the marketing mix into consideration, it is important to stress, that the
product instrument consists of three layers. The core product is the unique benefit that is being
marketed – it corresponds to the unique place in the mind of the consumer, which will be focused
upon (Pelsmacker, Patrick De, 2007). The brand is a summary or a visualisation of the core benefits
and all the associations it leads to. The core product itself (the electric car) has to be translated into
a tangible product. The descriptions above about the features, quality, benefits etc. are all
contributing in making the product tangible. The product has also to be augmented, and this can be
defined as the service layer on top of the tangible product. The service layer can be elements such
as prompt delivery, installation service, after-sales service and management of complaints. The
service layer for the electric car market is interesting, because it gives the certain infrastructure
firms the opportunity to make package-solutions (see from line 386 on page 16 in the transcript).
Two of the leading electric car operators (or firms responsible for handling the infrastructure for
electric cars) in Denmark are called ‘Better Place’ and ‘ChooseEV’. They are working with
establishing battery stations across the country. The first charging spots are already established in
Copenhagen, and more will follow as a result of the agreements between Better Place and Danish
municipalities and companies (IEA hybrid and electric vehicles annual report, March 2010). They
are also able to provide consumers with a charger at their residents (http://www.vindenelbil.dk/). JC
explains that if customers want to buy an electric car from Peugeot, they do not visit a Peugeot
dealer, but they have to contact e.g. ChooseEV. In a similar way a Smart electric can be bought
from the company ‘Clean-Charge’. In this way package solutions can be made, exactly how we see
it in the cell-phone market. The consumer is in this way forced to pay an additional fee for the
additional services, which are provided by the electric car operators. JC further stresses that
traditional car dealers have to be aware of this, because the more popular e-mobility gets, the less
service is provided compared to a gasoline car as electric cars only contain 20 % moveable parts
compared to a normal car.
The price is the only marketing instrument that does not cost anything, but it provides the resources
to spend on production and marketing activities (Pelsmacker, Patrick De, 2007). The list price is the
official price of a product. Discounts and incentives can be used to make the product more attractive.
Also systems of down payments and payment periods, combined with attractive interest rates can be
used to make the offering more attractive for the consumer. This can also ensure that the immediate
budget constraint is less a problem for the consumer.
Electric cars are at the moment quite expensive (see from line 195, p. 9). The most expensive part
of an electric car is the battery, which accounts for approximately 60 per cent of the total cost. JC
explains that the price will not drop before the firms are getting advantages connected to economies
of scale. This means that the batteries have to be produced on large-scale which make the price drop.
The problem is that the producers are not making a large-scale production of electric cars as long as
they are unsure if there is a good enough market for them.
Another factor which has to be considered is competition. Car manufacturers will adjust the price, if
other producers are threatening their market share. JC believes that the introduction of cars like the
two-person electric car introduced by a small company located on Bornholm (a small Island in
Denmark) called ‘My Castle’ which costs DKR 150.000 (≈ €19.841) could offset some things. Also
the introduction of the Nissan Leaf can cause a price reduction in general, because it can perform as
good as or even better than other cars. JC further explains that the more similar the electric car gets
to a gasoline car, the more relevant they will be as an alternative for consumers (see from line 64, p.
4 in the transcript).
Furthermore, incentives are used in the electric car market. A consumer grant will reduce the retail
price of electric and hybrid vehicles by 25 per cent in the UK, capped at £ 5000. This came into
effect on 1st January 2011 in the UK (http://www.hybridandelectriccars.co.uk/incentives). There are
good chances that potential customers will be tempted to buy electric cars because of these
incentives. To take the Nissan Leaf as an example again, which came on the market in the UK in
February 2011, it means that the retail price has dropped from £ 28.990 to £ 23.990, which is a
much more attractive price for the buyers. The Leaf will because of this be able to compete with a
hybrid car like the Toyota Prius. The Leaf has some advantages compared to the Prius, because it
offers zero emissions and much lower running costs.
JC also provides information about incentives which are used: from line 25, page 3 in the transcript
he explains that there is no tax on electric cars until the year 2015. In Norway there is no value-
added tax and in France the Government gives buyers a payment of € 5000 for buying an electric
In Norway, were most progress for electric carts can be seen, regulations for the tax of electric cars
for businesses has been made, so you can get DKR. 3,67 (≈50 cents) per kilometre of the
governments official transportation reimbursement, but in Norway they get DKR. 4,10 (≈55 cents)
for driving an electric car (see transcript line 43, p. 4). The conclusion of this is that there is a higher
transportation reimbursement for an electric car than for a gasoline car.
In line 345, p. 14 in the interview, I ask JC if it is difficult to tempt consumers, when the prices of
electric cars are so high. He says that it depends on how it is done. Instead of making a sales
argument on the official price of the car, the ads could instead highlight how expensive it is to drive
an electric car compared to a normal car in the long-term. The electric car will be less expensive
when looking at the costs which are associated with it in the long-term.
By means of place or distribution the company manages the process of bringing the product from
the production site to the customer (Pelsmacker, Patrick De, 2007). Transportation of the product,
keeping an inventory, selecting wholesalers and retailers, deciding on which types of outlet the
product will be distributed in, and the assortment of products in the outlets are essential parts of the
Toyota can be used as an example here. At a car exhibition called Biler for alle (In English: Cars
for everyone) which I participated in to gain knowledge about the electric car market, Toyota gave
me a magazine called ‘Toyota and the environment’. Toyota is producing electric and hybrid cars
(www.toyota.com) and are very environmentally aware when considering production, logistics etc.
They have made a model to describe this concern called ‘Toyota’s 360 degree approach’. In this
approach product design and development, the production process, logistics, sales, marketing and
the market and recycling of vehicles that are at the end of the life-cycle are the main components of
the model. This shows that Toyota is a company which is concerned with many aspects when
thinking of eco-awareness. They feel a responsibility in the whole chain before the product reaches
the customer. They are developing innovative eco-technologies, e.g. hybrid and hydrogen motors.
Also, they produce a huge assortment of vehicles which in the most effective way are utilizing the
fuel in gasoline and eco-friendly diesel motors. Toyota is also striving to minimize the production
facilities impact on the environment, minimize the consumption of resources and maximize
recycling of recyclable parts.
Since 2001, Toyota’s production sites in Europe have been successful in minimizing the average
water consumption with 42 per cent and the average energy consumption with 44 per cent for each
car produced. This makes them leaders in the area compared to other big producers in Europe.
Furthermore, zero garbage has been dumped on the waste site (Toyota and the environment, 2011).
To maintain an environmental friendly image it is important for companies like Toyota to be
sustainable across all actors in the distribution channel. Toyota has more than 3.000 Toyota- and
Lexus dealers. They state that they are demanding that all their actors should live up to the
environmental standards proposed by the ISO 14001 agreement by the year 2015. This will
guarantee that the cars are handled with the environment in mind. Products which have been seen as
damaging for the environment have been disposed, and waste from production has been sorted so
some parts can be recycled.
In 2008 Toyota launched a program called ‘Sustainable Dealers’ which is made to secure the
sustainable operations in the future at the dealers’ outlets. The outlets are now built with grass on
the rooftops and solar panels. New sustainable outlets are built in Malmö, Sweden and in La
Rochelle, France. These outlets are undergoing what Toyota calls the Kaizen-energy program
(Kaizen is a Japanese phenomenon which means contentious improvement and which is the core of
the way Toyota is making business).
The last instrument in the marketing mix is promotion. Promotion, also called marketing
communications, is the most visible instrument of the marketing mix. They involve all instruments
by means of which the company communicates with its target groups and stakeholders to promote
its products or the company as a whole (Pelsmacker, Patrick De, 2007).
JC gives information about the promotion mix in the transcript from line 354, p. 15. The first point
in the promotion tool is advertising. Advertising is considered a synonym of marketing
communications. However, a large variety of communication instruments exists, each with its own
typical characteristics, strengths and weaknesses (Pelsmacker, Patrick De, 2007). Advertising can
be seen as a non-personal mass communication using mass media such as TV, radio, newspapers,
magazines, billboards etc. In Danish TV a lot of car commercials are broadcasted, but not for
electric cars. In other countries however, electric cars commercials are aired. Again the Nissan Leaf
can serve as an example. In a TV commercial found on Youtube, it is shown how easy it is to
charge the battery from home. Also they say that the car has got the title ‘car of the year 2011’ and
that it is 100 per cent electric. No arguments are used about the environment and the commercial is
rather short – it lasts for 30 seconds (http://www.youtube.com/watch?v=e8yhPdK4Eh4). Nissan
could have omitted arguments about the environment in this case, because they want to capture
market shares from the regular car market. The core message in the ad is that it is easy to recharge
and that it is the car of they year in 2011.
Nissan also aired another one called ‘Gas powered everything’
( http://www.youtube.com/watch?v=pC_IjqYYf4o). The commercial shows how our lives would be
affected if hairdryers, computers etc. ran on gas instead of electricity. In the end a Nissan Leaf is
shown and the spokesman says: ‘What if everything ran on gas? And again; what if everything
didn’t’? The car is shown while it is being recharged at the road, surrounded by trees, which of
course are symbolizing the environment. In another Nissan Leaf TV ad called ‘Polar Bear’
(http://www.youtube.com/watch?v=BNeEVkhTutY) we follow a polar bear from an ice-landscape
where everything is melting. In this commercial the environment is in focus, and the polar bear
walks through a huge city. In the end of the commercial a guy goes from his home to his car where
he meets the polar bear. The man looks quite frightened, but the polar bear embraces him (because
he is driving a Nissan Leaf). A voice over end the commercial with the words: ‘100 % electric leaf
– innovation for the planet, innovation for all’. The TV ads are mainly from the UK and the US.
Magazines are also a good communication tool. JC explains in line 359 on page 14, that no electric
cars really are advertised in magazines, because the target group is not reading car magazines. He
advises me to look at a report by Etrans, which explains the target group in more detail. However
he says later (see line 375, p. 14) that one of the biggest car magazines in Denmark called
Bilmagasinet (in English: the Car Magazine) has given interested people the opportunity to rent an
electric car and test it. In this way articles can be written about the electric cars. Information about
electric car magazines has been found on the internet. It is possible to subscribe for a magazine
called ‘Green car journal’, which only is containing information about the latest news related to
electric vehicles (http://www.greencar.com).
Electric car ads often show an electric car embedded in a beautiful nature setting, and this is not a
coincidence. In appendix 4 four ads are shown. Clearly the environment is a central part of the
pictures. In the advertising of electric and hybrid cars, we further distinguish between substantive
and associative claims (Hartmann, Patrick, 2009). Substantive claims relate to the concrete, tangible
benefits and reflect intent to maintain or enhance consumers’ perception of an organisation as
environmentally responsible or of a product as environmentally sound. These claims are important
because they provide information that enables and facilitates individual consumption decisions that
are advantageous to the physical environment. The problem is that many consumers often do not
understand environmental claims correctly. A green ad is more effective if it is supplemented with
associative claims. Associative claims are claims that are image related or present environmental
facts. These claims are more intangible than substantive claims. The use of nature imagery in green
advertising falls into this category. Shifting from an urbanised to a more natural setting has been
considered an intrinsic value for human beings, and numerous research findings in health, medicine
and psychology appear to be supportive of the proposition that nature has some inherently positive
effects on physical and psychological well-being for humans. Elements of nature in the ads also
showed to influence the attitudes of consumers, were informational claims (product features) and
affective claims (nature imagery) leads to brand associations and virtual nature experiences, which
together influence brand attitude (see appendix 5, figure 1).
Considering the promotion tools even further, JC explains that he knows that billboards are used,
for example in Paris, were a huge billboard for Citroen C-Zero (see line 370, p. 14) can be seen by
thousands of people. Green energy is used to highlight the billboard, which makes it consistent with
the message they want to send. He further explains that he believes that the commercials are locally
advertised in the future. The commercials will primarily be shown where there is a market, which
will be in the inner cities.
He continues explaining that the companies want to use as much free marketing as possible. The
companies want all Public Relations (PR) they can get and communicate to customers via
newspapers, who can rent an electric car, which the readers can test-drive. An example of this is the
OPEL Ampera, which potential buyers could test-drive. PR consists of all the communications a
company instigates its audiences or stakeholders. Stakeholders are groups of individuals or
organisations with whom the company wants to create goodwill (stakeholders were dealt with in the
previous chapter). Publicity can be created by press releases and references. Publicity is however
impersonal mass communication in mass media, but it is not paid for by a company and the content
is written by journalists, which means that negative publicity is also possible (Pelsmacker, Patrick
Another communication tool is sponsorship. Sponsorship implies that the sponsor provides funds,
goods, services and/or know-how. The sponsee will help the sponsor with communication
objectives like building brand awareness or reinforcing the brand or the corporate image. Sports,
media, arts, education, science and social projects and institutions, and television programmes can
be sponsored. Events are often linked to sponsorship (Pelsmacker, Patrick De, 2007). In the 2010
Winter Olympics in Vancouver, Canada, the Chevy Volt, which is a hybrid car, was sponsored, and
people were able to test-drive it (http://news.discovery.com). Consistency was obtained between the
car and the event as the Olympics were dubbed ‘green Olympics’ because of the unusual warm
weather that year. Also the Nissan Leaf, which was not on the market at that time, was a sponsor at
the winter Olympics (http://www.insideline.com).
JC talks about sponsorship from line 405, p. 15 in the transcript. ChooseEV is sponsoring a project
called ‘Test an electric car’ where 2500 families are testing an electric vehicle.
Direct marketing is another important tool in the electric car market. Direct marketing
communications are a personal and direct way to communicate with customers and potential clients
and prospects (Pelsmacker, Patrick De, 2007). As mentioned previously on I visited a trade
fare/exhibition in Herning; Cars For Everyone. Several thousand cars were exhibited and there was
also a green-hall where electric and hybrid cars were shown. Each company had a stand where
interested visitors could have a talk with the employees of the automobile companies. The
companies could in this way establish a relationship with potential customers and show them the
advantages of their products. Companies that were present were Better Place, ChooseEV, Toyota,
Citroen, Renault and others. In regard to this, personal selling was also possible, because of the
close relationship the companies were able to establish with the customers.
Last but not least, the importance of viral marketing should not be underestimated. Viral marketing
is also called ‘word-of-mouth’ which is a technique that is used to spur brand users and lovers, game
participants or advocate consumers among the target group to promote their favourite brand to
friends and relatives (Pelsmacker, Patrick De, 2007). He explains that viral marketing dominated
the communication area five years ago (see line 359, p. 14 in the transcript) and that it is still very
important. A company’s image can of course also be damaged, if consumers who have had a
negative experience with a company are telling it to others.
One of the fastest growing promotion tools used in the electric car market is e-commerce. E-
communications offer new ways to communicate interactively with various stakeholders. The
internet can together with e-commerce combine the communication with mobile phones. Mobile
marketing uses the possibilities of text, video and sound transfer to mobile phones. Also interactive
digital television has the potential to transform traditional advertising into interactive
communication on television (Pelsmacker, Patrick De, 2007). The introduction of the smart phones
has made it possible for people to get applications made by car manufacturers, which makes people
able to be updated with the latest news. Also social media like Facebook gives car producers the
opportunity to reach a very broad target group. The car manufactures can show pictures and list
features and specifications on their websites which is connected to the social media. It is also
possible to compare prices and products across producers, which makes it far easier for a customer
to make the right choice.
Clearly there are a lot of possibilities for car producers to communicate their products to customers.
The tools can be used more extensively in the electric car market, when the market is more mature.
2.4. The Ansoff Matrix and the target group in the electric car market
The market for electric cars can be analysed via ‘The Ansoff Matrix’ (Czinkota, Michael R., 1997).
The matrix is shown in appendix 5, figure 2. There are four basic product strategies. These four
alternatives are simply the logical combination of the two available positioning variables of
products and markets. The risk increases the further the strategy moves away from known quantities
(the existing product and the existing market). The electric car market is currently in the phase of
product development, because new cars are produced to an existing market.
The electric car will in the first instance be an urban phenomenon. ‘Etrans’ has divided car drivers
into seven consumer types. First-mover customers will be enthusiastic technician, the guard of the
environment, the big-city bohemian and the design-lover (http://www.dongenergy.dk). ‘Etrans’
further explains that those who most probably will buy an electric car the next time will be persons
in the age of 35-65 years old who live in big cities, who are deep-pocketed and have a necessity to
drive less than 80 km per day. Furthermore, people from the target group is primarily weighting the
design so it matches the style of living and they consume on average 54 km per day – this
requirement is fully accomplished by the electric car.
From line 103, p. 6 in the transcript JC explains that there will be no movement in the market if
everyone is saying ‘that they are ready, when the market is ready’. Renault-Nissan, General Motors,
Opel, Tesla and Toyota are firms which are doing a lot in the product development stage. The
consumers are ready for the electric cars, but it is still unlikely that they can compete with gasoline
cars. Further, he mentions a paradox (see from line 121, p. 6 in the transcript). On one side the
customers are saying that they are willing to buy the electric car as soon as it is able to perform as
good as the traditional, gasoline car. The car manufacturers are on the other hand saying that the
cars already have been developed, but that they are not willing to mass-produce them as long as
there is no market for it. The government has to intervene and make regulations and incentives that
make it attractive for customers to buy the car. The conclusion of this is that the firms have to make
further product developments for their new products, which they can sell to an existing market of
buyers, who obviously are tied to their habits. With the help from government initiatives, the
electric car market will have the opportunity to be successful in the future.
2.5. Product life cycle (PLC) and MEC-theory for the electric car market
The electric cars can be seen as being in the introduction phase of the product life cycle (see
appendix 5, figure 3), but product development combined with initiatives from the government can
make the cars go into the growth phase. If the marketing mix is working effectively, the customers
will be aware of the product and its benefits (Czinkota, Michael R., 1997). In the introduction phase,
companies are having several strategies available to make the product a success. For example the
producers can invest in promotion or set a low price to gain the maximum share of a new market.
Skimming is another strategy, where the maximum short-term profit is derived, typically by a high
price. In the electric car market it seems that money is invested into promotion is combined with a
skimming strategy because of the high costs of the batteries.
Again it should be stressed that first-mover advantages are important in the introduction phase; the
pioneer retains the highest market share even after competition really enters. It has been shown that
pioneers usually hold double the share of later entrants, even over the longer term (cited in:
Robinson, W.T.; ‘Sources of Market Pioneer Advantage in Consumer Goods Industries’). Electric
car producers should stress the advantages which can be obtained by buying an electric car, with the
product life cycle in mind. When looking at the whole lifecycle, an electric car will cost
approximately cost 1/3 of the costs related to a gasoline car. In respect to the costs associated with
driving only, an electric car like the ‘MyCar’ costs DKR. 12 (≈ €1,60) to drive 100 km. while an old
SAAB drives 12 km./litre (see transcript, line 88, p. 5). An affective handling of the marketing mix
and government incentives has the possibility to influence and change the habits of consumers to be
One last thing, which has to be dealt with in this chapter, is the theory of ‘Means-end-chain’ (MEC).
The MEC figure is available in appendix 5, figure 4. The structure suggests a ladder of meanings
that a brand can deliver as a ‘means-end-chain’, with six levels/associations (Pelsmacker, Patrick
De, 2007). The theory is important because it can be used to enhance the promotion/advertising of
electric cars. An electric car brand suggests like all other brands both concrete and abstract
attributes. Concrete attributes can be the car itself – i.e. the body of the car or more precisely the
design. Abstract attributes can be the level of technology in the car – for example a good battery,
which makes customers able to drive long distances on a single charge. These attributes can lead to
a functional benefit: that you can drive without releasing greenhouse gases. The attributes also lead
to psycho-social benefits; to have a car that both delivers an attractive design and which also
performs well when thinking of the technology in the car, can lead to customers having a feeling of
joy when they drive it – both because of e.g. the reduced noise which is experienced when the car is
running. At the same time they will have a feeling of satisfaction if the car matches their style of
living. These benefits will lastly lead to instrumental values and end-values. Instrumental values
relates to a customers necessity to impress others. For some it will for example be important to
show others that they care for the environment, and that they are economically wealthy. The end-
values which will result from this will give buyers of electric cars a feeling of self-esteem, which
will make them able to see themselves as belonging to a group; a group of environmental aware
individuals which have money enough to afford an electric car, and who, at the same time, are
aware of their lifestyle.
The marketing mix has to be integrated, if the car producers’ goal is to awake these feelings in the
target group. Integrated marketing means that there should be consistency and synergy, when the
marketing mix is designed and implemented (Pelsmacker, Patrick De, 2007). The instruments have
to be combined in such a way that the company’s offerings are consistently marketed: all marketing
instruments have to work in the same direction, and not conflict with each other. For example it will
make no sense to sell a high-quality brand like the Nissan Leaf or a Tesla Roadster which lies in the
high premium segment, at a car-outlet which is known for selling cars in a bad quality. The cars
should instead be sold at outlets which focus on standards like the ISO 14001 agreement, or at the
infrastructure firms like Better Place or ChooseEV. In this way, consumers will know that they can
rely on the provider and at the same time get a feeling of superior customer support, while the firms
themselves have potential to earn a lot of money through subscription of different services, e.g.
through services like: loan possibilities for batteries, access to all the chargers located in the country,
installation-services etc. (source: flyer by Better Place – product specification). Customers who buy
an electric car through Better Place pay for a subscription based on how much they use the car e.g.
there is a special price if the consumer drives 10.000 km. per year). The more they drive, the better
the offering will be.
In the next chapter green advertising will be analyzed further by an experiment. Both an ad from
Apple and an ad for the electric car Tesla Roadster will serve as the basis for the experiment.
Chapter 3. The Experiment, Theory
Before the hypotheses and the results of the experiment are described, this chapter will explain the
underlying theory of the experiment.
The theory about attitudes and persuasion began as the central focus of social psychology (Petty,
Richard E. & Cacioppo, John T., 1986). Through the 1960’s researchers began to get interest into
the field as they wondered if attitudes were capable in predicting behaviour. Researchers where
unsure about the results they got in respect to attitude change and the connection to consumer
behaviour because of conflicting research and theory. But by the late 1970’s progress had been
made in addressing important methodological and theoretical issues regarding the first substantive
problem plaguing the field; the consistency between attitudes and behaviour. Conditions under
which attitudes would and would not predict behaviour were specified, and the number of
researchers interested in this field began to rise. However there was still disagreement of what was
responsible for attitude change, and if there was a link to consumer behaviour.
There is a lot of research about how low and high involvement groups’ attitudes are affected by
argument strength in advertisements. The most relevant theory for this experiment is described by
Richard E. Petty and John T Cacioppo’s Elaboration Likelihood Model (ELM). In the very last
section of this chapter, section 3.1.4, the underlying theory for variable 1, Q2 and Q3 from the
questionnaire will be explained.
Before, an explanation of the ELM is made, it will be necessary to explain what is meant by needs,
goals and motivation as they are prerequisites for motivation. In the end of the chapter a description
will be made of what it all leads to; affecting the attitudes of consumers.
Needs are the basis for motivation. If a consumer is motivated to buy a given product, his or her
attitude can be affected. The attitude can influence the consumer’s behaviour by a manifestation of
accepting or rejecting a product.
3.1.1. The Experiment, Theory; Needs as a prerequisite for motivation
Before motivation, which here refers to how highly involved an individual is in respect to ecology,
is explained, it is necessary to know that certain needs are prerequisites to motivation. Motivation is
influenced by consumer needs and goals (Pelsmacker, Patrick De, 2007). Consumer needs can be
classified into what is called functional, symbolic and hedonic needs. The functional needs are those
needs referring to solving consumer problems, e.g.: consumers buy electric cars to drive from one
place to another without exhausting the environment.
Symbolic needs refer to the way how people see themselves, and here status plays a major role.
Some people may buy an expensive or luxurious electric car or an eco-friendly laptop to show their
Lastly hedonic needs reflect the consumers’ desires for sensory pleasure; some will buy a laptop
from Apple, because of the quality or the user-friendliness it provides.
Needs/goals can be further classified into approach or promotion goals, and avoidance or
prevention goals (Pelsmacker, Patrick De, 2007). For example a customer can decide to shop at an
Apple store because it offers them a nice shopping experience (=approach, promotion) or because
they do not have to drive far (=avoidance, prevention). A consumer who on the other hand is
planning to buy an electric car is probably motivated to process marketing communications on
electric cars – however, the needs or higher order goals that this particular consumer is pursuing
have an impact on information processing and the benefits he or she is receptive to (Huffman, C.,
Ratneshwar, S. and Mick, D.G., 2000).
If the consumer is mainly driven by functional needs, he or she may want clear information on
safety, price, electricity consumption etc., while a status appeal or an advertisement showing
driving sensations is more relevant for a consumer who is driven by symbolic and hedonic needs.
By motivation is meant a willingness to engage in behaviour, make decisions, pay attention, process
information etc. Previous social psychology analyses of personal relevance have labelled motivation
as ‘personal involvement’, ‘ego-involvement’, ‘vested interest’, issue-involvement’ and ‘personal
meaning’ (Petty, Richard E. & Cacioppo, John T., 1986). Personal meaning, or relevance, is when
people expect the issue to have significant consequences for their own lives. Relevance/motivation
can in this respect vary in magnitude and the duration of consequences can also differ. For example,
some advocacies may remain high in personal relevance for many people over a long period of time
(e.g. changing the income tax structure in Denmark), while other advocacies may have personal
relevance for a more circumscribed period and/or audience (e.g. when college tuition is raised), and
still other advocacies may have personal relevance only under certain very transient conditions (e.g.
an electric car ad may only have personal relevance for a person, if he/she belongs to the specific
What is important for the experiment in the thesis is how people’s attitudes are affected by message
quality, when they differ in personal relevance for a product (laptop and electric car).
3.1.2. The Experiment, Theory; Elaboration Likelihood Model (ELM):
Richard Petty and John Cacioppo’s ELM is an influential model of attitude formation and change,
which describes how consumers make evaluations in both low- and high-involvement
circumstances (Kotler, Philip & Keller, Kevin Lane, 2006).
Motivation or personal relevance is especially important in considerations of the Elaboration
Likelihood Model (ELM). The level of consumer involvement becomes relevant because messages
in advertisements can affect consumers differently, when they differ in the level of involvement.
Consumer involvement can be defined in terms of the level of engagement and active processing
undertaken by the consumer in responding to a marketing stimulus e.g. from viewing an
advertisement or evaluating a product or service (Kotler, Philip & Keller, Kevin Lane, 2006).
In appendix 6, figure 1 it is possible to see a figure of how the ELM works. The model provides a
fairly general framework for organizing, categorizing, and understanding the basic processes
underlying the effectiveness of persuasive communications (Petty, Richard E. & Cacioppo, John T.,
Another advantage of the model is that it attempts to integrate the many seemingly conflicting
research findings under one conceptual umbrella. Petty and Cacioppo saw many of the different
empirical findings and theories in the field as emphasizing one of just two relatively distinct routes
to persuasion, the central and the peripheral route to persuasion.
In the central route of persuasion the attitude formation or change involves much thought and is
based on a diligent, rational consideration of the most important product or service information.
Issue-relevant information is used by the highly involved consumer who actively thinks about the
arguments in the message (De Mooji, 2005). According to the theory, consumers will follow the
central route only if they possess sufficient motivation, ability, and opportunity. In other words,
consumers must have the will to evaluate a brand in detail and have sufficient knowledge about the
brand, product or service in their memory. They must also be given sufficient time and the proper
setting to evaluate the brand. If any of these factors are missing, the consumer will tend to follow
the peripheral route and consider less central, more extrinsic factors in their decisions.
The other route is the peripheral route, where attitude formation or change involves comparatively
much less thought and is a consequence of the association of a brand with either positive or
negative peripheral cues. Examples on peripheral cues could be a celebrity endorsement, a credible
source, or any object that engenders positive (or negative) feelings. In theory, the peripheral route
generally also includes visual cues like the package, pictures, or the context of the message. In this
route of persuasion the attitude for the object/product can be changed without necessitating scrutiny
of the true merits of the information presented (Petty, Richard E. & Cacioppo, John T., 1986).
If we assume that we have created a control condition in which motivation or ability to process
issue-relevant arguments is rather low, subjects should show relatively little differentiation of strong
from weak arguments. If then on the other hand a manipulation is made where argument processing
is enhanced in a relatively objective manner, the subjects should have greater differentiation of
strong and weak arguments. A message with strong arguments which is pro-attitudinal should
produce more agreement when it is scrutinized carefully than when scrutiny is low. On the other
hand, a message with weak arguments should produce less overall agreement when scrutiny is high
rather than low. Consider now a situation where the subjects are processing the message arguments
quite diligently; these persons should show considerable differentiation of strong from weak
arguments. If an argument processing is disrupted, either because of reduced motivation or ability,
argument quality should be a less important determinant of persuasion (Petty, Richard E. &
Cacioppo, John T., 1986).
In respect to the experiment, the respondents will be seen as low in motivation if they are scoring
low on Q1 in the questionnaire; the New Ecological Paradigm (NEP). In other words, they are
hypothesised to be low in motivation to elaborate on the information of a green product, because
they have not obtained a sufficient high score on the items from the NEP. Therefore we can say that
those subjects who are not motivated can be grouped into a low-involvement group, where
argument quality does not lead to attitude formation through the central route of persuasion.
Petty and Cacioppo further suggested that as personal relevance increases, people become more
motivated to process the issue-relevant arguments presented. Because of the greater personal
implications people should become more motivated to engage in the cognitive work necessary to
evaluate the true merits of the proposal. These consumers, who are highly motivated and who see
the personal relevance in the arguments presented, are also more likely to be familiar with the issue
at hand and may have more topic-relevant knowledge. Thus in addition to having a greater
motivation to process the relevant arguments, they may also have greater ability to do so. This
means that when messages contain information that is consistent with the subjects’ initial attitudes,
high relevance/high involvement subjects should be more motivated and generally more able to
elaborate the strengths of the arguments.
Appendix 6, figure 2 shows in box number 5, the effects of personal relevance on attitudes
following pro- (strong) and counter-attitudinal (weak) messages. It is shown that a contrast effect is
present, because when the message is strong and pro-attitudinal, the attitudes will be greater for the
high involvement group, because they have strong beliefs which are confirmed by the message,
while a weak message which is counter-attitudinal makes the high involvement group’s attitude to
drop more than for the low involvement group. This drop is caused by the high involvement groups’
ability to form counter-arguments to a message which is not confirming the views this group has.
The low involvement group is not able to elaborate the message in the same way. Petty and
Cacioppo are firstly explaining that involvement/personal relevance was believed to be associated
with a greater probability of message rejection, because people were holding expanded latitudes of
rejection as personal involvement increased. Therefore incoming messages would more likely be
judged as unacceptable, and the attitude would not be increased after exposure to these arguments.
One significant distinction was made in regard to this: the nature of the argument. Counter-
attitudinal arguments caused resistance when involvement increased, because the arguments
presented were contrasted with the high-relevance person’s beliefs and attitudes. On the other hand
there was a higher probability that attitudes where increased if they were pro-attitudinal, so the
arguments instead of being rejected where assimilated, so the arguments were seen closer to one’s
own position and therefore more acceptable.
To check if the experiment holds when the arguments are not manipulated to be pro- and counter-
attitudinal, Petty and Cacioppo made a test where only counter-attitudinal messages where used.
The messages varied by their quality, so some of them were strong while others were weak. This is
what we can see in appendix 6, figure 2, box 6. A message-relevance interaction indicates here that
as the personal relevance increases, subject’s attitudes and thoughts shows greater discrimination of
strong from weak messages. More specifically, when the message is strong, increasing relevance
produces a significantly increase in attitudes, but when the message is weak, increasing relevance
produces a significant decrease in attitudes.
So what Petty and Cacioppo found was that motivational variables are important in affecting the
likelihood of message elaboration.
One thing we have to keep in mind is that people can derive different meanings from the same
message. Some are contextual people who will see more in the message than is intended by the
sender of the message. In high-context cultures people are used to contextual messages, and they
will read more in the pictures and derive ‘hidden’ meaning from the visual image – in this way they
will try to construct metaphorical meaning which is not intended by the sender of the message.
Furthermore a study has shown that there are differences of print advertisements, when making a
multi-country comparison (Cutler, B. D., Javalgi, R. G., 1992). The comparison was made by
looking at advertising practices in countries like the United States, United Kingdom, France, Korea
and India; differences where found with respect to the size of the visual, frequency of usage of
photographs and product portrayals, the size of the product, usage and metaphors, and frequency of
persons in general, here especially the use of women and children depicted in advertising. However,
it is assumed that the respondents in this experiment read the messages in a similar way, and that
they only are affected by their level of involvement in respect to the NEP.
So the goal of this experiment is to analyse the attitude formation by the respondents. The reason
why attitudes toward a given product are important for a marketer is because of the possibility that,
the more favourable the attitudes are, the more likely it is that a brand is chosen. The major
challenge for marketing communications is to change attitudes in favour of the company’s product.
3.1.3. The Experiment, Theory; Attitudes:
Attitudes can be defined as general evaluations people hold in regard to themselves, other people,
objects, and issues (Petty, Richard E. & Cacioppo, John T., 1986). These evaluations can then be
based on a variety of behavioural, affective, and cognitive components, which are capable of
influencing or guiding behavioural, affective, and cognitive processes. Attitudes play important
roles in hierarchy-of-effects models, but in these models attitudes are primarily defined as affective
reactions in a hierarchical setting (Pelsmacker, Patrick De, 2007). The cognitive component reflects
knowledge, beliefs and evaluations of the object, the affective component represents the feelings
associated with the object, and the behavioural component refers to what is called action readiness,
which is a behavioural intention with respect to the object. An example may clarify these distinct
components: you may love a laptop from Apple (affective component) because you know that it is
convenient to use, or because you like the design of the laptop. Because of this you will intend to
buy an Apple laptop the next time you go shopping (behavioural component).
Marketers can change the attitudes of customers by changing one of the three components. If a gap
exists, Apple may stress that their laptops are cool and stylish, thereby they are trying to influence
the feelings (affective component) associated with the laptops by image building. Communications
campaigns trying to influence the consumer with the affective component often use emotional
advertisements containing no or very few product arguments.
However, if the goal of the campaign is to address the benefits by a product, for example the
benefits of an electric car and its impact on the environment, marketing communications may use
many, strong arguments to illustrate the numerous benefits of the car, thereby affecting the
One of the early theories of how advertising works explained that people would first learn
something about a product or brand, then form an attitude or feeling, and consequently take action,
which means purchasing the product or at least going to the shop with the intention of buying (De
Mooji, 2005). This was summarized as ‘learn-feel-do’, but later it was seen as mainly applicable to
products of high-involvement, such as cars, for which the decision-making process was assumed to
be highly rational.
In contrast there are low-involvement products, such as detergents or other fast-moving consumer
goods, where the consumer is less involved in behaviour, because there is little interest in the
product. The low-involvement sequence was assumed to be ‘learn-do-feel’ instead of ‘learn-feel-do’.
Again, knowledge comes first after that purchase, and only after having used the product, one
would form an attitude. The FCB planning model suggests four sequences in the process by which
advertising influences consumers:
where the two first sequences are related to high involvement; the third and fourth sequences are
There are many other ways to influence consumers when using messages. Petty and Cacioppo
found variables like distraction, repetition of messages, the number of endorsers in the ads (to affect
the attitude of low-involvement groups through the peripheral route) and the need for cognition to
influence the attitudes. In this paper, the most relevant is the one explained about the degree of
In the experiment in this thesis, the goal is to analyse communication effects for an electric car and
for the new, sustainable laptops made by Apple. Both of these products can be seen as high-
involvement products, because consumers normally do not buy the products without engaging in an
extensive information search for product attributes. As explained earlier, respondents will in this
experiment instead be split into groups dependent on how involved they are in considerations of the
NEP. Exactly how this is done will be explained later.
3.1.4. Underlying theory for variable 1, Q2 and Q3 in the questionnaire:
Consumer behaviour is complex and it is a multidimensional phenomenon, which further is
complicated by the inclusion of sustainability concerns (K. Wells, Victoria et al., 2010).
Consumers’ sustainability related behaviours can be affected by their demographics, values,
attitudes, knowledge, goals and circumstances, and it is especially differences in the demographics
which are important here. It is furthermore unusual to find a behavioural influence (influence on the
environmental behaviour of a person) that is generic (means that it is not specific to a particular
environmental behaviour such as recycling), and it only appears in those models of consumer
behaviour developed to explain social or environmental consumption behaviour. One factor which
is generic is a sense of responsibility (K. Wells, Victoria, 2010). This means that consumer
perceptions of responsibility can influence their buying behaviour, which also was the reason that
the responsibility questions (Q3) are important to analyse. Consumer’s social responsibility has so
far received comparatively little attention in past research.
In Q2 respondents are asked about their pro-environmental behaviours (PEB’s) such as recycling,
energy saving, greener travel, and purchase habits.
In an experiment made by Victoria K. Wells, respondents were asked a range of questions regarding
consumers’ behaviours from reduction, reuse, and recycling behaviours to travel, shopping, and
energy-consumption behaviours. She used a measure, termed General Environmental
Responsiveness (GER), which has a minimum score of -47 and a possible maximum score of 79 (a
range of 126). Those scoring low reported more negative behaviours and those scoring high
reported more positive behaviours in regard to environmental friendliness. She and her colleagues
found among other things that females had higher GER scores than men. For the behavioural
questions (Q2) a comparison will be made between males and females in order to see if the
conclusion will be the same.
In yet another study conducted by Linda Steg, a theory called value-belief-norm (VBN) was tested.
Results confirmed a causal order of the variables in VBN theory, moving from relative stable,
general values to beliefs about human-environmental relations, which in turn affect behaviour
specific beliefs and norms, and acceptability judgements, respectively (Steg, Linda, 2006). The
model can be seen in appendix 7, figure 1. The VBN theory links value theory, the NEP and the
norm-activation model (NAM). The NAM focuses on the role of moral obligations to act in favour
of the common good, which is used to understand environmental behaviour. Like the NAM it has
been proposed that environmental behaviour results from person norms i.e. a feeling of moral
obligation to act pro-environmentally (Steg, Linda, 2006). In the VBN figure, the statements of the
NEP is used as an ecological theory from which beliefs about the adverse consequences of
environmental changes can be deduced. In the VBN figure we see a causal link where the chain
moves from relatively stable and general values to beliefs about human environment relation (NEP),
which in turn are believed to affect specific beliefs on consequences of environmental behaviour,
individual’s responsibility for these problems and for taking corrective actions. The very first
variable from the questionnaire ‘Climate change will negatively impact me during my life time’ can
be classified as an AC (awareness of consequences) variable (see figure 1, Appendix 10). The
model can in this study be used to look after differences in the mean values for the AR (ascription
of responsibility) variables from Q3. Hereafter a possibility is to analyze if there is any connection
to the actual behaviour of the respondents (Q2). Also this variable can be linked to the VBN theory,
so those who have high values on the variable ‘Climate change will negatively impact me during
my life time’ should also score higher on the AC variables (Q3). This can in turn lead to more
environmental-friendly behaviour. No regression analysis is made to test the causal order in the
model, but the mean values for each question will be analysed.
According to the scales used, it is necessary to make a few comments. Taciano L. Milfont and John
Duckitt explain in their paper, that there are hundreds of environmental attitude (EA) measures
available. In their paper they use the Environmental Attitudes Inventory (EAI) which has 12
specific scales that capture the main facets measured by previous research (L. Milfont, Taciano &
Duckitt, John, 2010). EA’s have in earlier research’s been seen as having a single higher order EA
factor structure, but the EAI takes two higher order factors into consideration, these are
preservation and utilization. Preservation refers to the general belief that priority should be given to
preserving nature and the diversity of natural species in its natural state, and protecting it from
human use and alteration. Utilization on the other hand, expresses the general belief that it is right,
appropriate and necessary for nature and all natural phenomena and species to be used and altered
for human objectives. EA can be characterized by its horizontal and vertical structure, where the
horizontal structure refers to the primary order or first order factors forming the structure of EA,
while the vertical structure refers to the higher order or secondary factors. There are several scales
measuring EA but no gold standard measure has been made. In the research there has been
consensus that EA structure is multidimensional and organised in a hierarchical fashion and there
has been no previous attempt to assess the overall structure of EA, therefore the EAI was developed
(L. Milfont, Taciano & Duckitt, John, 2010). The EAI is capturing both the vertical and horizontal
structure of EA by measuring 12 specific facets (the primary factors), which define a 2-dimensional
higher order structure of EA (preservation and utilization). The 12 facets are 12 different EA scales.
The variables used in Q2 are subscales to the scale from the EAI which is called ‘Personal
conservation behaviour’ and variable 1 in Q1, ‘Climate change will negatively impact me during
my life time’ is a sub-scale from EA scale ‘Environmental threat’. The responsibility variables from
Q2 are having, as already explained, the similar form as those used by Linda Steg to test the VBN
Chapter 3.2. The Experiment, Method
The analytical approach is a descriptive study, and gives the following guidelines of which
considerations one has to make in respect to the data (Arbnor, Ingeman, 1997):
1) designing methods for collecting the data
2) collecting the data
3) coding and arranging the data
4) reporting the results
We shall in this chapter look at the four points above, and an explanation will be given about the
design of the questionnaire. Furthermore, an exposition of the respondents and the data collection
process will be given.
The last point above; ‘reporting the results’ is devoted to a whole chapter (Chapter 3.3: The
experiment, Results’). This chapter begins after section 3.2.3: ‘Coding and arranging of the data’.
3.2.1. Designing methods for collecting the data
A questionnaire was developed in order to analyze AU students concern with environmental issues.
The first 15 questions in the questionnaire is a version of the New Ecological Paradigm, which has
been included in order to divide respondents into high and low ecological involvement groups to
test the theory of the ELM. The NEP was created by Dunlap and Van Liere (1978) to measure the
degree to which people view humans as part of nature rather than separate from nature, which is the
ecological worldview (www.conpsychmeasures.com).
Respondents were told in the beginning of the questionnaire, that their answers would be treated
completely anonymously, in order to get as unbiased results as possible. For each of the 15
statements from the NEP, respondents were told to rate their answers on a 5-point Likert scale
ranging from 1(strongly disagree) to 5(strongly agree). Likert scales were developed in 1932 as the
familiar five-point bipolar response, which most people are familiar with today (Allen, I. Elain,
2007). There is no wrong way to build a Likert scale, but the most important consideration is to
include at least five response categories. It is possible to increase the scale to a seven-point scale or
more. The seven-point scale has been shown to reach the upper limits of the scale’s reliability and
as a general rule it has been recommended that it is best to use as wide a range as possible. It would
always be possible to collapse the responses into condensed categories, if it is appropriate for the
analysis. The Likert scale in this paper was not extended to be a 7-point scale, because the 5-point
scale should be sufficient in accomplishing the purpose of this particular experiment.
The NEP consists of 15 items in total. Using a 5-point Likert scale therefore implies that scores can
get a minimum value of 15 and a maximum value of 75. A score of 75 indicates total acceptance of
There are five subscales within the Revised New Ecological Paradigm used in this questionnaire,
these are: reality of limits of growth, antianthropocentricism, rejection of exemptionalism, and the
possibility of eco-crisis. Each of these subscales consists of three items. The ‘reality of limits to
growth’ consist of items 1, 6 and 11 (keep in mind that these items correspond to variables 2, 7 and
12 in the questionnaire as the first item ‘climate change will negatively impact me during my life’
has to be treated separately, as it is not part of the NEP) and measures the respondents attitudes of
the reality of the limits of growth in the environment. The second subscale
‘Antianthropocentricism’ consists of the items 2, 7, and 12 (variables 3, 8 and 13 in the
questionnaire). It measures the respondents’ attitudes towards the fragility of natures balance. The
next subscale which is the ‘rejection of exemptionalism’ consists of items 4, 9, and 14 (variables 5,
10 and 15 in the questionnaire), which measures the respondents attitudes towards the rejection of
exemptionalism. In other words, these items describe the knowledge and ingenuity humans have,
which could prevent an ecological crisis.
Finally, the subscale ‘The possibility of Eco-crisis’ subscale contains items 5, 10 and 15 (items 6,
11 and 16 in the questionnaire) and measures the respondents attitudes towards the possibility of an
The second part of the questionnaire was made to analyze certain environmental behaviours by the
respondents. Again a 5-point Likert scale was used ranging from 1 (=never) to 5 (=always). An
example of a question in the second part of the questionnaire is: ‘I leave my electrically powered
appliances (TV, computer, stereo etc.) on stand-by.
In the third part of the questionnaire (Q3), respondents were asked to think about what to do about
the issue of increasing energy use and related carbon dioxide emissions, and how they felt about
whose responsibility it is to act upon it. Again a 5-point Likert scale was used, which ranged from 1
(=strongly disagree) to 5 (=strongly agree). This part consisted of five questions, e.g.; ‘I am jointly
responsible for the exhaustion of non-renewable resources’.
In the end of the questionnaire, the students were asked about their gender, age, nationality and
whether they were car-owners or not.
Although the main goal of the thesis is to test the Elaboration Likelihood Model (ELM), it is also
possible to see if there is a significant difference in regard to variable 1, environmental behaviours
from Q2 and the AR items from Q3 by comparing Scandinavian vs. foreign students and males vs.
The main goal of the thesis was to analyze how advertisements, by the use of weak and strong
arguments, affect the high/low involvement groups. In question 4a) respondents were asked to look
at the picture of an electric car (Tesla Roadster) and read the subtext, which in this case was made
with weak arguments. The Tesla roadster is an electric sports car which can drive from 0-60 mph in
only 3.7 seconds. Furthermore it can drive 245 miles on one single charge
(http://www.teslamotors.com). The subtext to the picture contained actual information, which can
be found on Tesla Motor’s website. Most importantly was to make arguments for Q4a) that were
weak. Inspiration of how weak arguments could be designed was obtained by looking at previous
research papers about high and low involvement groups. In an article by Marie Helweg-Larsen
called ‘Effects of Erotophobia on the persuasiveness of condom advertisements containing strong
and weak arguments’, weak arguments were of the more insignificant type: ease of use of the
condom, easy instructions, fashionable colours, and discrete packaging. These arguments have been
tested by the author to be weak, and therefore it was reasonable to make arguments, which had
somewhat the similar fashion. Therefore is the subtext to the Tesla Roadster ad in the weak
condition containing information about the convenience of the car with its heated seats, floor mats
and cup-holder. Some humour is also included in the ad by the sentence ‘burn rubber, not gasoline’.
Subjects who are not involved are more likely to focus on non-content cues as the rewards available
for adopting a certain attitude, the attractiveness, credibility, or power of the communication’s
source, and the number of others who advocate a certain position (Petty, Richard E. & Cacioppo,
John T., 1981). This allows a person to evaluate a message or decide what attitudinal position to
adopt without engaging in any extensive cognitive work relevant to the issue or product under
consideration. The credibility, power and attractiveness of the source (Tesla Motors) can serve as a
cue for the low involvement group. A celebrity could have been chosen, but as it was the desire not
to make the ad too artificial, the idea was ignored. As it was explained in the previous theory
section, the peripheral route generally includes visual cues like the package, pictures, or the context
of the message. In this route of persuasion the attitude for the object/product can be changed
without necessitating scrutiny of the true merits of the information presented (Petty, Richard E. &
Cacioppo, John T., 1986). The peripheral cue was in the experiment the picture combined with the
weak arguments of the Tesla Roadster car. The picture was meant to persuade the low involvement
group especially in the way of the design of the electric car ad.
After showing the advertisement, five questions were asked to it. Respondents were told to answer
how effective, how persuasive, how pleasant and how appealing the advertisement was, responding
on the 5-point Likert scale. Lastly they were asked how likely it would be that they would choose
the Tesla Roadster (given they wanted to buy a car).
In question 4b) the respondents were asked the same questions they answered for the Tesla
Roadster, but this time it was for the new, environmentally friendly family of Apple laptops. Again
a picture was shown in the ad for the low involvement condition. It has been chosen to include
Apple, because it is a completely different product than the electric car. Though are the two
products similar in one respect: they are both made in respect to minimize the impact they have on
the environment. It has also been necessary to add Apple in the experiment, because of the
respondents who were asked. As the respondents all were students, there is a possibility that many
of them do not own a car. Therefore it could be problematic for them to determine if the
advertisement for example is effective. Apple is on the other hand a firm which all students can
relate to, as almost everybody owns a laptop and has an opinion about Apple. The attractiveness,
credibility, or power of the communication’s source can again be seen as peripheral cues, which are
expected to influence the low involvement group. The picture in the advertisement is showing the
laptops with symbols that indicate the environmental friendliness of the computers. The subtexts
were again made, so they did not give too detailed explanations, and the text was made with
information containing non-significant product attributes in respect to environmental issues; colour,
size and design.
In question 4c) the respondents were told to look at the advertisement of the Tesla Roadster again.
The picture itself was not included in the high-involvement condition. Respondents were told to
read the subtexts carefully, as these were much more significant and detailed explanations of the
products; for example that the car has lithium ion batteries which reduce the energy used and
allowing driving 250 miles on one single charge. It is also three times better than the hybrid car and
four times better than the hydrogen fuel-cell car in respect to carbon dioxide emissions. Again the
subtexts contained information which is real.
Under high involvement, the focus of thought is on the content of the persuasive message. If the
communication presents arguments that are subjectively specious and subject to counter-
argumentation, resistance to persuasion and in worst circumstances boomerang will occur. If on the
other hand, the arguments are compelling for the respondents, which they are both in the case of the
Tesla Roadster and especially for Apple, respondents’ attitudes should be enhanced, so increasing
involvement can lead to either enhanced or reduced persuasion depending upon the quality of the
arguments presented in the message (Petty, Richard E. & Cacioppo, John T., 1981).
In respect to the Apple advertisement with strong arguments, question 4d, the arguments were also
designed as being much more detailed than in the previous, low-involvement, Apple-condition.
Here the main goal was to explain respondents that the body of the laptops did not contain any
harmful toxins, the glass was free of arsenic, the circuit board did not contain PVC and the LED
technology did not need any mercury. It was also written that the laptops were more energy
efficient, and that highly recyclable material in the body of the computers allowed reuse. Again the
same questions were asked in order to look after any differences to compare it with the Apple
advertisement containing weak arguments.
3.2.2. Collecting the data:
When the questionnaire was made, it was time to distribute them to the respondents. Most of the
questionnaires were distributed at different classrooms at the ‘Aarhus School of Business’, but some
were also answered by students from the department of ‘Anthropology, archaeology and
Linguistics‘. It was possible to get diversified answers by giving the questionnaire to students with
different study backgrounds. If only one particular class was asked to fill out the questionnaire, we
could not be sure that the answers given would be that different from each other.
161 answers were collected. The questionnaires were handed out manually as this was seen as the
fastest way to receive the data. Initially, more questionnaires were collected, but as some of them
only were partly answered, they were excluded from further analysis.
3.2.3. Coding and arranging the data:
The research groups are the high and low involvement groups. The respondents were divided into
one of the two groups in accordance to the answers they gave in respect to the NEP (items 2 – 16 in
Q1 of the questionnaire). It was initially the thought that respondents were split by a value of 2.5 as
this is the middle point in the Likert scale, so all respondents who fell under this threshold value
were considered low involvement. But the problem was that not enough respondents had a value
under 2.5, so in this case there would be less research units in the smallest group than dependent
variables, thereby a violation of the sample size assumption, which soon will be discussed in more
detail. It was a necessity to make a manipulation of the data. This was done by taking the average of
the respondent’s mean values of the NEP. This resulted in a value of 3,4485. Respondents who’s
average of NEP fell into values between 0 – 3,4485 were put into the low involvement group. Each
respondent was clustered so they belonged to the dummy variable ‘1’, which indicated that they
belonged to the low involvement group. The rest of the respondents were labelled with a ‘0’, which
indicates the high involvement group.
It was necessary to recode some of the variables into different variables, so that they were similar to
the other questions they are a part of. Tables from appendix 7 – 9 show all recoded variables and
their new variable name. Please note that recoded variables from the NEP can be found in appendix
8, table 1. In appendix 7, table 1 recoded items from Q2 are shown, and items from Q3 are shown in
appendix 9, table 1. Variables’ new names starting with ‘SMEAN’ are indicating that they have
been treated for missing values with the ‘series mean’ method is SPSS. This means that empty cells
has been replaced by a mean for the particular item. Variables labelled ‘Reversed’ are variables
which are recoded, e.g. a variable which originally had a value of 5 got a new value of 1. In this
way variables were recoded so they matched each other. Variables named SMEAN(Reversed) are
variables which have been both reversed and treated for missing values.
Chapter 3.3. The Experiment, Results
In this chapter the results of the analysis will be shown. First reliability and validity is discussed.
After this we will go through certain assumptions, which have to be fulfilled for using the profile
analysis for testing the theory of the ELM. Since the profile analysis is a straightforward method
used to look for differences between groups (profiles) it will not be necessary to give a detailed
description of it. What the profile analysis is all about will be apparent when going through the
three steps, which the method consists of.
Variable 1, Q2 and Q3 in the questionnaire will be analyzed by making independent-samples t-tests
in the end of the chapter. This is done by analyzing the sum of Q2 and Q3 in order to analyze
overall differences between males versus females and Scandinavians versus other nationalities. Also
an analysis will be made of each item from the questionnaire, as it is interesting to see on which
particular variable the groups differ.
3.3.1. Reliability and validity:
The analysis has to be checked for reliability and validity. Reliability is a test made to discover
whether the variance in our variables is at an acceptable level. The 5-point Likert scale used is said
to be reliable if it gives the same results under repeated use (Arbnor, Ingeman, 1997). To test for
reliability several options are available, for instance the test-retest method, where we conduct the
same test more than once. We can also conduct two similar, parallel tests at about the same time,
and yet another test is called split halve, which means that a test is divided into two halves, each
containing similar questions, and correlations are analyzed to see how well these two halves’s fit
In the thesis it has been chosen to use Cronbach’s alpha as a measure for reliability. The variables
are said to be reliable if they exceed a value of 0,7. Running the analysis, gives us the results which
can be seen in appendix 9, table 1-4. A value of 0,880 for the variables asked for the advertisements
is obtained, which is a high value for reliability. The N items are the dependent variables which will
be used in the profile analysis (10 variables for the electric car advertisement and 10 questions for
the Apple advertisements). The reliability test for the NEP, which was the basis for classifying the
respondents into groups is equal to 0,7 (by rounding it up from 0,699 to 0,7). Reliability tests for Q2
shows a value of 0,497 which is relatively low. This means that we can not be sure to get the same
results for Q2, if the same analysis is made for Q2 in another point of time. The same is true for Q3
which shows a value if 0,654. Although not all parts of the questionnaire are reliable, the
experiment will still be relevant, as we are measuring what the respondents answered at this
The data also has to be valid. Validity and reliability are closely related – reliability is a condition
for achieving valid measures. It is obvious that we cannot get precise measures if they are affected
by coincidence. Validity becomes in this way a measurement of the extent to which the results are
correct or true (Arbnor, Ingeman, 1997).
There are four ways of evaluating validity. From these four there are two ways of measuring
qualitative validity: face validity and content validity (Sartori & Pasini, 2006) and two ways of
measuring quantitative validity: criterion-related validity and construct validity. I will briefly go
through them here in relation to the report.
Criterion-related validity is based on the coefficient of correlations between two tests. Since I have
not been in possession of any other tests which measures exactly the same variables as I have in the
questionnaire, this option was not available.
Construct validity is used when theoretical, non-observable construct or trait is to be measured. In
this report, I am measuring the attitude of green advertisements on a 5-point Likert scale. Therefore
I am not measuring a non-observable construct, which is the reason why it is irrelevant here. This
leaves me to explain the quantitative measures of validity, face validity and content validity.
Face validity is the simplest form of evaluating validity. It is a subjective way of rating validity. In
this method we are investigating whether the questions asked in the questionnaire really are
measuring, what we want to measure. For example, do the questions asked for the liking of the
advertisements etc. represent the attitude of the respondents? Furthermore can it be seen as
reasonable to ask the respondents questions of the NEP, and use these answers to group them into
either a low involvement or high involvement group? The face validity criterion is fulfilled here;
when questions are asked about how effective the advertisement is, how much the respondents liked
it and if it was persuasive, these answers are also telling us about the attitude to each advertisement.
The NEP is also seen as relevant as it serves as an indicator of how environmental aware the
respondents are in general. Respondents who are mostly environmental aware, should also have
more favourable attitudes and have the ability to scrutinize advertisements with strong arguments.
In face validity, the goal was to test if the test appears as being valid from a subjective point of view.
In content validity one goes further to see if we have included all relevant variables. Haynes defines
content validity as ‘the degrees to which elements of an assessment instrument are relevant to and
representative of the targeted construct for a particular assessment purpose’ (Sartori & Pasini,
2006). This means that every single variable in the questionnaire should describe, what we want to
analyse. The coherence of the variables in the questionnaire is seen as satisfying when analyzing
their behaviour, attitudes etc. If for example only attitudes were asked for the electric car
advertisement, we could not be sure that the results could have been used solely to explain the
theory or the ELM, because many of the respondents, as discussed, can have problems in referring
to a car. Also it would not have been sufficient to include ad variables containing only weak
arguments for explaining the ELM model.
Normally test for content validity is conducted through expert opinions about each item. The NEP
measurement used to classify the respondents into groups is internationally recognised, and the
subtexts used for the advertisements are inspired by weak and strong arguments from other research
papers. In regard to the questions asked for each advertisement, no expert opinion as such has been
necessary. Still the layout of questionnaire has been discussed with my advisor in Denmark.
So based on face validity and content validity, the conclusion is that the data is valid. Overall the
data is assumed to be both reliable and valid. This can be illustrated by looking at appendix 9, figure
3.3.2. Assumptions for the profile analysis:
The statistical method used to analyse the differences between groups has been chosen to be the
profile analysis. Profile analysis is a special application of multivariate analysis of variance
(MANOVA) to a situation where there are several dependent variables, all measured on the same
scale (Tabachnick, Barbara G., 2007).
Sample Size, Missing Data, and Power (Tabachnick, Barbara G., 2007):
Firstly, it is important to have a sufficient sample size. The sample size in each group is important
in profile analysis (as in MANOVA), because there should be more research units in the smallest
group, than there are dependent variables. This is important both because of considerations of
power and for evaluation of the assumption of homogeneity of variance-covariance matrices.
The dependent variables correspond to the questions asked in the questionnaire. In this case the
questions are those asked for each advertisement in the questionnaire. For the car advertisement 10
questions were asked (5 for the ad with weak arguments and 5 for the ad with strong arguments)
and 10 for the Apple ad. 79 respondents were classified into the low involvement group which
leaves the rest 82 respondents belonging to the high involvement group. Thus, the sample-size
assumption is not violated, because there are more units in the smallest group (79) than dependent
Unequal sample size does not provide any difficulty when making the profile analysis. In this case
though, the cases are split almost in the middle (82 vs. 79).
When the data was typed into SPSS, it was obvious that there were some missing values. Missing
data is one of the most pervasive problems in data analysis. Missing data can be caused by some
respondents being unable to understand a question, or because they want to finish the surveys as
fast as possible. The seriousness of missing data depends on the pattern of missing data, how much
is missing, and why it is missing (Tabachnick, Barbara G., 2007). The pattern of missing values is
very important; if the missing values are scattered randomly the problem is less serious than if there
are non-randomly missing values. Non-randomly missing values are serious no matter how few of
them there are, because they affect the generalizability of the results. Fortunately, all the missing
values in this experiment have been randomly missing values (it simply seemed as some of the
respondents forgot to answer few of the items). As already explained in the thesis, all the missing
values were replaced by a method called ‘series-mean’ in SPSS.
Larger sample sizes produce greater power (Tabachnick, Barbara G., 2007). The power is not
lowered if there are more cases than dependent variables in every cell because of reduced degrees of
freedom for error. In this experiment there are more cases than dependent variables. Again it can be
argued that it could have been better with a larger sample size, but due to limited time the sample
size of 161 respondents has been considered to be sufficient.
Multivariate Normality (Tabachnick, Barbara G., 2007):
Profile analysis is robust to the violation of normality, which is also the case for other forms of
MANOVA. There are only problems with multivariate normality if there are fewer cases than
dependent variables in the smallest group and highly unequal sample size between the groups. If
these are not present, which they are not in this experiment, deviation from normality of sampling
distributions is not expected.
Frequency tables of the dependent variables can be seen in the appendices 10-12. Looking at these
histograms, it is possible to see that the populations are approximately following the normal
distribution. One reason why some of the variables show skewness could be that the sample size is
not big enough.
It is possible to transform some of the variables, for example by taking the logarithm to the variable
if there is substantial positive skewness or to take the square root of a variable which has moderate
positive skewness (Tabachnick, Barbara G., 2007), like variable 41: Tesla, weak; ‘How likely is it
that you would choose the product’. Other methods are applied when the distribution is negatively
skewed or J-shaped. However it has been decided not to transform any of the variables. The reason
is that transformations normally are recommended as a remedy for e.g. normality but they are not
universally recommended. Transformed variables can also be much harder to interpret.
It has been tried to make frequency histograms like those above after variable-transformation, but
the distributions were not that different from those included in appendices 10-12. After all the
variables do not diverge that much from the normal distribution, and therefore it is not seen as a
problem for analysing the mean values without getting biased results.
Absence of outliers (Tabachnick, Barbara G., 2007):
Profile analysis is extremely sensitive to outliers as in all MANOVA. As a 5-point Likert scale is
used here, the impact of outliers will not be large enough to be considered as being an issue.
Multivariate outliers can occur when several different populations are mixed in the same sample or
when some important variables are omitted that, if included, would attach the outlier to the rest of
the cases. Different populations are mixed in the same sample here, but the populations do not differ
that much from each other causing outliers to be present.
Box plots have been made and can be seen in appendices 12-13. The plots show that there are some
outliers present, indicated by the numbers underneath the box. Each number represents specific
respondents and each box plot refers to variables were outliers are present.
However it is assumed that the outliers reflect the true opinions of the respondents, and therefore
they are included in the analysis.
Homogeneity of Variance-Covariance Matrices (Tabachnick, Barbara G., 2007):
Homogeneity and its opposite heterogeneity arise in describing the properties of the dataset. It is
related to the validity of the assumption that the statistical properties of any part of an overall
dataset are the same as any other part. Homoscedasticity (called homogeneity of variance in
grouped data) is related to the assumption of normality, because when the assumption of
multivariate normality is met, the relationship between the variables is homoscedastic.
This assumption can be disregarded as well, because the smallest sample size is much bigger than
the number of depended variables, and also because the sample sizes are practically equal.
To be completely sure, Box’s M test has been made (see appendix 12, table 1 & 2). The assumption
is not violated if the p-values is above 0,05. Since the p-value for the electric car ad is 0,273 and
0,108 for Apple, it is concluded that homogeneity of the variance-covariance matrices exist, and
therefore the assumption is not violated.
Linearity (Tabachnick, Barbara G., 2007):
For the parallelism and flatness test performed later in the profile analysis, linearity of the
relationships among the dependent variables is assumed. The issue can be ignored because there are
many symmetrically distributed dependent variables and a relative large sample size. If the sample
size had been very small the major consequence of failure of linearity would be loss of power in the
Furthermore the variables are not considered to deviate that much from the normal distribution
which gives reason to say, that linearity exists as well.
Absence of Multicollinearity and Singularity (Tabachnick, Barbara G., 2007):
When there is correlation among the dependent variables and the correlation is high, one dependent
variable is a near-linear combination of other dependent variables: this means that the dependent
variable provides information that is redundant to the information available in one or more of the
other dependent variables. To check for this it is necessary to make a correlation matrix of the
relevant variables. This is shown in appendix 14, table 1.
Multicollinearity and singularity are problems with a correlation matrix that occur when variables
are too highly correlated, which is when a correlation exists that is 0,90 or above. Multicollinearity
and singularity cause both logical and statistical problems. It is a bad idea to include redundant
variables, because they are not needed and they inflate the size of error terms, so they actually
weaken the analysis. None of the correlations in the matrix is exceeding 0,4, and as this is far away
from the threshold level of 0,90, the assumption can be disregarded.
3.3.3; Results of the experiment:
The profile analysis will now be carried out, both for the electric car and the Apple condition. The
analysis is based on three steps (Tabachnick, Barbara G., 2007):
Step 1: Parallelism: Tests of parallelism and flatness is conducted through hypotheses
about adjacent segments of the profiles. The test of parallelism, for example asks if the
difference (segment) between the first dependent variable; Variable 37: ‘How effective do
you think the advertisement is, weak’ and the second dependent variable; Variable 38:
‘How persuasive do you think the advertisement is, weak’ is the same between the high
and low involvement group. The analysis of each segment will be carried out across all
dependent variables and for both conditions. The test is about analyzing if the groups
attach different weights to the factors (Cand. Merc. Manual for SPSS).
Step 2: Differences in levels: this test examines differences between the combined means
of the two groups (high vs. low involvement) combined over all 20 dependent variables.
First the level test examines the means for the electric car condition, were the first five
variables contain weak arguments and the last five variables contain strong arguments.
Exactly the same thing is done for the Apple condition afterwards. More precisely, the
level test makes an average of all the means from each dependent variable. If we in the
previous step find that both groups attach equal weight to the factors, there is still a
probability that there is a difference in the levels (Cand. Merc. Manual for SPSS).
Step 3: Flatness: This test examines if the profiles for the two groups are horizontal,
meaning that the slope of the profiles equals zero. If we find that the profiles are not
parallel (step 2), this test can be ignored. In other words we are testing if the groups find
that the factors have the same impact/weight on their attitudes from one dependent variable
to the other (Cand. Merc. Manual for SPSS).
The profile graphs for the Tesla Roadster and the Apple ad’s can are available in appendix 15,
figure 1 and 2.
Step 1: Parallelism:
The test results for the electric car ad shows the following SPSS output:
Table 2: Multivariate Test Results ‘Parallelism’, Tesla Roadster
Value F Hypothesis df Error df Sig.
Pillai’s trace .057 1.011 9.000 151.000 .433
Wilks’ .943 1.011 9.000 151.000 .433
Hotelling’s .060 1.011 9.000 151.000 .433
Roy’s largest .060 1.011 9.000 151.000 .433
The test results show that the hypothesis about parallelism can not be rejected (F= 1.011,
p=0.433>0.05). This means that there are no differences in the slopes between each segment, and
therefore the profiles are parallel in respect to the Tesla Roadster ad.
For the Apple ad we get:
Table 3: Multivariate Test Results ‘Parallelism’, Apple
Value F Hypothesis df Error df Sig.
Pillai’s trace .111 2.087 9.000 151.000 .034
Wilks’ .889 2.087 9.000 151.000 .034
Hotelling’s .124 2.087 9.000 151.000 .034
Roy’s largest .124 2.087 9.000 151.000 .034
The profiles for the Apple ad are not parallel (F=2.087, p=0.034<0.05). Therefore is the flatness test
in step 3 ignored in the Apple condition.
Step 2: Differences in levels:
After the syntax has been specified in SPSS, the following output is retrieved for the electric car
Table 4: Test Results ‘Differences in levels’, Tesla Roadster
Source Sum of Df Mean Square F Sig.
Contrast 6578.614 1 6678.614 1.206 .274
Error 867179.083 159 5453.956
The results for the Tesla Roadster indicate that there is no difference in the levels (F=1.206,
p=0,274>0,05). After the whole profile analysis has been carried out an independent t-test will be
made, to show if there are significant differences between the profiles on each item. This also
allows us to see the exact mean on each item.
Table 5: Test Results ‘Differences in levels’, Apple
Source Sum of Df Mean Square F Sig.
Contrast 16135.399 1 16135.399 2.930 .089
Error 875681.259 159 5507.429
The test results for the ‘difference in levels’ test in case of Apple shows that no difference in the
levels are found if we use a significance level of 5 %. However, results do indicate difference in the
levels if we use a significance level of 10 % (F=2.930, p=0.089<0,10). This means that with a
significance level of 10 %, we conclude that there are differences between the combined means
taking all the dependent variables from the Apple condition into consideration. Exactly which items
are different from each other will be analysed after the profile analysis is finished.
Step 3: Flatness:
As the profiles were found not to be parallel in consideration of the Apple ad it will be ignored. But
in respect to the Tesla Roadster ad’s, the flatness test will be performed, because the profiles in this
case were found to be parallel and congruent (no difference in the levels). This results in the
Table 6: Multivariate Test Results, ‘Flatness’, Tesla Roadster
Value F Hypothesis df Error df Sig.
Pillai’s trace .631 28.730 9.000 151.000 .000
Wilks’ .369 28.730 9.000 151.000 .000
Hotelling’s 1.712 28.730 9.000 151.000 .000
Roy’s largest 1.712 28.730 9.000 151.000 .000
From the results we can see that the profiles are not horizontal (F=28.730, p=0.000<0.01).
This means that the slopes between each segment are different from zero.
Before any conclusion can be made in respect to the ELM model, we have to test each item with the
independent-samples t-test. This will be done in the next section.
3.3.4: Assumptions for the independent-samples t-test, Tesla Roadster & Apple ad:
First, it is necessary to go through a number of assumptions for using the independent samples t-test,
these are (http://lap.umd.edu/psyc200/handouts/PSYC200_0812.pdf):
Independence: Observations within each sample must be independent. In other words they should
not influence each other. If two students for example were to fill out the questionnaire together, or if
they have discussed the issues of the questionnaire with their fellow students, independence can be
a problem. This could have been avoided to some degree if the questionnaire was sent to each
student’s individual e-mail, but in this case the questionnaires were handed out manually in the
lectures. However, the problems that can occur in a test when dependence prevail, is assumed to be
absent here. The students seemed focused in answering the questions by their own after the
questionnaires were handed out.
Normality: The scores in each population must be normally distributed. A normality test can be
done or histograms for the dependent variables can be examined to see if the assumption is violated.
The normality assumption has previously been discussed, and it is overall seen as not being violated.
Homogeneity of Variance: The two populations must have equal variances. This means that the
degree to which the distributions are spread out is approximately equal. When the t-test results are
shown, the Leviene’s test, which also is available in the original SPSS tables in the appendix, will
show if variance homogeneity is present or absent. The choice of the t-statistic in the table depends
on this assumption.
The hypothesis for variance homogeneity can be written as:
H0: σ21 = σ22 =…. σ2r
H1: At least two are different
3.3.5: Hypotheses and independent-samples t-test results for the Tesla Roadster & Apple ad’s:
The hypotheses to be tested for the means of the Tesla Roadster and the Apple ad can be defined as:
Where j represents the sample number; 0=high involvement, 1=low
involvement and k represents the variable number, μ symbolizes the mean
The first variable to be tested is variable 37, Tesla, weak: ‘How effective do you think the
advertisement is?’, weak arguments, and this can be defined as:
H0: μ0,37 = μ1,37
H1: μ0,37 ≠ μ1,37
In the similar way, the mean is tested for variable 38 – 41, which are variables for the Tesla
Roadster ad with weak arguments. For the Tesla Roadster ad with strong arguments variables 47-51
Table 7: Description of significant results p-value < sig. level Mean (M)
Profile analysis: ‘Parallelism’, Apple 0,034 < 5 % -
Profile analysis: ‘Difference in levels’, Apple 0,089 < 10 % -
Profile analysis: ‘Flatness’, Tesla Roadster 0,000 < 1 % -
Variable 48: Tesla, strong: ‘How persuasive do you t=1,891 High inv.: μ = 4
think the advertisement is?’
0,06 < 10 % Low inv.: μ =3,68
Variable 50: Tesla, strong: ‘How appealing do you t=2,338 High inv.: μ = 3,94
perceive the advertisement?’
0,021 < 5 % Low inv.: μ =3,54
Variable SMEAN(52_apple_how_effective_strong) t=2,388 High inv.: μ =4,183
0,018 < 5 % Low inv.: μ =3,823
Variable 53: Apple, strong: ‘How persuasive do you t=2,482 High inv.: μ =4,12
think the advertisement is?’
0,014 < 5 % Low inv.: μ =3,72
Variable 54: Apple, strong: ‘How pleasant do you t=1,818 High inv.: μ =4,10
perceive the advertisement?’
0,071 < 10 % Low inv.: μ =3,82
Variable 55: Apple, strong: ‘How appealing do you t=3,145 High inv.: μ =4,24
perceive the advertisement?’
0,002 < 1 % Low inv.: μ =3,75
Variable 56: Apple, strong: ‘How likely is it that you t=2,271 High inv.: μ =4,04
would choose this product (given you want to buy a
laptop)?’ 0,025 < 5 % Low inv.: μ =3,58
From appendix 16, table 1 a red circle shows where a significant mean is found when the high
involvement group is compared with the low involvement group. The corresponding means can be
seen from appendix 16, table 2. The original SPSS table has been included for all the tests
throughout the report, as documentation for the results, in the appendices, but tables with significant
results, like the one shown above, will be included in the main report.
Variable 48: ‘How persuasive do you think the advertisement is?, strong arguments’ is significant if
we use a 10 % significance level. As can be seen from the table above the high involvement group
scored the highest mean. In other words, the high involvement group is more persuaded than the
low involvement group, when the arguments are strong, which also was what was expected.
Variable 50: ‘How appealing do you perceive the advertisement?, strong arguments’ also shows that
the high involvement group think that the ad is more appealing, when strong arguments are used.
The means for the groups are relatively close to each other, but still significant results are found.
For the Apple ad we have for the weak arguments: variable 42-46 are tested. Finally, for the Apple
ad with strong arguments variables; 52-56 are tested.
Now, it is interesting to see if the results differ when Apple is taken into account. The original
independent-samples t-test can be seen in appendix 17, table 1. Again significant differences are
marked with a red circle, but all significant results are shown in the table above. All variables with
strong arguments is this time showing a significant mean difference; variable SMEAN_52: ‘How
effective do you think the advertisement is?, strong arguments’, variable 53: ‘How persuasive do
you think the advertisement is?, strong arguments’, variable 54: ‘How pleasant do you perceive the
advertisement?, strong arguments’, variable 55: ‘How appealing to you perceive the advertisement,
strong arguments’ and lastly variable 56: ‘How likely it is that you would choose this product,
This time we see that argument strength has an impact on the high involvement group, which also
was expected according to the theory proposed by the ELM. Again it is important to stress that the
mean differences are not large, but there is evidence that the more environmental aware the
respondents are, the higher the chance is that they are affected by ads with strong environmental
A clearer confirmation of the theory proposed by the ELM is present for the Apple ad compared
with the Tesla Roadster ad, because the high involvement group’s attitudes are increased on all
dependent variables with strong arguments, whereas significant differences only are found for 2 of
the variables with strong arguments from the Tesla Roadster ad. A reason could be that more
respondents are able to relate to a product that is more relevant for them.
Before all the hypothesis are summarized, we need to test for differences the very first variable in
the questionnaire; variable 1: ‘Climate change will negatively impact me during my life time’,
variables 17 – 29 from Q2 (the behavioural questions) and variables 30 – 36 from Q3 (the questions
which reflect the environmental responsibility of the respondents). This will be done in order to see
if there is a connection to the VBN theory.
It will not be possible to test for causal links between the variables with a t-test. But because
previous research has confirmed that causal links exist between awareness of consequences,
ascription of responsibility and pro-environmental behaviour, it can be useful to see if females
differ from males on the variables from the VBN theory. The same tests will be made comparing
Scandinavians respondents with those from other nationalities.
The assumptions for the above described t-tests are the same as when items for the ads were tested
in for the items from the two ads, as the respondents are the same as well as the way the
questionnaires were answered. Instead of making histograms to check for normality, as this would
involve too many tables in the appendix, I will argument that normality is present with the ‘Central
Limit Theorem’ in mind (Tabachnick, Barbara G., 2007). The Central Limit Theorem says that with
a sufficient large sample size, sampling distributions of means are normally distributed regardless of
the distribution of the variables. Furthermore, there should be more than 20 degrees of freedom. In
this case we have 159 degrees of freedom, so the tests are robust to violations of normality.
As it has been discussed already, some of the variables from Q2 and Q3 were reversed; those who
for example answered 5 to a specific variable got the value of 1 instead. In this way, all questions
are expressed with a concern for the environment. This makes it possible to make the test by
summing the variables, so we can see if there overall is a difference between for example males vs.
females and Scandinavians vs. Other nationalities.
3.3.6. Independent-samples t-test for variable 1, group: ‘females’ vs. ‘males’ and
‘Scandinavians’ versus ‘Other nationality/foreigners’:
We start testing variable 1. The hypotheses, when comparing males with females, can be defined as:
Here j stands for the group, where 0=females and 1=males, k represents the
variable number, which in this case is variable 1.
For variable 1, the alternative hypothesis can be defined as: Halt.: μ0,1 ≠ μ1,1
In appendix 18, table 1 we see that there are no differences in the means (M) on the variable 1
(t=0,370, p=0,712>0,05). The μ =3,61 for males and μ =3,55 for females, indicating that they on
average were ‘unsure’ – ‘mildly agree’ on the question (appendix 18, table 2).
The same variable is tested, but this time we look at Scandinavians versus foreigners:
H23: μ0,1 ≠ μ1,1 Similar as before the first number, 0 = corresponds to foreigners and 1 to
Scandinavians and the last number corresponds to the variable number.
The test results can be seen in appendix 18, table 3. No significant difference was found either (t=-
0,571, p=0,569>0,05). Again respondents answers correspond to the area of ‘unsure’ – ‘mildly
agree’ with μ = 3,64 for other nationality and μ = 3,55 for Scandinavians.
3.3.7. Independent-samples t-test for variable SUMQ2, Q2 items, groups: ‘females’ vs. ‘males’
and ‘Scandinavians’ versus ‘Other nationality’:
The alternative hypothesis for the summed items from Q2, when females are compared with males
can be defined as:
H24: μ0, SUMQ2 ≠ μ1, SUMQ2 0 represents females and 1 represents males. The last letter corresponds
again to the variable, which here is the sum of the variables from Q2.
Table 8: Description of significant results p-value < sig. Mean (M)
SUMQ2 t=4,152 Females: μ =40,86
0,000 < 1 % Males: μ =37,6
Items Q2: variable 17: ‘I use the backside of used prints t=5,273 Females: μ =3,29
for taking notes’
0,000 < 1 % Males: μ =2,37
Items Q2: variable ‘SMEAN(V18_If-I_am_offered)’ t=3,08 Females: μ =3,26
0,002 < 1 % Males: μ =2,69
Items Q2: variable ‘Reversed_V20_I_leave_my’ t=2,162 Females: μ =3,19
0,032 < 5 % Males: μ =2,73
Items Q2: variable ‘SMEAN(V22_As_the__last_person)’ t=1,836 Females: μ =4,5
0,068 < 10 % Males: μ =4,28
Items Q2: variable ‘Reversed_24_Ieatmeatless’ t=-4,801 Females: μ =3,1
0,000 < 1 % Males: μ =3,9
Items Q2: variable ‘SMEAN(V26_I_actively_support)’ t=2,986 Females: μ =2,6
0,003 < 1 % Males: μ =2,1
Items Q2: variable 28: ‘Environmental or social t=1,833 Females: μ =3,06
credentials of companies influence my buying decisions’
0,069 < 10 % Males: μ =2,78
Items Q2: variable 29: ‘I buy certified organic products’ t=3,267 Females: μ =2,93
0,001 < 1 % Males: μ =2,39
The original table can be found in appendix 18, table 1 and the table for the mean values is placed
in appendix 19, table 2. In the table above we see that there is a significant difference on the
variable ‘SUM_Q2’ when females are compared with males. Females scored higher on the sum of
the behavioural questions, μ = 40,8 for females and μ = 37,6 for males. So overall on Q2, females
obtain a higher score for pro-environmental behaviours compared with males.
No significant differences were found (t=0,9, p=0,360>0,05) when Scandinavians were compared
with those belonging to ‘other nationality’ (appendix 19, table 3). ‘Other nationality’ shows μ =40
and Scandinavians μ =39,26 (appendix 19, table 4).
Checking for differences between the means for the individual items from Q2, when females are
compared with males, result in independent-samples t-tests for variables 17 – 29.
In appendix 20, table 1 we see the original table for the results of the t-test. The original SPSS table
for the mean values can be seen in appendix 20, table 2.
Looking at the table above, we see that significant differences are found in variable 17: ‘I use the
backside of used prints for taking notes’, even when using a significance level of 1 %. Here females
scored higher than males, with μ =3,29 versus μ =2,37 for males. Furthermore, females scored
higher on variable SMEAN18: ‘If I am offered a plastic bag in a store I say I do not need one’
where μ for females= 3,26 and μ for males=2,69 which also is significant using a significance level
of 1 %. Females also scored higher on the variable reversed_V20: I (do not) leave my electrically
powered appliances on stand-by’ where μ for females=3,19 and μ for males=2,73, this time by
using a significance level of 5 %. The variable has been reversed; in this way the variable gives an
indication that females are more inclined to switch off their electrically powered appliances
completely. On average mean values for females correspond to somewhere between ‘Now and then’
and ‘often’ taking the Likert scale into consideration. Males have on average answered that they
switch off the appliances ‘Seldom’ – ‘Now and then’. Females again scored highest on variable
SMEANV22_As_the_last_person with μ for females=4,5 versus μ for males=4,28, with a
significance level of 10 %. The variable reversed_24__Ieatmeatless (because it has been reversed; I
do not eat meatless warm dishes, men scored higher than females, as μ for females=3,1 and μ for
males=3,9. The variable has been reversed because the new variable now serves as an indication of
the level of energy used to warm up meatless dishes. If respondents are agreeing that they to some
degree do not eat warm dishes, we also expect that they are concerned with environmental-friendly
In SMEAN(V26_I_actively_support), females show a higher score than males, as μ for females=2,6
and μ for males=2,1 using a significance level of 1 %. However, these results correspond to a
relatively low mean value. Both females and males in the sample seldom support environmental
causes on average. Variable 28: ‘Environmental or social credentials of companies influence my
buying decisions’ shows that μ for females=3,06 and μ for males=2,78, where the result is
significant with a level of 10 %.. This means that environmental companies are more successful in
influencing the females’ buying behaviours than males buying behaviours. Finally variable 29: ‘I
buy certified organic products’ shows a significant difference, where μ for females=2,93 and μ for
males=2,39 with a significance level of 1 %. Although the mean difference is low for both groups,
we see that females are more inclined to buy organic products than males. All in all, females are
more environmental aware in their behaviour than men.
No differences were found for SUMQ2 when Scandinavians were compared with ‘Other
nationality’. However, there is still a possibility that significant differences can be found for single
Table 9: Description of p-value < sig. level Mean (M)
Items Q2: variable 17: ‘I use t=3,232 Other nationality: μ =3,32
the backside of used prints for
taking notes’ 0,001 < 1 % Scandinavian: μ =2,70
Items Q2: Variable 23: ‘I buy t=1,984 Other nationality: μ =2,85
paper (e.g. as printing paper or
writing pad) made by recycled 0,049 < 5 %
paper Scandinavian: μ =2,51
Items Q2: t= -2,826 Other nationality: μ =3,679
0,005 < 1 % Scandinavian: μ =4,157
Variables 17 – 29 from Q2 are analyzed in the same way as above. In appendix 21, table 1 we see
the original SPSS results for the independent-samples t-test, and the original table for the mean
values can be seen in appendix 21, table 2.
A significant difference was found for variable 17: ’I use the backside of used prints for taking
notes’ even with a significance level of 1 %. Here the groups ‘other nationality’ scored the highest
with μ =3,32, whereas Scandinavians μ =2,70.
Also variable 23: ‘I buy paper (e.g. as printing paper or writing pad) made from recycled paper’
showed significant differences, this time with a significance level of 5 %. Foreigners scored a bit
higher than the Scandinavians with μ=2,85 versus μ=2,51 for Scandinavians.
The last variable that was significant is variable Reversed_27_I_use_a_car (since it has been
reversed: ‘I do not use a car for going short distances’). The variable is significant even with a
significance level of 1 %, with μ=3,679 for foreigners and μ=4,157 for Scandinavians. This means
that Scandinavians are not driving for short distances in a car in the same degree as people from
‘other nationality’ do.
However, we have to keep in mind that the differences for all variables are minor.
Before any conclusions can be made of the results, the responsibility questions Q3 are to be
analysed. We start looking for significant differences for SUMQ3; first for foreigners compared
with Scandinavians and secondly the same test is made comparing females with males.
3.3.8. Independent-samples t-test for SUMQ3, Q3 items, groups: ‘females’ vs. ‘males’ and
‘Other nationality’ vs. ‘Scandinavians’:
The variables to be tested are variables 30 – 36, starting with variable 30: ‘I am jointly responsible
for the exhaustion of non-renewable energy sources’. The original tables can be seen in appendix 22,
Table 10: Description of p-value < sig. level Mean (M)
SUMQ3 t=3,394 Females: μ =23,89
0,001 < 1 % Males: μ =21,5
Items Q3: variable 33: ‘I feel t=2,56 Females: μ =3,93
morally obliged to save as
much energy as possible’ 0,01 < 5 % Males: μ =3,52
Items Q3: variable 34: ‘I feel t=3,86 Females: μ =3,61
guilty when I waste energy’
0,000 < 1 % Males: μ =2,84
Items Q3: variable 35: ‘I talk to t=1,745 Females: μ =3,15
others about saving energy’
0,083 < 10 % Males: μ =2,81
Items Q3: variable 36: ‘I would t=2,96 Females: μ =3,44
be a better person if I saved
energy’ 0,004 < 1 % Males: μ =2,82
The test for SUMQ3, where females are compared with males, shows that a significant difference
exists. We are very certain of this result as it is significant on the 1% significance level. Females
again scored higher with μ=23,89 and μ for males=21,5, which means that they overall have a
greater feeling of responsibility in respect to the environment.
No significant difference was found testing SUMQ3 (see appendix 22, table 3) with the groups
being ‘other nationality’ and ‘Scandinavians’ (t=,530, p=0,597>0,05). The summed μ was 23,17 for
‘other nationality’ and 22,726 for ‘Scandinavians (see appendix 22, table 4).
Now let us see if any significant differences in respect to the items from Q3 are found when females
are compared with males. Females have shown to be more environmental aware than the males in
mind of Q2. To find out if this is the case for Q3, let us have a look at the table above (see original
independent-samples t-test in appendix 23, table 1).
Variables 33-36 were all significant. Variable 33: ‘I feel morally obliged to save as much energy as
possible’ was significant using a significance level of 5 % and μ=3,93 for females and μ=3,52 for
males (see original SPSS table in appendix 23, table 2). This means that more females feel morally
obliged to save as much energy as possible.
For variable 34: ‘I feel guilty when I waste energy’ females also scored higher with μ=3,61 whereas
μ=2,84 for males.
The same is true for variable 35: ‘I talk to others about saving energy’, which is significant when a
significance level of 10 % is used. Females scored higher with μ=3,15 versus μ=2,81 for males.
Lastly, variable 36: ‘I would be a better person if I saved energy’ was significant even on a
significance level of 1 % with μ=3,44 for females and μ=2,82 for males.
Now, we will have a look at the items in order to see on which variables the group ‘other
nationality’ differs from Scandinavians. In appendix 24, table 1 we see the results show no
significant differences for any of the results. The respondents don’t show that much of a sense of
responsibility for the environment on the items – they are ranging around 2 – 4 (see appendix 24,
table 2). This means that they mildly disagree on the questions asked and that they are unsure of
what to answer. For example respondents are somewhat agreeing that they feel morally obliged to
save as much energy as possible (μ=3,74 for ‘other nationality’/ μ=3,77 for Scandinavians) but they
express uncertainty when it comes to feeling guilty when they waste energy (μ=3,28 for ‘Other
nationality’)/ μ=3,29 for ‘Scandinavians’). Also the respondents are unsure when it comes to
speaking with others about saving energy, so it does not seem that the environment is a topic, they
think so much about (μ=2,89 for ‘Other nationality’/ μ=3,06 for ‘Scandinavians).
Chapter 4.1; The experiment, summary of results:
The goal of the results section was firstly to show if the attitudes of the high involvement group
were enhanced when argument strength was manipulated. This analysis was carried out with the
profile analysis, which was applied to show if any differences existed comparing the two profiles
(low versus high involvement). The Elaboration Likelihood Model proposed that high involvement
groups’ attitudes were increased for ads with strong arguments, if the products advertised contained
pro-environmental statements with which the group could agree. As the personal relevance
increases, the attitudes and thoughts of the subject show greater discrimination of strong from weak
messages. More specifically, when the message is strong, increasing relevance produces a
significantly increase in attitudes, but when the message is weak, increasing relevance produces a
significant decrease in attitudes. The profile analysis showed that there was greater discrimination
between respondents from the two groups’ attitudes when strong arguments were used. The
discrimination was smaller in considerations of ads with weak arguments, which is confirming that
respondents with the highest environmental awareness according to the NEP also have the highest
attitudes when argument quality is enhanced.
The ELM theory was only partly confirmed by the Tesla Roadster ad, as only two variables were
significant when low involvement and high involvement groups were compared, these were
variable 48: ‘How persuasive do you think the advertisement is’ for the Tesla Roadster ad with
strong arguments and variable 50: ‘How appealing do you think the advertisement is’ for Tesla
Roadster with strong arguments. Because the high involvement group did not differ significantly on
the other variables with strong arguments in respect to the Tesla Roadster ad, it is difficult to
conclude that the effects are caused by the level of involvement.
Evidence for the ELM became much clearer after analysing the Apple ad’s with strong and weak
arguments, as all ads with strong arguments caused the high involvement group’s attitudes to
increase relative to the attitudes of low involvement group.
Variable 1: ‘Climate change will negatively impact me during my life’ is corresponding to the
category ‘awareness of consequences’ from the VBN theory. Here no significant differences were
found comparing females with males and Scandinavians with respondents from other nationalities.
Both groups were unsure of what to answer on this particular variable.
Still it seems that both males and females acknowledge that eco-friendly behaviour is important,
and that climate change will impact humanity in some way or the other.
The next objective of the thesis was to analyze respondents according to the behavioural items and
the variables relating to environmental responsibility. Two pairs of groups were compared with
each other, these were: ‘females’ versus ‘males’ and ‘Scandinavians’ versus ‘other nationalities’.
Females were expected to score highest on the environmental items from Q2 and Q3 in the
questionnaire. As written in the theory section of this paper, an experiment has previously been
made by Victoria K. Wells, which found that when respondents were asked questions related to
shopping, travel, reuse and energy-consumption behaviours, females were the most environmentally
aware. This was also the case in the experiment for this thesis; females firstly scored highest on
SUMQ3, which means that they overall feel more responsible for the environment than males.
Females feel guiltier when wasting energy than males (variable 34). Furthermore, they talk more to
others about saving energy than males (variable 35). Finally, more females are agreeing that they
would be better persons if they saved energy. The results should be taken with caution, because the
significant differences between the groups are minor. Still the differences show that females are
thinking more about the environment than males.
The VBN theory suggests a causal order where ascription of responsibilities leads to pro-
environmental behaviour. Results from Q2 in the questionnaire showed that females also scored
highest on pro-environmental behaviour, both overall and on the items. This became apparent as
more females are agreeing that they do not need a plastic bag while shopping (variable 18), females
do not leave their powered appliances on stand-by as much as males (variable 20), more females are
inclined to switch off the light before leaving a room (variable 22). Also females scored higher than
men in relation to supporting environmental causes (variable 26), and they are to a larger extent
than males influenced by environmental credentials of companies in their buying behaviour
(variable 28). Lastly, females also scored the highest value in respect to buying organic products.
The only variable where males obtained the highest score was on variable reversed 24: ‘I do not eat
meatless warm dishes’. This means that males, on average, do not use energy to warm food that
does not contain meat at the same scale as females from the sample do it. Again we have to keep in
mind that the mean values are not that different from each other (but still different though) and
therefore should the discrimination of females from males be taken with the minor mean differences
Although the goal has not been to analyze if a causal order exists between ascription of
responsibility and pro-environmental behaviour, as this already has been dealt with in previous
research, and because the VBN theory has only been dragged into the report to show where the
variables from the questionnaire stem from, we see that respondents are following the same sort of
pattern as proposed by the VBN theory. Because females ascribe more responsibility for the
environment, they are also engaging in more pro-environmental behaviour than men.
Variable 1, Q2 and Q3 were again tested comparing ‘Other nationality’ with Scandinavians. This
time it was expected that Scandinavians would show both more responsibility for the environment
and more pro-environmental behaviour than people from other nationalities. The reason is
Scandinavians often manifest themselves as leaders for environmental causes. This was apparently
not the case for this experiment. Variable 1 showed no significant difference, and both groups were
unsure/mildly agree, meaning that respondents to some extent agree that climate change could have
a negatively impact on them during their lifetime. Respondents who answered that they were from
other nationalities scored highest on variables like variable 17: ‘I use the backside of used prints for
taking notes’ and variable 23: ‘I buy paper made by recycled paper’. However, both Scandinavians
and people from other nationalities only seldom buy recycled paper. Scandinavians got the highest
score on variable reversed 27, indicating that they do not use a car as much as respondents from
other nationalities to drive small distances.
No significant differences were found on items from Q3. This means that there are greater
variations in the results when respondents are grouped by sex than by nationality.
Chapter 4.2; Conclusion:
The thesis has mainly been focusing on the electric car market, as cars today are contributing
significantly to global warming. Apple has been characterized as well to show how firms can
minimize the CO2 footprint in the production, packaging and transportation processes.
In the beginning of this paper, the importance of environmental communication between all actors
in the market was discussed in order to show the numerous players who are involved. These players
have to work together before consumption behaviour is getting more sustainable. In consideration
of the electric car market, when we look ahead the next five to ten years, there are some challenges.
For example there has to be a shift from traditional powertrains to more electric powertrains. With a
well-trained and motivated team, like the IA-HEV combined with the strong awareness of the
likelihood for unplanned events and cooperation with the car industry, success can be obtained.
Consumers have to be informed how to shift away from the internal combustion engine and towards
hybrid and electric vehicles. If 5 % of the cars sold annually from 2012 on were EVs, the growth in
the EV share of the market for new vehicles will still be so slow that even in 2020, 95 % of all
vehicles on the road will still not be electric. The challenge of the media and politicians will be to
communicate this fact to the consumer while the IA-HEV will have to moderate this change process
(IEA hybrid and electric vehicles annual report, March 2010).
The importance of collaboration among actors like politicians, automobile companies and
consumers was discussed in a documentary film called ‘Who killed the electric car’. The conclusion
was that both oil companies, automobile companies, the consumers and the government can be seen
as guilty in killing the electric car. Governments are trying to make incentives, but still they are
influenced by major oil companies who feel threatened by the introduction of hybrid and electric
vehicles. Oil companies have much power. They can easily create demand by dropping the oil price.
The problem is that this only leaves the junkie (the consumer) imprisoned in the addiction to oil.
Former president Jimmy Carder said that oil imports should never exceed the 1977 level of 8.8
million barrels per day. This resulted in oil companies making price drops, which made the US
import 13,5 million barrels per day in 2005.
Consumers are not innocent either, because they have to change their habits to being more
Lastly, the automobile manufacturers have to look at the long-term instead of the short-term.
Automobile manufacturers are not willing to produce as long as they cannot see any profits in the
market. The companies also have to be more effective in the marketing efforts to educate consumers
about the benefits by an electric car. Past marketing efforts have not been good enough, as the focus
was on the limitations instead of the advantages (‘Who killed the electric car’). As the thesis has
shown, the attitudes of the consumer can be enhanced if the marketing strategy is effective.
Marketing efforts should be done by using as many marketing instruments as possible so a wider
audience can be reached.
The paper has also shown that consumers can be affected by the level of arguments strength. For a
beginning the companies should make ads that contain strong arguments to those who initially will
buy the product. By focusing on high involvement groups a market will be created for green
products. This market can very well grow, meaning that more consumers will be attracted. In this
way can our consumption behaviour in the future be changed to be more environmental friendly.
As the experiment in the thesis also showed, women were more environmental aware than men.
This information can be used to take the target groups of green products into consideration and to
design ad campaigns that are targeting women. Future research should seek to investigate why
females are more environmentally aware. Future research should also analyze the marketing of
green products even further, by investigating how consumers grouped by nationality, sex and
involvement are affected by green advertising. If useable results are found, which companies can
apply, there will be opportunities to positively influence the market for green products.
We have to be optimistic because both economic and technological forces can give the opportunity
to ensure a greener future. Still, the struggle is tough. Major ‘Goliath’s’ like the oil companies have
to be fought with. But with enough ‘Davids’ it may be possible. Furthermore, the economic crises
can help us to follow the right path: JC lastly explains in the interview (see transcript line 420, p. 16)
that the financial crisis is both negative and positive. It is positive because the price of gasoline has
increased, forcing consumers to think of their consumption.
Further he acknowledges that other technologies like the fuel cell technology also can be possible as
substation for the gasoline motor. Still the electric car has more advantages as it is less expensive
and because battery innovations are constantly developed.
Looking at the automobiles possibilities to minimize the impact on the environment starts by an
examination of the figures starting from appendix 25, figure 1. This figure shows a roadmap study
which was based on two scenarios: the baseline with CO2 emissions continuing to grow along
current trends, and the more ambitious BLUE Map scenario. According to the BLUE Map scenario,
global CO2 emissions could be reduced by 50 % from 2005 levels by the year 2050. In the figure,
we see the respective shares of the reduction from four different energy users: power production,
industry, buildings, and transport (IEA hybrid and electric vehicles annual report, March 2010).
In appendix 25, figure 2 we see that in the BLUE Map Scenario, transport energy consumption can
decrease almost to the level of 2005, including a more than 50 % share of low CO 2 fuels. The
baseline scenario shows an 80 % increase of fuel usage for transportation only by the year 2050.
This mainly depends on car sales and freight, which is sensitive to economic growth. Besides the
increasing fuel economy, it seems key to improve vehicle electrification where factors like the price
dropping in costs of the batteries, commercial production of EVs, plug-in hybrids as the transition
from the compulsion engine to the electric vehicle as a transition strategy all are factors which can
make us decrease the level of green-gas emissions. Additional reductions can be achieved from
changes in the nature of travel (IEA hybrid and electric vehicles annual report, March 2010).
In appendix 26, figure 1, we see that advanced technologies must play a major role in minimizing
the emissions. The figure shows the BLUE Map scenario for passenger light-duty vehicles. We see
that there is a strong growth of passenger vehicle sales per year, and which almost is tripled by the
year 2050 compared to 2005. The growth will happen in developing countries like India, China,
Brazil, and Indonesia, and this may very well result in shortages in raw materials (IEA hybrid and
electric vehicles annual report, March 2010). They key point in this figure is that we can see rapid
light-duty vehicle technology evolution over time.
In appendix 26, figure 2 we see the Blue Map scenario for the EV and PHEV (plug-in hybrids) sales
trajectory to 2050. The two propulsion technologies of the future must increase to 100 million units
per year in 2050. The goal is very ambitious, because it requires that EV and PHEV sales reach a
significant level by 2015 and rise rapidly thereafter. The magnitude is huge as it can be compared
with today’s situation, where EVs and PHEVs still have to break out of niche markets into the
So there is no doubt that there are a lot of challenges, both in the electric car market and for
electronic providers like Apple. It is important that companies try to take sustainability into the core
of their business processes instead of using ‘green’ as a marketing ploy.