SlideShare ist ein Scribd-Unternehmen logo
1 von 60
Downloaden Sie, um offline zu lesen
[1]
The Effect of Transparency on
Information Need, Purchase Intent and
Customer Loyalty, moderated by Involvement
and Risk Aversion
Author: Patrick Heeremans
Student number: 10660623
Supervisor (1): Joris Demmers
Supervisor (2): Dr. Meg Lee
Qualification: MSc. Business Studies – Marketing Track
Institution: Amsterdam Business School – UVA
Document: Master Thesis
Version: Final version
Date of submission: 19/08/2014
[2]
Table of contents
Abstract ...................................................................................................................................... 3
Introduction ................................................................................................................................ 4
Literature Review....................................................................................................................... 8
Data and Method ...................................................................................................................... 19
Results ...................................................................................................................................... 29
Discussion ................................................................................................................................ 37
Conclusions .............................................................................................................................. 43
Limitations ............................................................................................................................... 45
References ................................................................................................................................ 46
Appendices............................................................................................................................... 52
A. The actual experiment (treatment group) ........................................................................ 53
B. Risk aversion Scale.......................................................................................................... 54
C. Information needs scale................................................................................................... 55
D. Revision letter 1............................................................................................................... 56
E. Revision letter 2............................................................................................................... 58
[3]
Abstract
The growing interest for transparency comes from the accountancy sector, and is fed by
different business scandals (Cohn & Wolfe, 2012). Next to this, consumers get a growing
interest in being able to verify good business, in which transparency plays a central role
(Bhaduri & Ha-Brookshire, 2011). Despite growing interest transparency is a concept in
marketing science, which has not been researched much. This leaves us with many question
marks and undefined opportunities for business practice. Marketing science knows many
conflicting views on transparency. For example, Hultman & Axelsson (2007) and Koslow
(2000) think transparency can be a bad practice resulting in problems. Meanwhile, most other
researchers in this field argue for the opposite.
In order to find answers a survey experiment was conducted and to analyze the
different responds different statistical methods were used.
In this research an effect of transparency on the dependent variables information need,
purchase intent and customer loyalty could not be confirmed. However, let us not forget no
negative effects of transparency were found as well. Based on this, business practice should
not be anxious to become more transparent, because this is what consumers want (Bhaduri &
Ha-Brookshire, 2011). On top of this, the current developments, like the digitalization, almost
force business practice to disclose product information (Cohn & Wolfe, 2012; Cohn & Wolfe,
2013). Furthermore, when balancing the scale research arguing for more transparency
outbalances research being skeptical about transparency. So, nothing should hold businesses
from becoming more transparent about their products.
Finally, no significant moderating effects of risk aversion or involvement were found.
[4]
Introduction
Recently, we have encountered a real life example of transparency in combination with
information need, purchase intent, customer loyalty, involvement and risk aversion, which
inspired this research: Chinese manufacturers of baby milk used the dangerous ingredient
melamine in their products, resulting in a business scandal: The usage of this ingredient
coming to the front was resulting in scattered trust amongst Chinese consumers. These
Chinese parents try to avoid possible risks connected to Chinese baby milk and started to
purchase foreign baby milk, because Western firms are more transparent about their products
and the ingredients they use (Trouw, 2013). On top of this, Chinese consumers are even
willing to pay significantly more for milk if information about the product is provided, thus
showing the important role transparency can play for both consumers and companies (Zhang
et al., 2012). However, hardly any research on Western customers, and their purchase intent,
customer loyalty and information need related to disclosure is conducted while this could be
of great importance to business practice. In this real life example six key constructs are
playing a central role: transparency, information need, purchase intent, customer loyalty, risk
aversion and involvement. In this case the fact that melamine was used in producing baby
milk and this practice coming to the front resulted in risk aversion, moderating information
need, purchase intent and customer loyalty, because the Chinese consumers do not trust
national manufacturers anymore. These Chinese consumers are most likely to be parents, thus
determining their involvement. Shortage of information, lowered purchase intent and lowered
customer loyalty for Chinese baby milk made Chinese consumers seek their salvation in
Western products.
Transparency causes an impact on information need, purchase intent and customer
loyalty and can therefore be seen as the independent variable of the relationships in this real
[5]
life example. Different researchers also consider transparency to be an independent variable in
the relationships with the dependent variables information need (Shroff et al., 2005; Xiang, J.
Zhou, X. Zhou, & Ye, 2012), purchase intent (Koslow, 2000; Bhaduri & Ha-Brookshire,
2011) and customer loyalty (Clemons, 2008; Cohn & Wolfe, 2012; Cohn & Wolfe, 2013).
However, this research is still limited in numbers and not conducted on Dutch consumers,
whom are an unique audience (Hofstede, 1983), leading up to the first contribution of this
thesis. Subsequently, no research has dealt with all three dependent variables of the real life
example in one study despite these three variables playing an important role in the Baby-milk
drama, at the same time. However, some researchers do mention some linkages between the
dependent variables. For example, information need is argued to influence purchase intent
(Bhaduri & Ha-Brookshire, 2011); purchase intent is the start of behavioral loyalty (Backman
& Crompton, 1991); and information need could influence customer loyalty (Bhaduri & Ha-
Brookshire, 2011). This research will bring a second contribution to knowledge by
researching if correlations between the dependent variables exist, because based on current
literature strong conclusions cannot be drawn yet.
Furthermore, the moderating roles of involvement and risk aversion have not been
researched much. Actually, only Carey et al. (2008) and Mittal (1988) have conducted
thorough research on the moderating role of involvement in relation to transparency. Despite,
involvement seemingly playing an important role in the Chinese baby milk example no break-
through research on involvement as moderator is to be found. Carey et al. (2008) share this
view and argue for more research on this construct: Parents have started to think totally
different about the world and ethics (Carey et al., 2008), and transparency could play an
important role in this changed perception (Bhaduri & Ha-Brookshire, 2011). Furthermore,
involvement is important to research, because parents can be considered to be a completely
different audience than other consumers (Carey et al., 2008). Hereby, this research can bring a
[6]
third contribution to knowledge about involvement and its relationship with transparency. As
fourth contribution, this research provides knowledge about involvement to help business
practice to find out how to manage transparency for two complete customer groups; parents
and non-parents (Carey et al., 2008).
Risk aversion on the other hand has been researched much in the past, but not in
combination with the independent variable transparency. Especially, the moderating role of
risk aversion in consumer settings has not been researched much, which is a potentially
important gap because of the real life illustration of risk aversion as a moderator in the
Chinese baby milk gate. Chu & Chintagunta (2011) and Pardo (2013) studied the impact of
risk aversion as a moderator for the relationship of transparency on information need; Shimp
& Bearden (1982), Clemons (2008) and Kim et al. (2008) studied the moderating impact of
risk aversion on the relationship of transparency on purchase intent, and only one research
was conducted on the moderating impact of risk aversion on customer loyalty by Clemons et
al. (2008). By using risk aversion as moderating variable, as in the Chinese baby milk
example, this research addresses a gap in marketing literature and results in a fifth
contribution.
Customer loyalty (retention), possibly influenced by transparency, is the most
important factor of the Customer Lifetime Value formula and therefore profit (Rust et al.,
2004), which underlines the value of this research for business practice. This resulted in the
sixth major contribution of this research.
In order to make all contributions this research used an experimental survey design to
create purchase simulations in which respondents were confronted with highly transparent
products and low transparent products. The respondents were divided into parents and non-
parents in order measure the impact of involvement. Risk aversion, not yet used as moderator
in current Marketing literature, will be measured by creating Likert scales.
[7]
Respondents’ choices as a result of the degree of product transparency will show the impact
of transparency on information need, purchase intent and customer loyalty, moderated by risk
aversion and involvement.
The goal of this research is to find out how the independent variable transparency
influences de dependent variables information need, purchase intent and customer loyalty.
Furthermore, this research strives to confirm a moderating role of risk aversion and
involvement. By melting all these constructs together we can see that this research addresses
an interesting relationship model and an important gap in Marketing science, and could help
business practice increase consumers’ purchase intents, lower information needs and even
more attractive; making customers return.
[8]
Literature Review
The overarching topic for this master thesis will be ‘Transparency in Marketing’.
Transparency, in its simplest form, means sharing information and is an element of buyer-
seller relationships (Hultman & Axelsson, 2007). Subsequently, transparency has to be ‘free
from delusion’ (Cohn & Wolfe, 2012) and is increasingly becoming more important. This
importance comes from the accountancy sector and is fed by different business scandals of
which the banking crisis is probably the best known example (Cohn & Wolfe, 2012). On top
of this, the digitalization, and increasingly improving communication channels and
information spreading really puts pressure on companies to become more transparent, because
consumers are increasingly becoming better able at finding information themselves and have
a growing awareness (Cohn & Wolfe, 2012). Cohn & Wolfe (2012) even call transparency a
new form of currency, which is information based and builds trust with customers and other
stakeholders. This is supported by Zhang et al. (2012) who claim transparency increases
consumers’ willingness-to-pay for at least the Chinese market. Despite, transparency comes
from the accounting sector this does not mean consumers experience financial transparency,
like annual reporting, as being very important. Product transparency, essentially openness of
a brand about the products it offers, is considered to be far more important to consumers
(Cohn & Wolfe, 2012). This thesis used this product transparency to operate the broad
concept of transparency, because of its importance to consumers. Cohn & Wolfe (2012) argue
that the ongoing economic unrest and instability facilitates the growing need for transparency.
Bhaduri & Ha-Brookshire (2011) add that the growing concern for how businesses perform,
fed by environmentalists and other nongovernment organizations spreading awareness about
the current abuses of planet earth, increases consumers’ demands for transparent brands.
[9]
Slavin (2009) adds today’s consumers are conscious about their society and environment,
demanding transparent and sustainable products: The market for such products is expected to
grow by as much as 19% by 2014.
Transparency knows different stages: A buyer-seller relationship can be based on
complete transparency, a middle path of ‘translucency’, information can be partially shared
and a situation is possible in which no information is shared at all (Lamming et al., 2001).
This master thesis used two extremes: a high degree of product transparency and a low degree
product transparency. These extremes were chosen to be able to draw strong conclusions on
consumer choices by leaving out the middle path Lamming et al. (2001) write about, because
two extremes force consumers to really favor product transparency or not.
Cohn & Wolfe (2012) consider 9 types of transparency being important as indicators
of a transparent organization and have ranked them as followed:
Nr. 1: Information about where an organization gets its materials/ingredients from.
Nr. 2: Source of funding and ownership of the business.
Nr. 3: Information about profits and losses.
Nr. 4: Transparency regarding the suppliers of the company.
Nr. 5: Transparency regarding the partners of the company.
Nr. 6: Information about the payment of senior management.
Nr. 7: Information about the payment of all layers of employees.
Nr. 8: Transparency regarding community activities and charities the company is involved in.
Nr. 9: Information regarding political alignment.
[10]
However, this long list mentioned by Cohn & Wolfe (2012) is less operational compared to
the four types of transparency Hultman & Axelsson (2007) distinguish. This research is based
on a Business to Business (B-to-B) setting and names the following four types of
transparency: technological transparency, supply (chain) transparency, cost/price transparency
and organizational transparency. These four types will be modified to a Business to Consumer
(B-to-C) environment into different types of product transparency. The reason to chose for the
Hultman & Axelsson (2007) types of transparency over the Cohn & Wolfe (2012) list of
transparency types is that these leave out the types that are of lesser importance to consumers.
Most previous research considers transparency to be a good practice for both
businesses and consumers. An example of this is the study of Carter & Rogers (2008) which
suggests that businesses need to be transparent with their operations to maintain business
legitimacy and to build reputation. Only few researchers consider negative outcomes:
Hultman & Axelsson (2007) and Koslow (2000) find proof for transparency resulting in
problems. Hultman & Axelsson think Transparency in a B-to-B setting can result in a power
shift and increases uncertainty. On top of this, transparency demands a high level of trust,
because otherwise the relationship between buyer and seller will become less trustworthy and
more opportunistic (Hultman & Axelsson, 2007). All this makes transparency a fragile
concept not easily to reap the benefits from. Koslow (2000) argues consumers get more
skeptical when confronted with ‘honest’ transparency in the form of advertising. This
skepticism flows forth from brands sometimes lying and providing sales arguments lacking
credibility, but even more important skepticism is a mechanism that protects consumers from
brands that provide information to sale rather than inform. In this sense, skepticism can be
considered to be a defense mechanism. This mechanism can be considered to be a problem,
because skepticism limits sales.
[11]
Information need
This negative impact of limited sales can be nullified when brands provide verifiable
information (Ford, Smith, & Swasy, 1990). Loewenstein et al. (2013) and Rijswijk & Frewer
(2012) even go a step further in stating that providing traceable (product) information can
even lower information need. On the other hand, consumers also tend to become more and
more skeptic towards information provided, even though this information is surprisingly
honest (Koslow, 2000). Traceability is not only important from a consumer perspective,
because it starts with brands being able to trace all the ingredients of their products
themselves. On top of this, developing tracing mechanisms is one of the most important tools
in order to regain consumers’ confidence (Golan et al., 2004).
Information need is the first dependent variable of this research. Information need can
be lowered, if consumers´ information asymmetry opposed to brands is narrowed (Xiang, J.
Zhou, X. Zhou, & Ye, 2012). Information asymmetry means that either the buyer or the seller
has superior knowledge over the other. In most cases it is the seller who has most information
resulting in a disadvantage, or information need for the buyer (Xiang, J. Zhou, X. Zhou, &
Ye, 2012). Transparency lowers information asymmetry (Shroff et al., 2005), and in that
sense reduces peoples´ information need as well: All this resulted in the first hypothesis.
H1a: Transparency decreases information need.
Purchase intent
Consumers actively search for needed information and will use this information in order to
select products (McNeal, 1987). Transparency is able to provide this needed information and
therefore enhances purchase intent. Koslow (2000) adds advertising, and providing
information in that sense, reduces search costs and enhances consumers welfare. The same
accounts for companies being transparent, thus providing information. Economically spoken
[12]
transparency should increase purchase intent, because it benefits consumers searching for
product information. If this theory is true will be tested with the second hypothesis.
Another reason to test this hypothesis is that for consumers who are concerned about
environment and society, product transparency can be important for their purchase intent.
This is the case, because transparency creates trust as a basis for a good buyer – seller
relationship (Bhaduri & Ha-Brookshire, 2011). Consumers’ perceived value perspectives
towards ‘good business’ and the theory of reasoned action makes transparency key in their
purchase intent as well (Bhaduri & Ha-Brookshire, 2011). The theory of reasoned action
argues people’s intention to perform a certain action, like purchasing a product, is a function
of these people’s subjective norms (Bhaduri & Ha-Brookshire, 2011). In other words, a
person will have a high purchase intent when he or she gets his or her beliefs of good business
confirmed by the transparency of this certain product. A person can confirm his or her beliefs,
because transparency gives the opportunity to evaluate good business practices (Bhaduri &
Ha-Brookshire, 2011).
An important distinction has to be made between positive and negative information
disclosure. Negative information even has a stronger impact on consumer decision making,
meaning consumers will not purchase a product as a result from it (Arndt, 1967; Reynolds &
Darden, 1972). Despite clearly being advocates of transparent business practices Bhaduri &
Ha-Brookshire (2011) also conclude disclosure of negative information will result into a
lowered purchase intent. To exclude this effect this thesis’ survey experiment did not use
(strong) negative information.
Purchasing a product can be considered to be the first step to customer loyalty in terms
of at least behavioral loyalty. Backman & Crompton (1991) call this spurious loyalty.
H2a: Transparency positively influences purchase intent.
[13]
Transparency is ‘love’..?
Besides spurious loyalty customer loyalty knows 3 additional levels; high, low and latent,
which are driven by emotion and behavioral factors (Backman & Crompton, 1991).
Transparency is the third most important driver of customer loyalty, after quality and price,
and is even more important in the decision making process than brand appeal, and
recommendations online –and from friends are (Cohn & Wolfe, 2012). The importance of
transparency has even grown in 2013 (Cohn & Wolfe, 2013). People become more loyal
towards a brand, because of transparency, because people are very skeptic about companies
ever since recent scandals, such as the banking crisis. So, companies that are being transparent
reduce risks for consumers and lower skepticism and distrust, and therefore increase customer
loyalty (Cohn & Wolfe, 2012). On top of this, people will only buy brands they love. This
love essentially is customer loyalty and can come forth from information disclosure, because
transparency is an important basis of resonance marketing (Clemons, 2008). Resonance
marketing means managing and profiting from hyperdifferentiation, by ‘creating’ loyal
customers. Clemons (2008) describes hyperdifferentiation as the ability of a brand to offer
numerous products that fit many different needs and wants. This would not make sense if
people do not know what products a brand offers, what these products are, where these
products come from and how these are produced. In other words, consumers need product
transparency to evaluate –and optimize choices, which will result in brand love –or customer
loyalty (Clemons, 2008). This view on transparency and customer loyalty resulted in the third
hypothesis.
H3a: Transparency positively influences customer loyalty.
[14]
Involvement
After mentioning the three dependent variables; information need, purchase intent and
customer loyalty the moderating variables risk aversion and involvement are being discussed
now. An example of a study in which involvement is conceptualized is Olsen’s study (2003):
This research is speaking of health involvement mediating the consumption (i.e. preferences)
of sea food. In order to conceptualize involvement in this thesis being a parent or not was
chosen, because parents can be considered to be a completely different audience than other
consumers (Carey et al., 2008). Furthermore, reducing risks of purchases for your baby or
child might be even more important to parents than to others. Risk and involvement seem to
be interrelated, and risk has a strong connection to transparency. Namely, transparency is able
to reduce risks (Hultman & Axelsson, 2007).
Involvement facilitates the need for information processing (Mittal, 1988). The higher
a person’s involvement the more important cognitive processing becomes for buying a
product opposed to affective based motives (Mittal, 1988). In other words, high involvement
increases peoples’ information need (Mittal, 1988) and therefore moderates the impact of
transparency on the dependent variable information need, resulting in the following
hypothesis:
H1b: Transparency decreases information need, but this relationship is moderated by
involvement, in the sense that the impact of transparency on information need is stronger
under high involvement conditions.
Ethical issues are very important to consumers willing to buy products, and
transparency plays an important role in verifying ‘good business’ and eventually consumers’
purchase intents (Bhaduri & Ha-Brookshire, 2011). However, this grown interest for ‘good
practices’ cannot be explained by a grown awareness of ethics alone, but is for a large part
[15]
facilitated by the arrival of a new born child (Coughlan, 2002; Foster, 2004; Hickman, 2004;
in Carey et al., 2008). Parenthood as a determinant of involvement makes ethical decision
making even more important (Carey et al., 2008) and in that sense moderates the relationship
of hypothesis 2. Ethical decision making gets facilitated more by involvement, because the
birth of a child ‘awakes’ peoples’ sense of ethics, and changes peoples’ view of the world. In
other words, parents want to contribute to a better world (Carey et al., 2008).
H2b: Transparency positively influences purchase intent, but this relationship is moderated
by involvement, in the sense that the positive impact of transparency on purchase intent is
stronger when involvement is high.
Involvement also plays an important role in customer loyalty, because being able to
evaluate choices is more important when involvement is high (Carey et al., 2008). In this
sense, the importance of transparency for hyperdifferentiation grows, thus making Clemons’
(2008) ‘resonance marketing’ more important under high involvement conditions. Being able
to make ‘good’ consumer choices, and being loyal towards a selection of these choices means
parents can make a trade off and balance their less ethical consumer decisions, which is an
important driver for their consumer choices (Carey et al., 2008). Transparency helps making
such trade offs (Clemons, 2008).
H3b: Transparency positively influences customer loyalty, but this relationship is moderated
by involvement, in the sense that the positive impact of transparency on customer loyalty is
stronger under high involvement conditions.
Risk aversion
Risk, an old –and well known concept in research, is what consumers perceive as a possible
negative outcome of a purchase. In other words, it is the possibility a purchase will not meet
[16]
the expectations or anticipated benefits (Bauer, 1960). The way a company communicates
information about their products or services determines the risk aversion of a consumer
(Shimp & Bearden, 1982).
An important driver of this second concept, (perceived risk) reduction, is the
verifiability of given information, because transparency without the possibility of tightening
the given information is useless (Loewenstein et al., 2013). Different studies have talked
about pictures (of negative impacts of smoking) increasing the awareness of risk (Borland et
al., 2009; Hammond et al., 2006; Thrasher et al., 2011).
Information need has much to do with perceived uncertainty of consumers; and
increased transparency, as a source of information, can lower this uncertainty (Shimp &
Bearden, 1982). The same view is shared by Hultman & Axelsson (2007). Pardo (2013)
researched the moderating role of risk aversion in entrepreneurial investment decisions. His
study found that the independent variable information asymmetry, determined by
transparency, influences entrepreneurial investment decisions and is moderated by risk
aversion. Risk aversion as a moderator has also been researched in consumption settings. For
example, Chu & Chintagunta (2011) claim risk aversion is an important moderator for the
relationship of warranties and information asymmetry. Information asymmetry is determined
by transparency (Shroff et al., 2005), and the higher information asymmetry is the more
important a warranty is, because warranties provide a sense of protection against knowing less
than the seller (Chu & Chintagunta, 2011). Risk aversion moderates this relationship, because
the sense of feeling protected becomes more important when risk aversion is higher. So,
transparency lowers information need and lowers the importance of warranties. However, risk
aversion moderates the impact on this, because a higher risk aversion demands more
transparency to lower information asymmetry and subsequently lower information need, in
order to lower the importance of warranties.
[17]
H1c: Transparency decreases information need, but this effect is moderated by risk aversion,
in the sense that a higher degree of risk aversion demands a higher degree of transparency in
order to decrease information need.
Shimp & Bearden (1982) claim that if consumers are able to assess products, they will
have a lowered risk aversion: This has mostly to do with increased trust according to Kim et
al. (2008). Subsequently, this lowered risk aversion results in an increased purchase intent
(Shimp & Bearden, 1982; Clemons, 2008; Kim et al., 2008). Transparency plays an important
role in offering consumers the ability to assess products (Clemons, 2008). So, risk aversion
has a moderating role on the relationship between transparency and purchase intent, because a
higher degree of risk aversion demands a higher degree of transparency to increase purchase
intent (Kim et al., 2008).
H2c: Transparency positively influences purchase intent, but this relationship is moderated
by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree
of transparency in order to increase purchase intent.
In case of the dependent variable customer loyalty, risk aversion even plays a more important
moderating role opposed to purchase intent: A product towards which a person is loyal to can
be considered to be an optimal choice (Clemons, 2008). Risk aversion strongly influences this
optimal choice, because in case of a high degree of risk aversion, a person will not bond to a
certain product (Clemons, 2008). So, transparency increases customer loyalty, but only when
a person’s risk aversion is low (Clemons, 2008).
H3c: Transparency positively influences customer loyalty, but this relationship is moderated
by risk aversion, in the sense that it has positive influence under low risk aversion conditions
only.
[18]
This research will be conducted in the Netherlands, which has an impact on the
importance of risk aversion for consumers, because risk aversion tends to be different per
country. A study conducted by Weber & Hsee (1998) finds that respondents from China, the
U.S.A., Germany and Poland are different in their risk perceptions towards different financial
options. Another study conducted by Pennings et al. (2002) finds that consumers from
different countries react differently to food crises like the mad cow disease. The study also
finds that Dutch consumers are significantly less risk averse compared to German consumers
and slightly less risk averse compared to North American consumers. On top of this, the study
finds that involvement is negatively related to risk aversion: the more you consume a product
the less risk averse you will be in case of a crisis. In general Dutch people score relatively low
on risk aversion (Hofstede, 1983).
In order to find out if product transparency has negative or positive consequences for
business practice and to conclude if consumers really value this construct; the following
research question was answered: ‘How does transparency impact 1) information need, 2)
purchase intent and 3) customer loyalty; and how do involvement and risk aversion
moderate these effects?’
Scheme 1: Relations of the research question
Transparency
Involvement Risk aversion
Information need
Purchase intent
Customer Loyalty
[19]
Data and Method
Method
The method used for this thesis is the experimental design (in a controlled setting). An
experiment is a “research strategy that involves the definition of a theoretical hypothesis;
selection of samples of individuals from known populations; the allocation of samples to
different experimental conditions; the introduction of planned change on one or more
variables; and measurement on a small number of variables and control of other variables”
(Saunders & Lewis, 2012, pp 114).
A major disadvantage of conducting an experiment is the lack of external validity,
which is the possibility to generalize findings beyond the research setting. However,
experiments tend to be high on internal validity (MCDaniel & Gates, 2012). Internal validity
is important to the research question of this thesis, because this study wants to proof the
impact of product transparency (disclosure of companies about their products) on information
need, purchase intent and customer loyalty, moderated by risk aversion and involvement. A
major benefit of experiments is that different variables can easily be adjusted in this
controlled setting (MCDaniel & Gates, 2012). On top of this, the experimental design is ideal
to test the theoretical hypotheses (Saunders & Lewis, 2012).
The lack of external validity could be overcome by conducting a field experiment
(MCDaniel & Gates, 2012). This could have been an interesting possibility, but it would have
been too difficult to adjust information in a real life setting.
[20]
Respondents
The sample consisted out of students, family, friends and employees of different Albert Heijn
stores. Despite, the fact these people where easily accessible the experiment did not suffer
from common biases related to convenience sampling. This was due to the fact that opposed
to regular convenience sampling the people selected where selected randomly (Saunders &
Lewis, 2012).
Sample size
The experiment was based on a 2 x 2 between subjects design. Based on a general rule this
means 25 participants per condition where needed. By using high (product) transparency, low
(product) transparency, being a parent, and not being a parent as conditions this experiment
needed a minimum of 100 participants.
Independent variable (IV)
The ‘degree of product transparency’, based on the two extremes of transparency of Lamming
et al. (2001), was the independent variable used for this experiment. This experiment will
distinguish 4 groups, of which 2 consist out of a group to which products with low
transparency were shown and a group to which products with a high transparency were
shown. Each respondent was shown 4 products and each of these products contained another
type (manipulation) of transparency. In the low transparency condition hardly any info was
given about the product, opposed to a high degree of disclosure in the high transparency
condition. Each respondent in both the high –and low transparency condition was shown the
same 4 products in order to really measure the impact of transparency on their information
need, purchase intent and customer loyalty. In the meanwhile, other causes than transparency
could be excluded, because every respondent was confronted with the same scenario except
[21]
from the degree of transparency.
The manipulations used were based on three out of the four types of transparency of
Hultman & Axelsson (2007). Each manipulation was separated and assigned to one question
so that the differences in impact could be seen as well. The four types of transparency
Hultman & Axelsson (2007) mention were technological transparency, supply (chain)
transparency, cost/price transparency and organizational transparency, of which 3 were used
in this research. Each of these types of transparency were slightly adapted to fit a Business to
Consumer setting, opposed to the Business to Business setting Hultman & Axelsson (2007)
published about. The fourth type of transparency used was health based, which was especially
important to research in relation to involvement (Olsen, 2003).
Stimuli
The product used in the experiment was ‘Unox Soup in pack’. This is a fairly new product
and was therefore unlikely to have associations that could interfere the results of the
experiment. Having tried a product before might influence peoples’ purchase intent and
customer loyalty beforehand. The choice for this product tries to account for this. Each of the
stimuli was about disclosure of product information by the Unox brand, making these stimuli
a form of product transparency. This type of transparency was mentioned in a study by Cohn
& Wolfe (2012).
[22]
Pictures 1-4: Unox ‘Soup in pack’ pictures used in the experiment (without manipulations).
Health transparency (“Erwtensoup” flavour)
In the health transparency manipulation the salt intake per serving was shown on the package
to the respondents in the high transparency condition. Each serving of 250 ml contains 0,88
grams of salt, which corresponds with 37% of the daily need of salt of an adult.
Technological transparency (“Tomatensoep” flavour)
In the technological transparency manipulation the package of Unox ‘Soup in pack’ was
shown in combination with which technology the product was manufactured. The following
text was communicated to the respondents of the experiment, which was derived from the
Unox website: “Soup in bag and canned is pasteurized and sterilized longer. The time in the
[23]
production of soup in suit is significantly shorter. In addition, the soups are canned and bag
snatched heated, while the soup in a pack is first heated before it is taken away. Due to the
short and different manner of heating, the flavor and the color of the ingredients remains
optimally preserved. The shelf life is the same (unox.nl, 2014).”
Supply (chain) transparency (“Chinese Tomatensoep” flavour)
In the supply chain manipulation in the high transparency condition was communicated were
the product was produced, namely in the Netherlands. Despite, Unox is a Dutch company it
offers Chinese tomato soup, not being produced in China at all or having a reference to the
country in relation to the ingredients used. Knowing this could influence people’s purchase
intent. However, Dutch people could prefer Dutch made products as well.
Cost/price transparency (“Pompoensoep” flavour)
This experiment has used price transparency in the high transparency situation by showing
how many percent of the profits goes to farmers delivering the pumpkins for the Unox ‘Soup
in pack’ product.
Dependent variables (DV)
Information need, purchase intent and customer loyalty were the dependent variables of the
research question, because these were tested on the influence of (product) transparency,
moderated by involvement and risk aversion. In order to measure the information needs of the
respondents an information needs scale consisting out of 4 items was constructed based on the
research of Rijswijk & Frewer (2012), which focusses on information traceability of products.
This research mentions different categories of information traceability of products, of which 3
were incorporated in a self-constructed scale to measure information needs in this research.
The three chosen categories were related to gathering information from authorities, relatives
[24]
and points of sale. The fourth item was a general statement about looking for information. It
was not chosen to use an existing information needs scale, because each of these are based on
a very different type of information need than used in this thesis. This scale was important to
measure what the impact of transparency on information needs is. Increased transparency
might facilitate a higher information need and therefore decrease the purchase intent (Bhaduri
& Ha-Brookshire, 2011). The opposite could work as well. See appendix C. ‘Information
need scale’ for the items used in the experiment.
Purchase intent was measured by using a 5-point Likert scale. Respondents were
confronted with each of the four earlier mentioned stimuli in either a high –or low
transparency condition. From an economist perspective product transparency should increase
purchase intent, because it benefits consumers by reducing search costs for information
(Koslow, 2000). Secondly, Bhaduri & Ha-Brookshire (2011) studied transparency in the
apparel industry and concluded transparency positively influences people’s purchase intent. If
the same accounts for the food industry was researched by this thesis’ survey experiment.
The third and final dependent variable customer loyalty was measured by a 5 – point
Likert scale as well. Abougomaah, Naeim, Schlater, and Gaidis (1987) argue for price as
being very important in terms of purchase intent towards a brand or product, and therefore
brand loyalty as well. Cohn & Wolfe (2012) argue transparency is close to the importance of
price for brand loyalty. Thus, making the relationship between customer loyalty and
transparency relevant according to previous research, but still covering an existing research
gap as well.
Moderators
In order to conceptualize involvement in this thesis being a parent or not was chosen, because
the real life example of the Baby milk drama has clearly illustrated the impact of involvement
[25]
on the different relationships between transparency and the DV’s. Furthermore, different
research supports this view, such as the studies of Mittal (1988) and Carey et al. (2008). In
order to see if parenthood, transparency and the DV’s relate to each other the Pearson
correlations for these variables were analyzed.
Risk aversion was the second moderating variable. It is important to find out how the
respondents rate their risk aversion in order to use this as a moderating variable. The chosen
questions to asses respondents’ risk aversion were based on a 5-point Likert scale and were
derived from the research of Weber et al. (2002) on domain specific risk-attitude scales. The
research of Weber et al. (2002) distinguishes 5 types of risk-attitude domains, which are
ethical, financial, health/safety, recreational and social. The health/safety scale was chosen for
this experiment, because the transparency of the products in this research was mainly
health/safety based. This was the case, because people were confronted, with more or less,
information regarding a consumption/eatable product. See appendix B. ‘Risk aversion scale’
for the items used in the experiment.
Treatments
This experiment distinguished 4 groups, of which 2 consisted out of a group to which
products with low transparency were shown and a group to which products with high
transparency were shown. The third and fourth group are respectively a group consisting out
of parents, and a group consisting out of non-parents. The treatments as such were based on
involvement (being a parent or not) and on product transparency (low –and high
transparency), resulting in 4 treatments. The treatment group was confronted with a high
transparency condition and the control group with a low transparency condition to find out the
differences.
[26]
Tool
The experiment was conducted in a survey format online using the program Qualtrics.
Qualtrics is a program which enables researchers and students to make online surveys. This
program is suited for this research, because it facilitates the opportunity for creating a survey
based experiment.
Procedure
The following procedure was used to conduct the experiment. This procedure resulted in
answering the research questions by testing the mentioned hypotheses.
Various people were contacted by dropping a link to the experiment on Facebook. The
respondents were not selected, but reacted to the link themselves to keep the sample free from
biases related to selecting the sample.
Before participating the experiment each respondent had the opportunity to read about
the research and the goal of the experiment. As being a good practice in conducting an
experiment these respondents were told they were not obliged to cooperate and were ensured
that their responds would be operated in trust.
The experiment was divided in two parts; the ‘Low transparency condition’ and the
‘High transparency condition’: Qualtrics assigned each participant randomly to one of these
conditions. Before each experiment respondents were asked if they are a parent or not,
determining the ‘route’ of the experiment they took. This resulted in the division of a parent
and a non-parent group.
[27]
Data analyses
David Hayes
In order to analyze the data the David Hayes (2014) paper was used primarily. This paper
facilitates many models to choose from in order to make ‘prefab’ regression analyses
accounting for, for example moderators and mediators.
Scheme 2. Overall scheme of the different relations
The scheme as shown above was split in 3 models, because analysis of the dependent
variables showed these DV’s were not correlating. The first model ran the effect of
transparency on consumers’ information need, moderated by involvement and risk aversion.
The second model ran the effect of transparency on purchase intentions, moderated by
involvement and risk aversion. The third model ran the effect of transparency on customer
loyalty, moderated by involvement and risk aversion.
Transparency
(nominal/ IV)
Involvement
(Moderator)
Risk aversion
(Moderator)
Information need
(continuous/ DV)
H1 / Research Question
Purchase intent
(continuous/ DV)
H2 / Research Question
Customer Loyalty
(continuous/ DV)
H3/ Research Question
[28]
Cronbach’s Alpha
Two scales were used in the experiment. The first was the risk aversion scale and the second
one was the scale measuring information need. For each of these scales the Cronbach’s Alpha
was calculated showing if the scales were internally consistent, thus showing if each of the
items used in the scales measured the same concept (Dalen & Leede, 2009).
Correlation calculation
The last type of statistical analyses conducted was calculating the correlations in order to find
out if certain variables correlated strongly, and had to be combined; and in order to analyze
positive and negative correlations between variables possibly contributing to answering the
hypotheses. Finally, if applicable, the Pearson correlations were indicated as being significant
in the correlations matrix, which shows if strong conclusions could be drawn based on the
correlations.
[29]
Results
Data cleaning and missing values
In total 102 people responded to the survey experiment. At first, the missing data values were
replaced by using Hotdeck variables. Hotdeck variables in a sense use donor values of a
similar respondent in the dataset to replace missing data values (Myers, 2011). Hotdeck
variables are typically useable in case of discrete values, rather than continuous ones (Myers,
2011). Fortunately, all data in this survey experiment were of discrete nature. The most
important advantage of Hotdeck variables is that they are both valid (under most conditions)
and easy to use (Myers, 2011). However, before being able to run the HotdeckMacro it should
be checked if the missing data is < 10% (Myers, 2011). After analyzing the ‘missing system’
percentages in the various Frequency tables only one statement out of 25 had a higher missing
data score than 10%. This was for the following statement ‘Q7: I would buy Unox Soup in
bag repeatedly in the future considering my children will consume this product as well’. This
was due to the fact respondents only were confronted with this question in case they are a
parent, which was one of the conditions of the experiment. Because of the high missing data
score, and the fact non-parents could not answer this question in the first place, Q7 was not
used in the Hotdeck treatment.
Recoding counter-indicative items
All scales were analyzed to see if reverse coding was needed. This survey experiment used 2
different scales. None of the items of the information needs scale had to be reverse coded.
However, items 6, 7 and 8 of the risk aversion scale did need reverse coding, because these
[30]
items used the word ‘never’ in their statements. Thus, resulting in an opposite answering of
the question by respondents compared to the other items.
Item 6: Never using sunscreen when you sunbathe.
Item 7: Never wearing a seatbelt.
Item 8: Not having a smoke alarm in or outside of your bedroom.
Computing reliability
As mentioned earlier the experiment used 2 scales. The Cronbach’s Alpha of the information
needs scale was 0,814 and for the risk aversion scale 0.392.
A rule of thumb is that a Cronbach’s Alpha of 0.7 is enough to show if items within a
scale correlate (Dalen & Leede, 2009). Cronbach’s Alpha’s of scales sometimes can be
improved by deleting certain items. However, researchers have to be careful in doing that,
because this means deleting possible valuable information/data as well (Dalen & Leede,
2009).
In case of the information needs scale the Cronbach’s Alpha if deleted can be 0.831 at
most. However, it was argued that in this case the increase was too little to justify deleting
data. Furthermore, the Cronbach’s Alpha was above 0,7 in first place after all.
In case of the risk aversion scale the Cronbach’s Alpha could become 0.511 in case
items 8 ‘Not having a smoke alarm in or outside of your bedroom’ and 9 ‘Regularly riding
your bicycle without a helmet’ were deleted. Deleting these items resulted in a significant and
justifiable (in terms of data loss) increase of the Cronbach’s Alpha of >0.10.
Despite the rule of thumb, which states a Cronbach’s Alpha of 0.7 is needed to justify
the use of a scale, the analysis was continued, because a low Cronbach’s Alpha is not always
a problem (Boyle, 1991).
[31]
Correlation matrix
In total three significant correlations were found. When looking at the Pearson correlations it
could be seen that when a consumer knows how a product is produced this positively impacts
customer loyalty at the 0.05 level with a Pearson correlation of 0.515*, thus making
technological transparency very important. No significant correlations were found for
transparency and one of the moderators or dependent variables.
When looking at the correlations of the dependent variables it could be seen that
almost every correlation was insignificant. The only significant effect could be found for
customer loyalty and purchase intent correlating positively with a Pearson correlation of
0.753** at the 0.01 significance level.
Other than this, the dependent variable purchase intent correlated negatively at the
0.05 level with the moderator involvement (-0.236*).
Looking at the manipulation checks we could see respondents did not know about the
disclosed product information in advance meaning this research yielded significant outcomes.
The final results can be found in a table on the subsequent page:
[32]
[33]
Regression analyses
In total 3 models were run, because the dependent variables were not correlated. Otherwise,
all variables would have been combined into one model. All models were matched with
models shown in the ‘Process Guide’ of Andrew Hayes (2014). The models were based on a
95% confidence interval. By observing the models offered in this Process Guide eventually
model number 2 was chosen, which is an additive moderation showing the conditional effect
of X on Y (Hayes, 2014). This conditional effect consists out of the following formula: Y =
c'1 + b7M + c'4W (Hayes 2014). All variables in the model were tested, which means a F-
distribution was used to test the hypotheses (Dalen & Leede, 2009). The regression analyses
of the different models resulted in various coefficients mentioned in schemes 3, 4 and 5:
These were unstandardized coefficients (Hayes, 2014).
The first model had transparency as IV and purchase intent as DV, with both
involvement and risk aversion as moderators. The R square of model 1 was 0.3365. The
adjusted R square was 0.1132 (<0.3365), which essentially means the model could not be
improved (Dalen & Leede, 2009). The critical value was 2.33. Because of the lower F-value
H0 is accepted, which means the model did not have a significant effect on purchase intent.
The positive unstandardized coefficient of 5.815 as shown in scheme 3 had a p-value of
0.4244. Because of this p-value it could not be confirmed there was a positive direct impact
of transparency on purchase intent. The moderators play a small role, making the model
insignificant (p=0.9445). Next to this, the moderators resulted in a very small increase in the
R squared of 0.0013, showing the minor role of them. On top of this, the impact of the
moderators was insignificant as well with a p-value of 0.9554.
Model 2 had transparency as IV and information need as DV, with involvement and
risk aversion as moderators. This model had an R square of 0.1763. The adjusted R square
[34]
was 0.0311 (<0.1763). The F-value in the output was 0.5066. The critical value was 2.33 and
because of the lower F-value H0 was accepted, which means the model did not have a
significant effect on information need. The moderators played a small role, making the model
insignificant (p=0.9216). Besides, the moderators resulted in a very small increase in the R
squared of 0.0020, showing the minor role of these moderators. No significant direct effect of
transparency on information need could be confirmed, because the (positive) unstandardized
coefficient of 0.2780 had a p-value of 0.9698.
The final model was about the impact of transparency (IV) on customer loyalty (IV),
with involvement and risk aversion as moderators. The model had an R squared of 0.1840.
The adjusted R square was 0.0338. The F-value in the output was 0.5538 and the critical value
was 2.33. Because of the significantly lower F-value H0 was accepted, which means the
model did not have a significant effect on customer loyalty. Again, the moderators played an
insignificant role making the whole model’s impact negligible (0.0032 increase in R squared
and p value of 0.8786). The positive unstandardized coefficient of 1.5294 did not show
transparency increases customer loyalty, because of the p-value of 0.4918 making this direct
effect insignificant. In other words, a (positive) effect of transparency on customer loyalty
could not be confirmed.
Because of the very wide LLCI and ULCI’s of the different regression models no
conclusions could be drawn on these (Hayes, 2014).
Concluding, it could be said hypothesis 1, 2 and 3 could not be confirmed, because the
direct effects between transparency and the DV’s were not significant. On top of this, the
moderators did not play a significant role.
[35]
Table 2
Interaction effects transparency on purchase intent, moderated by involvement and risk
aversion
R2-chng F df1 df2 p-value LLCI ULCI
Int_1 Transparency x Risk aversion 0.0003 0.0294 1.000 79.000 0.8642 -0.9315 0.7837
Int_2 Transparency x Involvement 0.0006 0.0508 1.000 79.000 0.8223 -6.9705 5.5525
Both interactions 0.0013 0.0571 2.000 79.000 0.9445
R R-sq F df1 df2 p-value
Model complete 0.3365 0.01132 2.0173 5.000 79.000 0.0852
Table 3
Interaction effects transparency on information need, moderated by involvement and
risk aversion
R2-chng F df1 df2 p-value LLCI ULCI
Int_1 Transparency x Risk aversion 0.0019 0.1537 1.000 79.000 0.6960 -1.0518 0.7056
Int_2 Transparency x Involvement 0.0007 0.0577 1.000 79.000 0.8109 -5.7539 7.3326
Both interactions 0.0020 0.0817 2.000 79.000 0.9216
R R-sq F df1 df2 p-value
Model complete 0.1763 0.0311 0.5066 5.000 79.000 0.7704
Table 4
Interaction effects transparency on customer loyalty, moderated by involvement and
risk aversion
R2-chng F df1 df2 p-value LLCI ULCI
Int_1 Transparency x Risk aversion 0.0007 0.0561 1.000 79.000 0.8133 -0.2973 0.2341
Int_2 Transparency x Involvement 0.0013 0.1066 1.000 79.000 0.7449 -2.3030 1.6538
Both interactions 0.0032 0.1296 2.000 79.000 0.8786
R R-sq F df1 df2 p-value
Model complete 0.1840 0.0338 0.5535 5.000 79.000 0.7352
[36]
Scheme 3. Model 1, H1.
Scheme 4. Model 2, H2.
Scheme 5. Model 3, H3.
Involvement
(W)
Risk Aversion
(M)
Purchase
Intent (Y)
Transparency
(X)
5.815
9
-0.0470-0.1756
Involvement
(W)
Risk Aversion
(M)
Information
Need (Y)
Transparency
(X)
0.2780
0.1125-0.1075
Involvement
(W)
Risk Aversion
(M)
Customer
Loyalty (Y)
Transparency
(X)
1.5294
-0.02900.0064
[37]
Discussion
Let us first come back to the research question as stated in the Literature Review section of
this thesis: ‘How does transparency impact 1) information need, 2) purchase intent and 3)
customer loyalty; and how do involvement and risk aversion moderate these effects?´
The aim of this research was to find out if transparency of companies about their
offerings/products results in a decreased information need, and an increased purchase intent
and customer loyalty. Based on different scientific articles, and the researchers’ opinions in
these articles, it was argued that involvement and risk aversion could play a moderating role
in this interesting relationship. In order to find an answer to the research question 9 different
hypotheses were formulated, which provided an interesting view on transparency in
marketing practice.
H1a: Transparency decreases information need.
H1b: Transparency decreases information need, but this relationship is moderated by
involvement, in the sense that the impact of transparency on information need is stronger
under high involvement conditions.
H1c: Transparency decreases information need, but this effect is moderated by risk aversion,
in the sense that a higher degree of risk aversion demands a higher degree of transparency in
order to decrease information need.
H2a: Transparency positively influences purchase intent.
H2b: Transparency positively influences purchase intent, but this relationship is moderated
by involvement, in the sense that the positive impact of transparency on purchase intent is
stronger when involvement is high.
[38]
H2c: Transparency positively influences purchase intent, but this relationship is moderated
by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree
of transparency in order to increase purchase intent.
H3a: Transparency positively influences customer loyalty.
H3b: Transparency positively influences customer loyalty, but this relationship is moderated
by involvement, in the sense that the positive impact of transparency on customer loyalty is
stronger under high involvement conditions.
H3c: Transparency positively influences customer loyalty, but this relationship is moderated
by risk aversion, in the sense that it has positive influence under low risk aversion conditions
only.
However, first of all is has to be mentioned no relationships between (product) transparency
and the dependent variables could be confirmed. This essentially means none of the
hypotheses could be attested. Let us start off with information need. The regression analyses
based on model number 2 from the David Hayes Process Guide (2014) showed that the direct
relationship between transparency and information need had a positive unstandardized
coefficient of 0.2780. This coefficient was coupled with a p-value of 0.9698. Because of this
insignificant p-value a direct relationship could not be confirmed. Previous studies have
shown information need is determined by information asymmetry. Lowering information
asymmetry by transparency (Shroff et al., 2005) reduces information need (Xiang, J. Zhou, X.
Zhou, & Ye, 2012). In other words, increased transparency should lower information need.
Nevertheless, the positive standardized coefficient found (0.2780) suggested the opposite, but
the insignificant p-value (0.9698) hinders decisive claims on this. Respondents were not able
to verify the disclosed information from the Unox soup in pack manipulations which probably
resulted in the case no effect of transparency on information need could be found. Different
[39]
researchers have argued for the importance of verifiable disclosed information in order to
actually lower information need (Loewenstein et al., 2013; Rijwijk & Frewer, 2012).
Transparency is ought to reduce search costs and therefore benefits consumers from an
economic perspective (Koslow, 2000). Next to this, consumers actively search for information
and will use this information in order to select products (McNeal, 1987). Transparency was
expected to provide this needed information and therefore enhance purchase intent. Bhaduri &
Ha-Brookshire (2011) claim transparency creates trust as a basis for a good buyer – seller
relationship and helps consumers evaluate ‘good business’, which makes transparency key in
increasing consumers’ purchase intents. On top of this, consumers are even willing to pay
significantly more for transparent products (Zhang et al., 2012). This research has not found
additional support to these claims of previous researchers. Because of a p-value of 0.4244, the
positive unstandardized coefficient of 5.815 found with a regression analyses, could not
confirm a direct relationship between transparency and purchase intent.
The third and final hypothesis about the relationship between transparency and
customer loyalty was initially substantiated by the research of Cohn & Wolfe (2012).
Subsequent research of these researchers in 2013 showed transparency is increasingly
becoming more important. Furthermore, Clemons (2008) found transparency is important for
resonance marketing and hyperdifferentiation leading to customer loyalty. The regression
analyses did not find an effect. The positive unstandardized coefficient of 1.5294 was coupled
with a p-value of 0.4918, making the anticipated relationship insignificant.
When looking at the separate Pearson correlations it could be seen that only two
dependent variables correlated: Customer loyalty and purchase intent correlated positively
with a Pearson correlation of 0.753** at the 0.01 significance level. In fact this was to be
expected, because purchase intent is the first step to at least the spurious loyalty Backman &
Crompton (1991) write about. There were no other significant relations between any of the
[40]
dependent variables.
The moderating variables risk aversion and involvement surprisingly did not have a
significant moderating role. Involvement, being a parent or not, was argued to be of influence
because parents are a completely different audience than other consumers (Carey et al., 2008).
The results of this thesis argue against this view based on this research showing that
involvement does not play a significant role in the relationship of transparency on information
need, purchase intent and customer loyalty. However, the dependent variable purchase intent
correlated negatively at the 0.05 level with the moderator involvement (-0.236*) showing
being a parent or not influences peoples’ purchase intent.
Risk aversion was expected to be of great importance, because studies conducted by
Shimp & Bearden (1982), Clemons (2008) and Kim et al. (2008) find that risk aversion as a
moderator positively influences consumers’ purchase intentions and via the intentions the
actual purchase. Next to this, the study of Kim et al. (2008) finds that risk is a more important
determinant of purchasing behavior than trust and benefit, which underlines the importance of
this concept (for business practice). Risk aversion can play en even more important
moderating role in the relationship with transparency and customer loyalty (Clemons, 2008).
In this research risk aversion did not play a significant moderating role. The negligible impact
of the moderators was substantiated by the different changes in the R’s square: In the
relationship of transparency on information need the moderators increased the R square with
0.0020. In the relationship of transparency on purchase intent the moderators increased the R
square with 0.0013, and in the final relationship between transparency and customer loyalty
the moderators increased to R square with only 0.0032. Finally, all F-values of the complete
regression models were below the critical values showing the unimportant role of the
moderators as well.
[41]
Despite, the Cronbach’s Alpha of the risk aversion scale was below 0.7 the analysis was still
continued, because low Cronbach’s Alpha scores are not always a problem (Boyle, 1991).
Business practice now knows involvement based on family composition does not have
to be taken into account, because being a parent or not does not impact any relationship
between transparency and the dependent variables. Because no effects of transparency on the
dependent variables could be confirmed additional research is needed. Many researchers think
transparency is a good practice, but this research cannot support them. On the other hand,
opposing views from Hultman & Axelsson (2007) and Koslow (2000) cannot be endorsed
either. The one thing that can be said based on this research is that transparency at least does
not seem to harm business practice, so why would business not be more transparent about
their products and reach for the more transparent and sustainable world Bhaduri & Ha-
Brookshire (2011) write about. Just do good business and let transparency help consumers
evaluate good business (Bhaduri & Ha Brookshire, 2011). In order to avoid skepticism as a
result of transparency (Koslow, 2000), make sure information is verifiable (Loewenstein et
al., 2013; Rijwijk & Frewer, 2012) in order to avoid possible disadvantages as a business.
Business practice should for now hold on to positivism around the construct transparency,
because nothing indicated us we should be anxious about it. Future research can possibly help
transparency rise as a ‘must’.
[42]
Future research
Further research should be conducted on the four types of transparency as mentioned in this
thesis. These were derived from Hultman & Axelsson (2007) and transformed from a B-to-B
setting to a B-to-C setting. Because of the lack of research on transparency dimensions in a B-
to-C setting such a research could be worthwhile. Cohn & Wolfe (2012) did develop
dimensions of transparency for a B-to-C setting, but these are too extensive and not
operational.
Secondly, more extensive research should be done on the moderator involvement.
Except for being a parent –or not, involvement is supposed to know many other forms
possibly influencing or moderating transparency.
Finally, more research should be conducted on the relationship between
(product) transparency and the three dependent variables information need, purchase intent
and customer loyalty. Just now this research has not been able to confirm a relationship either
being positive or negative. To start, a study about the importance of verifiable disclosed
(product) information could be worthwhile, because respondents were not able to verify
information in this research’ survey experiment resulting in a too wide perceived information
asymmetry: This could be a reason why no relationship between transparency and information
need could be confirmed. Because of the growing attention in the current world for
transparency (Bhaduri & Ha-Brookshire, 2011; Cohn & Wolfe, 2012; Cohn & Wolfe, 2013)
we must find decisive answers: A challenge lays ahead.
[43]
Conclusions
The research question ‘How does transparency impact 1) information need, 2) purchase
intent and 3) customer loyalty; and how do involvement and risk aversion moderate these
effects?´ was not answered decisively by this research.
At first, all relevant literature related to this research question was reviewed resulting
in an operational theoretical framework and substantiation of the hypotheses. Furthermore, the
appropriate research method was chosen. A survey experiment was created in Qualtrics and
distributed amongst respondents by using Facebook. All surveys were analyzed by using
SPSS 20 in order to find out about the correlations and to test the hypotheses.
When looking at the correlations and direct effects of transparency on the dependent
variables information need, purchase intent and customer loyalty; no relationship could be
confirmed based on this research. This is essentially meaning none of the hypotheses could be
confirmed or rejected. The belief of different researchers that transparency has a positive
impact on all three mentioned dependent variables do not have to be ‘thrown away’, but
additional research is needed.
The different analyses showed the moderating variables risk aversion and involvement
had no significant impact on the direct effect of transparency on purchase intent, information
need and customer loyalty. In the first place, because of the small contribution to the R
squared and the insignificant p-values. Because of this, all F-values of the three regression
models (Hayes, 2014) were below the critical values resulting in rejecting H0. This meant the
models did not have an impact on the dependent variables purchase intent, information need
and customer loyalty.
Altogether, this research offered some contributions to business practice and
[44]
marketing science. Firstly, a significant gap in marketing science was closed concerning the
relationship between transparency, purchase intent, information need, and the moderating
variables involvement and risk aversion. Namely, because the anticipated relationship of
transparency on the dependent variables has not been researched much in the past. Secondly,
business practice now knows involvement does not play a significant moderating role. No
negative consequences of transparency were found, so business practice should listen to the
fast majority of researchers whom are claiming transparency is a good practice, beneficial for
both businesses and consumers. Not being able to confirm the hypotheses should not hinder
businesses in becoming more transparent as long as they take the possible important role of
verifiability of the disclosed information into account. This could be important to avoid
skepticism (Koslow, 2000) and increase information need (Loewenstein et al., 2013; Rijwijk
& Frewer, 2012).
In the end, three interesting possibilities for future research were provided. The first
opportunity is doing more extensive research on the relationship between (product)
transparency, and the dependent variables information need, purchase intent and customer
loyalty. Since, this research has not confirmed a relationship yet a challenge lays ahead. The
first breakthrough could be found in researching verifiable product information. Secondly,
research should be conducted on transparency dimensions in a Business to Consumer setting.
Unfortunately, existing literature does not provide good and operational dimensions of
transparency in a B-to-C setting yet. Finally, the moderating variable involvement should be
researched more thoroughly, because this variable possibly knows many interesting forms
influencing transparency, and its impact on other variables.
[45]
Limitations
This research had one possible limitation related to the respondents selected, which can be
named ‘subject error’. Subject error is a research error which derives from chosen
respondents, which may give different responses as normal: This is due to the fact they are for
example night shift workers (Saunders & Lewis, 2012). Many respondents selected for this
research were familiar with retail stores, because they work at Albert Heijn. This could have
resulted in the case that these people were possibly familiar with the products used in the
experiment. For this reason these people could have had product knowledge without the
products used in the experiment even disclosing any information. To account for this a fairly
new product was chosen: Unox Soup in pack. A second possibility is that these people
developed preferences for certain products while handling these during their work, thus
possibly influencing their choices in the experiment.
Secondly, the low Cronbach’s Alpha as discussed in the Results –and Discussion part
of this thesis was a limitation of this research.
Thirdly, this research was not built upon transparency types designed for B-to-C
relationships. Due to lack of research on transparency in B-to-C settings, types scientifically
built for B-to-B settings had to be altered. Therefore, a strong scientific base for these
dimensions lacked. However, this means interesting possibilities for future research as well.
Final and most important limitation is that this research could not confirm a
relationship between (product) transparency, information need, purchase intent and customer
loyalty, moderated by involvement and risk aversion. Despite, many previous research
claiming transparency has a positive impact on information need, purchase intent and
customer loyalty more additional research is needed to irrefutably confirm this.
[46]
References
Abougomaah, Naeim H., John L. Schlater, and W. Gaidis. (1987). Elimination and choice
phases in evoked set formation. The Journal of Consumer Marketing, 4 (4), 67-73.
Arndt, J. (1967). Perceived Risk, Sociometric Integration, and Word-of-Mouth in the
Adoption of a new Food Product. Risk Taking and Information Handling in Consumer
Behavior. Harvard University, 289-316.
Backman, S.J., & Crompton, J.L. (1991). Differentiating between high, spurious, latent and
low loyalty participants in two leisure activities. Journal of Park and Recreation
Administration, 9(2), 1-17.
Bannister, E.M., & Hogg, M.K. (2004). Negative Symbolic Consumption and Consumers’
Drive For Self-Esteem. European Journal of Marketing, 38(7), 68-850.
Bauer, R.A. (1960). Consumer Behavior as Risk Taking. American Marketing Association,
389-398.
Bhaduri, G., & Ha-Brookshire, J. E. (2011). Do Transparent Business Practices Pay?
Exploration of Transparency and Consumer Purchase Intention. Clothing and Textiles
Research Journal, 29(2), 135-149.
Boyle, G.J. (1991). Does Item Homogeneity Indicate Internal Consistency or Item
Redundancy in Psychometric Scales?. Personality and Individual Differences, 12(3), 291-294.
Borland R., Wilson N., Fong G.T., Hammond D, Cummings, K.M., et al. (2009). Impact of
graphic and text warnings on cigarette packs: findings from four countries over five years.
Tobacco Control, 18(5), 64-358.
[47]
Bruce, D. (2001). Me and My Brands: How the Stories of the Brands We Use Contribute to
Our Own Personal Stories. Marketing Magazine, 106(8), 30.
Carey, L., Shaw, D., & Shiu, E. (2008). The impact of ethical concerns on family consumer
decision-making. International Journal of Consumer Studies, 32, 553-560.
Carter, C. R., & Rogers, D. S. (2008). A framework of sustainable supply chain management:
Moving toward new theory. International Journal of Physical Distribution & Logistics
Management, 38, 360-387.
Chu, J. &, Chintagunta, P.K. (2011). An Emperical Test of Warranty Theroies in the U.S.
Computer Server and Automobile Markets. Journal of Marketing, 75, 75-92.
Clemons, E.K. (2008). How Information Changes Consumer Behavior and How Consumer
Behavior Determines Corporate Strategy. Journal of Management Information Systems,
25(2), 13-40.
Cohn & Wolfe. (2012). Transparency and Authenticity: A New Communications Landscape?.
Corporate Affairs, 1-15.
Cohn & Wolfe (2013). From Transparency to Full Disclosure?. Corporate affairs, 1-23.
Coughlan, S. (2002). Easy money. The Guardian, 23 November.
Dalen van, J., & Leede de, E. (2009). Statistisch onderzoek met SPSS voor Windows. Den
Haag: Lemma
Ford, G.T., Smith, D.B., & Swasey, J.L. (1990). Consumer Skepticism of Advertising Claims:
Testing Hypotheses from Economics of Information. Journal of Consumer Research, 16(4),
433- 441.
[48]
Foster, L. (2004). Socially stylish. Drapers Record & Menswear, 22,
34-36.
Golan, E., Krissof, B., Kuchler, F., Calvin, L., Nelson, K., & Price, G. (2004). Traceability in
the U.S. food supply: Economic theory and industry studies. Agricultural Economic Report.
Washington DC: U.S. Department of Agriculture, ERS.
Hammond, D., Fong, G.T., McNeill, A., Borland, R., & Cummings, K.M. (2006).
Effectiveness of cigarette warning labels in informing smokers about the risks of smoking:
findings from the International Tobacco Control (ITC) Four Country Survey. Tobacco
Control, 15(3), 19-25
Hayes, D. (2014). Process Documentation. May 2nd
, 2014, Afhayes.com: www.afhayes.com
Hayes, D. (2014). Frequently asked questions. June 12th
, 2014, Afhayes.com:
www.afhayes.com.
Hickman, L. (2004). How green is your house? The Guardian, G2. 29
January, 2-3.
Hofstede, G. (1983). The cultural relativity of organizational practices and theories. Journal of
International Business Studies, 14, 75-89.
Hultman, J., & Axelsson. B. (2007). Towards a typology of transparency for marketing
management research. Industrial Marketing Management, 36(5), 627-635.
Kim, D.J., Ferrin, D.L., & Rao, H.R. (2008). A trust-based consumer decision-making model
in electric commerce: The role of trust, perceived risk, and their antecedents. Decision
Support Systems, 44(2), 544-564.
[49]
Koslow, S. (2000). Can the Truth Hurt? How Honest and Persuasive Advertising Can
Unintentionally Lead to Increased Consumer Skepticism. The Journal of Consumer Affairs,
34(2), 245-265.
Lamming, R. C., Caldwell, N. D., Harrison, D. A., & Phillips, W. (2001). Transparency in
supply relationships: Concept and practice. Journal of Supply Chain Management, 37(4),
4−10.
Lee, L., Frederick, S., & Ariely, D. (2006). Try It, You'll Like It: The Influence of
Expectation, Consumption, and Revelation on Preferences for Beer. Psychological Science,
17(12), 1054‐1058.
Lee, M.S.W., Conroy, D., Motion, J. (2009). Brand Avoidance: A Negative Promises
Perspective. Advances in Consumer Research, 36(1), 421-429.
Loewenstein, G., Sunstein C.R. & Golman, R. (2013). Disclosure: Psychology Changes
Everything. Annual Review of Economics, 1-33.
MCDaniel, C., & Gates, R. (2012). Marketing Research. John Wiley & Sons, inc
McNeal, J.U. (1987). Children as Consumers: Insights and Implications. Lexington, MA:
Lexington Books.
Mittal, B. (1988). The Role of Affective Choice Mode in the Consumer Purchase of
Expressive products. Journal of Economic Psychology, 9, 499-524.
Myers, T.A. (2011). HOTDECK: An SPSS Tool for Handling Missing data 1.
Communication Methods and Measures, 5(4), 297-310.
[50]
Olsen, S.O. (2003). Understanding the relationship between age and seafood consumption: the
mediating role of attitude, health involvement and convenience. Food Quality and Preference,
14(3), 199-209.
Pardo, C. (2013). Entrepeneurial Risk Aversion, Net Worth Effects and Real Fluctuations.
Review of Financial Economics, 22, 158-168.
Pennings, J.M.E., Wansink, B., & Meulenberg, M.T.G. (2002). A note on modelling
consumer reactions to a crisis: The case of the mad cow disease. International Journal of
Research in Marketing, 19(1), 91-100.
Reynolds, F. D., & W. R. Darden (1972). Why the Midi Failed. Journal of Advertising
Research. 12(8), 39-46
Rijswijk, W. & Frewer, L.J. (2012). Consumer needs and requirements for food and
ingredient traceability information. International Journal of Consumer Studies, 36, 282-290.
Rust, R.T., Lemon, K.N., & Zeithaml, V.A. (2004). Return on Marketing: Using Customer
Equity to Focus Marketing Strategy. Journal of Marketing, 68, 109-127.
Saunders, M., & Lewis, P. (2012). Doing Research in Business & Management: An Essential
Guide in Planning Your Project. Prentice Hall, Pearson
Shimp, T.A., & Bearden, W.O. (1982). Warranty and Other Extrincic Cue Effects on
Consumers’ Risk Perceptions. Journal of Consumer Research, 9(6), 38-46.
Slavin, M. (2009). Commentary: Transparency increases firms’ credibility. Daily Journal of
Commerce, 5.
[51]
Shroff, N., Sun, A.X., White, H.D., Zhang, W. (2005). Voluntary Disclosure and Information
Asymmetry: Evidence from the 2005 Securities Offerings Reform. Journal of Accounting
Research, 51(5), 1299-1345.
Thrasher, J.F., Rousu, M.C., Hammond, D., Navarro, A., & Corrigan, J.R. (2011). Estimating
the impact of pictorial health warnings and “plain” cigarette packaging: evidence from
experimental auctions among adult smokers in the United States. Health Policy, 102(1), 8-41.
Trouw. (2013). ‘Voor mijn baby alleen buitenlandse melk’. January 10th, 2013, Trouw.nl:
http://www.trouw.nl/tr/nl/4496/Buitenland/article/detail/3411121/2013/03/18/Voor-mijn-
baby-alleen-buitenlandse-melk.dhtml
Unox. (2014). ‘Vragen over soep in pak’. April 10th, 2014. Unox.nl:
http://www.unox.nl/nl/verder-op-unoxnl/vragen-over-soep-pak
Weber, U., & Hsee, C. (1998). Cross-cultural Differences in Risk Perception, but Cross-
cultural Similarities in Attitudes Towards Perceived Risks. Management Science, 44(9),
1205-1217.
Weber, E.U., Blais, A-R., & Betz N.E. (2002). A Domain-specific Risk-attitude Scale:
Measuring Risk Perceptions and Risk Behaviors. Journal of Behavioral Decision Making, 15,
263-290.
Xiang, P., Zhou, J., Zhou, X., & Ye, K. (2012). Construction Project Risk Management Based
on the View of Asymmetric Information. Journal of Construction Engineering &
Management, 138(11), 1303-1311.
Zhang, C., Bai, J., & Wahl, T.I. (2012). Consumer’s willingness to pay for traceable pork,
milk, and cooking oil in Nanjing, China. Food Control 27, 21-28.
[52]
Appendices
[53]
A. The actual experiment (treatment group)
(*in reality this experiment was conducted in Dutch)
Q1: I am a parent Yes / No
*Dear respondent,
thank you in advance for participating. Your participation is voluntary and you can opt out whenever you like. Your
answers will be treated trustworthy. This research wants to find out if transparency of products matters in your
purchasing behavior. I want to stress there are no bad answers, just be honest. This research will take 5 to 10 minutes.
Q2: I would buy this product (5-point Likert scale)
Manipulation 1 (high transparency)
Q3: I would buy this product (5-point Likert scale)
Manipulation 2 (high transparency)
Q2: I would buy this product (5-point Likert scale)
Manipulation1 (high transparency)
Q3: I would buy this product (5-point Likert scale)
Manipulation 2 (high transparency)
Q7: I would continue buying the product repeatedly in the
future (considering my children will consume them as
well).
(5-point Likert scale)
Q5: I would buy this product (5-point Likert scale)
Manipulation 3 (high transparency)
Q5: I would buy this product (5-point Likert scale)
Manipulation 3 (high transparency)
Q8: Risk aversion scale. – See appendix B –
Q4: Information needs scale. – See appendix C –
Q6: I would buy this product (5-point Likert scale)
Manipulation 4 (high transparency)
Q4: Information needs scale. – See appendix C –
Q6: I would buy this product (5-point Likert scale)
Manipulation 4 (high transparency)
Q7: I would continue buying the product repeatedly in the
future.
(5-point Likert scale)
Q9 – Q12: Did you know about manipulation 1 to 4?
Yes/No – See appendix C –
Q9 – Q12: Did you know about manipulation 1 to 4?
Yes/No – See appendix C –
Q8: Risk aversion scale. – See appendix B –
Read ‘low transparency’ instead of high transparency
for control condition.
Each 5-point Likert scale consists out of the following
options: Firmly disagree / disagree / neutral / agree
/ firmly agree
[54]
B. Risk aversion Scale
The risk aversion scale is derived from the research of Weber et al. (2002) on domain specific
scales for risk aversion. The following Likert scale based questions were used for this
research’ questions:
(*in reality this experiment was conducted in Dutch)
*Indicate the likelihood of engaging in each activity. Provide a rating from 1 to 5 using the
following scale:
1. Extremely unlikely
2. Unlikely
3. Unsure
4. Likely
5. Extremely likely
Eating ‘expired’ food products that still look okay.
Frequent binge drinking.
Ignoring some persistent physical pain by not going to the doctor.
Taking a medical drug that has a high likelihood of negative side effects.
Engaging in unprotected sex.
Never using sunscreen when you sunbathe.
Never wearing a seatbelt.
Not having a smoke alarm in or outside of your bedroom.
Regularly riding your bicycle without a helmet.
Smoking a pack of cigarettes a day.
[55]
C. Information needs scale
The information needs scale is derived from the research of Rijswijk & Frewer (2012) by
modifying the categories of searching for information by consumers into a scale. The
following Likert scale based questions were used for this research’ questions:
(*in reality this experiment was conducted in Dutch)
*Indicate the likelihood of engaging in each activity. Provide a rating from 1 to 5 using the
following scale:
1. Extremely unlikely
2. Unlikely
3. Unsure
4. Likely
5. Extremely likely
From now on I will search as much information as possible about products I buy (general statement).
From now on I will contact points of sale to gain information about products (based on Rijswijk & Frewer).
From now on I will contact the authorities to gain information about products (based on Rijswijk & Frewer).
From now on I will ask people I know to provide me information about products I want to buy (based on
Rijswijk & Frewer).
[56]
D. Revision letter 1
June 23, 2014
Joris Demmers, first supervisor Master thesis
Amsterdam Business School – UVA
Dear Joris Demmers,
Hereby I send you my revised master thesis in its final version. I believe we had a good understanding,
and I value your often quick and detailed feedback. Especially, the face-to-face sessions were of great
help in completing this final manuscript. Below, you can find de headlines of your feedback on which
I commented as well. Thanks for your support.
1. Your fist comment was about the writing style being to personal.
- After your explanation in person I understood what you meant by this. I have replaced my
personal writing style for an academic writing style, and removed my personal
experiences from the thesis.
2. Secondly, you commented on the literature review. It was not coherent enough and lacked a
certain “flow”. Other than this, you would appreciate stronger argumentation to the
hypotheses and a more in dept analysis of the construct transparency.
- I have revised the whole literature review and started off with a ‘writing plan’ as you
suggested. I tried to combine all lose pieces of information by finding connections and
used more compound sentences. I went back into the literature as well and found
interesting additional argumentation for the dependent variables information need and
purchase intent. Finally, I provided a more in dept view on transparency, were this
[57]
independent variable comes from, and how I operate this construct in my thesis. ‘Product
transparency’ is the way in which the broad construct of transparency was operated.
3. The contributions in the thesis’ introduction lacked clear indications.
- I have restructured the introduction and clearly mentioned the different contributions.
4. The methods section had to be revised in terms of building.
- I have restructured the methods sections as you indicated in the detailed feedback.
5. The Data section had to be matched with an existing study, because it was not structured well
and you did not like the lay out. Next to this, I initially found separate correlations for high –
and low transparency which cannot be.
- I have looked for an article with a comparable method and based my thesis’ lay out and
structure on it. I have made a mistake in SPSS which resulted in separate correlations, but
have revised this and created a single correlation matrix.
6. You do not believe involvement plays a moderating role and thought proper argumentation
lacked for using this variable.
- However, research by Carey et al. (2008) suggests parenthood does influence consumers’
choices. The Pearson correlations showed involvement had a negative influence on
purchase intent, so choosing this variables was not all to strange after all. Eventually,
involvement did not turn out to play an important role when looking at the regression
analyses.
Yours sincerely,
Patrick Heeremans
10660623
[58]
E. Revision letter 2
August 12, 2014
Joris Demmers, first supervisor Master thesis
Amsterdam Business School – UVA
Dear Joris Demmers,
Hereby I send you my revised master thesis in its final version. After receiving my grade I felt
disappointed, but your feedback helped me revise my thesis in a good way. On top of this, I believe
my thesis became a better whole. Thanks again for your support, invested time and effort.
1. Your fist comments were about the introduction. The proposed contributions did not exceed
‘never been researched before’. The relation between the different constructs was not clear.
- I have explained the relationships between the constructs more clearly and used previous
research to support my claims about a construct being an independent, dependent –or
moderating variable. Furthermore, I have explained the contributions better and stressed
the ‘never been researched before’ factor.
2. Secondly, you commented on the absence of argumentation and hypotheses for the moderator
relations.
- I have revised the literature review and looked for argumentation for the moderator
relationships in the literature. Next to this, I have produced hypotheses for all separate
moderator relationships.
[59]
3. The argumentation of hypothesis 1 was not strong enough and only made intuitively sense.
- I have followed your advice to look at the connection between information asymmetry and
information need.
4. Hypothesis 2 lacked nuance.
- I have added literature about the negative impact of negative transparency on purchase
intent.
5. Argumentation for hypothesis 3 had to be sharpened.
- I have read the article of Clemons (2008) again and rewrote my argumentation for
hypothesis 3 based on this article. Now, the thesis also explains why transparency leads to
customer loyalty and not only that it does.
6. The Research Question (RQ) had to be revised.
- The RQ was revised in order to show the three dependent variables were separate
relations.
7. There was no need for limitations in the methods section.
- The limitations were deleted from the methods section and are only to be found in the
limitations section of the thesis.
8. Argumentation for low Alpha scores had to be revised.
- Argumentation was revised.
9. It was written ANOVA was used, but this was untrue.
- I have mistaken to believe I did use ANOVA. After your explanation that my statistical
analyses actually did not use ANOVA I deleted the announcement for using these.
[60]
10. It was unclear why the results section used 3 separate models.
- After testing the correlations of the dependent variables (DV) it was shown these were not
correlated. Because of this it was right to use separate models instead of 1 model including
all DV’s. An explanation of this was also added to the thesis itself.
Yours sincerely,
Patrick Heeremans
10660623

Weitere ähnliche Inhalte

Ähnlich wie Master THESIS UVA P.J. Heeremans

The Wisdom of Some Do We Always Need HighConsensus to Shape.docx
The Wisdom of Some Do We Always Need HighConsensus to Shape.docxThe Wisdom of Some Do We Always Need HighConsensus to Shape.docx
The Wisdom of Some Do We Always Need HighConsensus to Shape.docxpelise1
 
Contract cheating detection workshop materials
Contract cheating detection workshop materialsContract cheating detection workshop materials
Contract cheating detection workshop materialsOlumidePopoola
 
Importance of Food Packaging and Its Relation to the Consumer's Demographic ...
	Importance of Food Packaging and Its Relation to the Consumer's Demographic ...	Importance of Food Packaging and Its Relation to the Consumer's Demographic ...
Importance of Food Packaging and Its Relation to the Consumer's Demographic ...inventionjournals
 
Entrepreneurial Intention
Entrepreneurial IntentionEntrepreneurial Intention
Entrepreneurial IntentionNavy Savchenko
 
Contents1.0Introduction11.1 Research Objectives11.docx
Contents1.0Introduction11.1 Research Objectives11.docxContents1.0Introduction11.1 Research Objectives11.docx
Contents1.0Introduction11.1 Research Objectives11.docxbobbywlane695641
 
An analytical framework on deceptive Advertising with reference to Hyderabad ...
An analytical framework on deceptive Advertising with reference to Hyderabad ...An analytical framework on deceptive Advertising with reference to Hyderabad ...
An analytical framework on deceptive Advertising with reference to Hyderabad ...ijtsrd
 
Consumer Involvement and Shopping Behaviour
Consumer Involvement and Shopping BehaviourConsumer Involvement and Shopping Behaviour
Consumer Involvement and Shopping BehaviourAssignment Work Help
 
Demographic analysis of factors influencing purchase of life insurance produc...
Demographic analysis of factors influencing purchase of life insurance produc...Demographic analysis of factors influencing purchase of life insurance produc...
Demographic analysis of factors influencing purchase of life insurance produc...Alexander Decker
 
Froehling Johnson Satisfaction Loyalty
Froehling Johnson Satisfaction LoyaltyFroehling Johnson Satisfaction Loyalty
Froehling Johnson Satisfaction Loyaltyhfroehling1
 
NICOLEBRITTON - DISSERTATION FINAL
NICOLEBRITTON - DISSERTATION FINALNICOLEBRITTON - DISSERTATION FINAL
NICOLEBRITTON - DISSERTATION FINALNicole Britton
 
Research Paper InstructionsYou will write a Research Paper on a .docx
Research Paper InstructionsYou will write a Research Paper on a .docxResearch Paper InstructionsYou will write a Research Paper on a .docx
Research Paper InstructionsYou will write a Research Paper on a .docxronak56
 
Consumers evaluation of unethical marketing behaviors
Consumers evaluation of unethical marketing behaviorsConsumers evaluation of unethical marketing behaviors
Consumers evaluation of unethical marketing behaviorsLeslie showalter
 

Ähnlich wie Master THESIS UVA P.J. Heeremans (20)

AD 504 Final Paper
AD 504 Final PaperAD 504 Final Paper
AD 504 Final Paper
 
The Wisdom of Some Do We Always Need HighConsensus to Shape.docx
The Wisdom of Some Do We Always Need HighConsensus to Shape.docxThe Wisdom of Some Do We Always Need HighConsensus to Shape.docx
The Wisdom of Some Do We Always Need HighConsensus to Shape.docx
 
Contract cheating detection workshop materials
Contract cheating detection workshop materialsContract cheating detection workshop materials
Contract cheating detection workshop materials
 
Importance of Food Packaging and Its Relation to the Consumer's Demographic ...
	Importance of Food Packaging and Its Relation to the Consumer's Demographic ...	Importance of Food Packaging and Its Relation to the Consumer's Demographic ...
Importance of Food Packaging and Its Relation to the Consumer's Demographic ...
 
Impact of Advertisement on Behaviour Of Children As Consumers
Impact of Advertisement on Behaviour Of Children As ConsumersImpact of Advertisement on Behaviour Of Children As Consumers
Impact of Advertisement on Behaviour Of Children As Consumers
 
Entrepreneurial Intention
Entrepreneurial IntentionEntrepreneurial Intention
Entrepreneurial Intention
 
Buy Essay Paper
Buy Essay PaperBuy Essay Paper
Buy Essay Paper
 
Contents1.0Introduction11.1 Research Objectives11.docx
Contents1.0Introduction11.1 Research Objectives11.docxContents1.0Introduction11.1 Research Objectives11.docx
Contents1.0Introduction11.1 Research Objectives11.docx
 
An analytical framework on deceptive Advertising with reference to Hyderabad ...
An analytical framework on deceptive Advertising with reference to Hyderabad ...An analytical framework on deceptive Advertising with reference to Hyderabad ...
An analytical framework on deceptive Advertising with reference to Hyderabad ...
 
Ostraa - Overview
Ostraa - OverviewOstraa - Overview
Ostraa - Overview
 
Consumer Involvement and Shopping Behaviour
Consumer Involvement and Shopping BehaviourConsumer Involvement and Shopping Behaviour
Consumer Involvement and Shopping Behaviour
 
Research proposal
Research proposalResearch proposal
Research proposal
 
Demographic analysis of factors influencing purchase of life insurance produc...
Demographic analysis of factors influencing purchase of life insurance produc...Demographic analysis of factors influencing purchase of life insurance produc...
Demographic analysis of factors influencing purchase of life insurance produc...
 
Froehling Johnson Satisfaction Loyalty
Froehling Johnson Satisfaction LoyaltyFroehling Johnson Satisfaction Loyalty
Froehling Johnson Satisfaction Loyalty
 
NICOLEBRITTON - DISSERTATION FINAL
NICOLEBRITTON - DISSERTATION FINALNICOLEBRITTON - DISSERTATION FINAL
NICOLEBRITTON - DISSERTATION FINAL
 
Research Paper InstructionsYou will write a Research Paper on a .docx
Research Paper InstructionsYou will write a Research Paper on a .docxResearch Paper InstructionsYou will write a Research Paper on a .docx
Research Paper InstructionsYou will write a Research Paper on a .docx
 
Cereal product
Cereal productCereal product
Cereal product
 
Ps12
Ps12Ps12
Ps12
 
Consumers evaluation of unethical marketing behaviors
Consumers evaluation of unethical marketing behaviorsConsumers evaluation of unethical marketing behaviors
Consumers evaluation of unethical marketing behaviors
 
Ready To Eat Foods
Ready To Eat FoodsReady To Eat Foods
Ready To Eat Foods
 

Master THESIS UVA P.J. Heeremans

  • 1. [1] The Effect of Transparency on Information Need, Purchase Intent and Customer Loyalty, moderated by Involvement and Risk Aversion Author: Patrick Heeremans Student number: 10660623 Supervisor (1): Joris Demmers Supervisor (2): Dr. Meg Lee Qualification: MSc. Business Studies – Marketing Track Institution: Amsterdam Business School – UVA Document: Master Thesis Version: Final version Date of submission: 19/08/2014
  • 2. [2] Table of contents Abstract ...................................................................................................................................... 3 Introduction ................................................................................................................................ 4 Literature Review....................................................................................................................... 8 Data and Method ...................................................................................................................... 19 Results ...................................................................................................................................... 29 Discussion ................................................................................................................................ 37 Conclusions .............................................................................................................................. 43 Limitations ............................................................................................................................... 45 References ................................................................................................................................ 46 Appendices............................................................................................................................... 52 A. The actual experiment (treatment group) ........................................................................ 53 B. Risk aversion Scale.......................................................................................................... 54 C. Information needs scale................................................................................................... 55 D. Revision letter 1............................................................................................................... 56 E. Revision letter 2............................................................................................................... 58
  • 3. [3] Abstract The growing interest for transparency comes from the accountancy sector, and is fed by different business scandals (Cohn & Wolfe, 2012). Next to this, consumers get a growing interest in being able to verify good business, in which transparency plays a central role (Bhaduri & Ha-Brookshire, 2011). Despite growing interest transparency is a concept in marketing science, which has not been researched much. This leaves us with many question marks and undefined opportunities for business practice. Marketing science knows many conflicting views on transparency. For example, Hultman & Axelsson (2007) and Koslow (2000) think transparency can be a bad practice resulting in problems. Meanwhile, most other researchers in this field argue for the opposite. In order to find answers a survey experiment was conducted and to analyze the different responds different statistical methods were used. In this research an effect of transparency on the dependent variables information need, purchase intent and customer loyalty could not be confirmed. However, let us not forget no negative effects of transparency were found as well. Based on this, business practice should not be anxious to become more transparent, because this is what consumers want (Bhaduri & Ha-Brookshire, 2011). On top of this, the current developments, like the digitalization, almost force business practice to disclose product information (Cohn & Wolfe, 2012; Cohn & Wolfe, 2013). Furthermore, when balancing the scale research arguing for more transparency outbalances research being skeptical about transparency. So, nothing should hold businesses from becoming more transparent about their products. Finally, no significant moderating effects of risk aversion or involvement were found.
  • 4. [4] Introduction Recently, we have encountered a real life example of transparency in combination with information need, purchase intent, customer loyalty, involvement and risk aversion, which inspired this research: Chinese manufacturers of baby milk used the dangerous ingredient melamine in their products, resulting in a business scandal: The usage of this ingredient coming to the front was resulting in scattered trust amongst Chinese consumers. These Chinese parents try to avoid possible risks connected to Chinese baby milk and started to purchase foreign baby milk, because Western firms are more transparent about their products and the ingredients they use (Trouw, 2013). On top of this, Chinese consumers are even willing to pay significantly more for milk if information about the product is provided, thus showing the important role transparency can play for both consumers and companies (Zhang et al., 2012). However, hardly any research on Western customers, and their purchase intent, customer loyalty and information need related to disclosure is conducted while this could be of great importance to business practice. In this real life example six key constructs are playing a central role: transparency, information need, purchase intent, customer loyalty, risk aversion and involvement. In this case the fact that melamine was used in producing baby milk and this practice coming to the front resulted in risk aversion, moderating information need, purchase intent and customer loyalty, because the Chinese consumers do not trust national manufacturers anymore. These Chinese consumers are most likely to be parents, thus determining their involvement. Shortage of information, lowered purchase intent and lowered customer loyalty for Chinese baby milk made Chinese consumers seek their salvation in Western products. Transparency causes an impact on information need, purchase intent and customer loyalty and can therefore be seen as the independent variable of the relationships in this real
  • 5. [5] life example. Different researchers also consider transparency to be an independent variable in the relationships with the dependent variables information need (Shroff et al., 2005; Xiang, J. Zhou, X. Zhou, & Ye, 2012), purchase intent (Koslow, 2000; Bhaduri & Ha-Brookshire, 2011) and customer loyalty (Clemons, 2008; Cohn & Wolfe, 2012; Cohn & Wolfe, 2013). However, this research is still limited in numbers and not conducted on Dutch consumers, whom are an unique audience (Hofstede, 1983), leading up to the first contribution of this thesis. Subsequently, no research has dealt with all three dependent variables of the real life example in one study despite these three variables playing an important role in the Baby-milk drama, at the same time. However, some researchers do mention some linkages between the dependent variables. For example, information need is argued to influence purchase intent (Bhaduri & Ha-Brookshire, 2011); purchase intent is the start of behavioral loyalty (Backman & Crompton, 1991); and information need could influence customer loyalty (Bhaduri & Ha- Brookshire, 2011). This research will bring a second contribution to knowledge by researching if correlations between the dependent variables exist, because based on current literature strong conclusions cannot be drawn yet. Furthermore, the moderating roles of involvement and risk aversion have not been researched much. Actually, only Carey et al. (2008) and Mittal (1988) have conducted thorough research on the moderating role of involvement in relation to transparency. Despite, involvement seemingly playing an important role in the Chinese baby milk example no break- through research on involvement as moderator is to be found. Carey et al. (2008) share this view and argue for more research on this construct: Parents have started to think totally different about the world and ethics (Carey et al., 2008), and transparency could play an important role in this changed perception (Bhaduri & Ha-Brookshire, 2011). Furthermore, involvement is important to research, because parents can be considered to be a completely different audience than other consumers (Carey et al., 2008). Hereby, this research can bring a
  • 6. [6] third contribution to knowledge about involvement and its relationship with transparency. As fourth contribution, this research provides knowledge about involvement to help business practice to find out how to manage transparency for two complete customer groups; parents and non-parents (Carey et al., 2008). Risk aversion on the other hand has been researched much in the past, but not in combination with the independent variable transparency. Especially, the moderating role of risk aversion in consumer settings has not been researched much, which is a potentially important gap because of the real life illustration of risk aversion as a moderator in the Chinese baby milk gate. Chu & Chintagunta (2011) and Pardo (2013) studied the impact of risk aversion as a moderator for the relationship of transparency on information need; Shimp & Bearden (1982), Clemons (2008) and Kim et al. (2008) studied the moderating impact of risk aversion on the relationship of transparency on purchase intent, and only one research was conducted on the moderating impact of risk aversion on customer loyalty by Clemons et al. (2008). By using risk aversion as moderating variable, as in the Chinese baby milk example, this research addresses a gap in marketing literature and results in a fifth contribution. Customer loyalty (retention), possibly influenced by transparency, is the most important factor of the Customer Lifetime Value formula and therefore profit (Rust et al., 2004), which underlines the value of this research for business practice. This resulted in the sixth major contribution of this research. In order to make all contributions this research used an experimental survey design to create purchase simulations in which respondents were confronted with highly transparent products and low transparent products. The respondents were divided into parents and non- parents in order measure the impact of involvement. Risk aversion, not yet used as moderator in current Marketing literature, will be measured by creating Likert scales.
  • 7. [7] Respondents’ choices as a result of the degree of product transparency will show the impact of transparency on information need, purchase intent and customer loyalty, moderated by risk aversion and involvement. The goal of this research is to find out how the independent variable transparency influences de dependent variables information need, purchase intent and customer loyalty. Furthermore, this research strives to confirm a moderating role of risk aversion and involvement. By melting all these constructs together we can see that this research addresses an interesting relationship model and an important gap in Marketing science, and could help business practice increase consumers’ purchase intents, lower information needs and even more attractive; making customers return.
  • 8. [8] Literature Review The overarching topic for this master thesis will be ‘Transparency in Marketing’. Transparency, in its simplest form, means sharing information and is an element of buyer- seller relationships (Hultman & Axelsson, 2007). Subsequently, transparency has to be ‘free from delusion’ (Cohn & Wolfe, 2012) and is increasingly becoming more important. This importance comes from the accountancy sector and is fed by different business scandals of which the banking crisis is probably the best known example (Cohn & Wolfe, 2012). On top of this, the digitalization, and increasingly improving communication channels and information spreading really puts pressure on companies to become more transparent, because consumers are increasingly becoming better able at finding information themselves and have a growing awareness (Cohn & Wolfe, 2012). Cohn & Wolfe (2012) even call transparency a new form of currency, which is information based and builds trust with customers and other stakeholders. This is supported by Zhang et al. (2012) who claim transparency increases consumers’ willingness-to-pay for at least the Chinese market. Despite, transparency comes from the accounting sector this does not mean consumers experience financial transparency, like annual reporting, as being very important. Product transparency, essentially openness of a brand about the products it offers, is considered to be far more important to consumers (Cohn & Wolfe, 2012). This thesis used this product transparency to operate the broad concept of transparency, because of its importance to consumers. Cohn & Wolfe (2012) argue that the ongoing economic unrest and instability facilitates the growing need for transparency. Bhaduri & Ha-Brookshire (2011) add that the growing concern for how businesses perform, fed by environmentalists and other nongovernment organizations spreading awareness about the current abuses of planet earth, increases consumers’ demands for transparent brands.
  • 9. [9] Slavin (2009) adds today’s consumers are conscious about their society and environment, demanding transparent and sustainable products: The market for such products is expected to grow by as much as 19% by 2014. Transparency knows different stages: A buyer-seller relationship can be based on complete transparency, a middle path of ‘translucency’, information can be partially shared and a situation is possible in which no information is shared at all (Lamming et al., 2001). This master thesis used two extremes: a high degree of product transparency and a low degree product transparency. These extremes were chosen to be able to draw strong conclusions on consumer choices by leaving out the middle path Lamming et al. (2001) write about, because two extremes force consumers to really favor product transparency or not. Cohn & Wolfe (2012) consider 9 types of transparency being important as indicators of a transparent organization and have ranked them as followed: Nr. 1: Information about where an organization gets its materials/ingredients from. Nr. 2: Source of funding and ownership of the business. Nr. 3: Information about profits and losses. Nr. 4: Transparency regarding the suppliers of the company. Nr. 5: Transparency regarding the partners of the company. Nr. 6: Information about the payment of senior management. Nr. 7: Information about the payment of all layers of employees. Nr. 8: Transparency regarding community activities and charities the company is involved in. Nr. 9: Information regarding political alignment.
  • 10. [10] However, this long list mentioned by Cohn & Wolfe (2012) is less operational compared to the four types of transparency Hultman & Axelsson (2007) distinguish. This research is based on a Business to Business (B-to-B) setting and names the following four types of transparency: technological transparency, supply (chain) transparency, cost/price transparency and organizational transparency. These four types will be modified to a Business to Consumer (B-to-C) environment into different types of product transparency. The reason to chose for the Hultman & Axelsson (2007) types of transparency over the Cohn & Wolfe (2012) list of transparency types is that these leave out the types that are of lesser importance to consumers. Most previous research considers transparency to be a good practice for both businesses and consumers. An example of this is the study of Carter & Rogers (2008) which suggests that businesses need to be transparent with their operations to maintain business legitimacy and to build reputation. Only few researchers consider negative outcomes: Hultman & Axelsson (2007) and Koslow (2000) find proof for transparency resulting in problems. Hultman & Axelsson think Transparency in a B-to-B setting can result in a power shift and increases uncertainty. On top of this, transparency demands a high level of trust, because otherwise the relationship between buyer and seller will become less trustworthy and more opportunistic (Hultman & Axelsson, 2007). All this makes transparency a fragile concept not easily to reap the benefits from. Koslow (2000) argues consumers get more skeptical when confronted with ‘honest’ transparency in the form of advertising. This skepticism flows forth from brands sometimes lying and providing sales arguments lacking credibility, but even more important skepticism is a mechanism that protects consumers from brands that provide information to sale rather than inform. In this sense, skepticism can be considered to be a defense mechanism. This mechanism can be considered to be a problem, because skepticism limits sales.
  • 11. [11] Information need This negative impact of limited sales can be nullified when brands provide verifiable information (Ford, Smith, & Swasy, 1990). Loewenstein et al. (2013) and Rijswijk & Frewer (2012) even go a step further in stating that providing traceable (product) information can even lower information need. On the other hand, consumers also tend to become more and more skeptic towards information provided, even though this information is surprisingly honest (Koslow, 2000). Traceability is not only important from a consumer perspective, because it starts with brands being able to trace all the ingredients of their products themselves. On top of this, developing tracing mechanisms is one of the most important tools in order to regain consumers’ confidence (Golan et al., 2004). Information need is the first dependent variable of this research. Information need can be lowered, if consumers´ information asymmetry opposed to brands is narrowed (Xiang, J. Zhou, X. Zhou, & Ye, 2012). Information asymmetry means that either the buyer or the seller has superior knowledge over the other. In most cases it is the seller who has most information resulting in a disadvantage, or information need for the buyer (Xiang, J. Zhou, X. Zhou, & Ye, 2012). Transparency lowers information asymmetry (Shroff et al., 2005), and in that sense reduces peoples´ information need as well: All this resulted in the first hypothesis. H1a: Transparency decreases information need. Purchase intent Consumers actively search for needed information and will use this information in order to select products (McNeal, 1987). Transparency is able to provide this needed information and therefore enhances purchase intent. Koslow (2000) adds advertising, and providing information in that sense, reduces search costs and enhances consumers welfare. The same accounts for companies being transparent, thus providing information. Economically spoken
  • 12. [12] transparency should increase purchase intent, because it benefits consumers searching for product information. If this theory is true will be tested with the second hypothesis. Another reason to test this hypothesis is that for consumers who are concerned about environment and society, product transparency can be important for their purchase intent. This is the case, because transparency creates trust as a basis for a good buyer – seller relationship (Bhaduri & Ha-Brookshire, 2011). Consumers’ perceived value perspectives towards ‘good business’ and the theory of reasoned action makes transparency key in their purchase intent as well (Bhaduri & Ha-Brookshire, 2011). The theory of reasoned action argues people’s intention to perform a certain action, like purchasing a product, is a function of these people’s subjective norms (Bhaduri & Ha-Brookshire, 2011). In other words, a person will have a high purchase intent when he or she gets his or her beliefs of good business confirmed by the transparency of this certain product. A person can confirm his or her beliefs, because transparency gives the opportunity to evaluate good business practices (Bhaduri & Ha-Brookshire, 2011). An important distinction has to be made between positive and negative information disclosure. Negative information even has a stronger impact on consumer decision making, meaning consumers will not purchase a product as a result from it (Arndt, 1967; Reynolds & Darden, 1972). Despite clearly being advocates of transparent business practices Bhaduri & Ha-Brookshire (2011) also conclude disclosure of negative information will result into a lowered purchase intent. To exclude this effect this thesis’ survey experiment did not use (strong) negative information. Purchasing a product can be considered to be the first step to customer loyalty in terms of at least behavioral loyalty. Backman & Crompton (1991) call this spurious loyalty. H2a: Transparency positively influences purchase intent.
  • 13. [13] Transparency is ‘love’..? Besides spurious loyalty customer loyalty knows 3 additional levels; high, low and latent, which are driven by emotion and behavioral factors (Backman & Crompton, 1991). Transparency is the third most important driver of customer loyalty, after quality and price, and is even more important in the decision making process than brand appeal, and recommendations online –and from friends are (Cohn & Wolfe, 2012). The importance of transparency has even grown in 2013 (Cohn & Wolfe, 2013). People become more loyal towards a brand, because of transparency, because people are very skeptic about companies ever since recent scandals, such as the banking crisis. So, companies that are being transparent reduce risks for consumers and lower skepticism and distrust, and therefore increase customer loyalty (Cohn & Wolfe, 2012). On top of this, people will only buy brands they love. This love essentially is customer loyalty and can come forth from information disclosure, because transparency is an important basis of resonance marketing (Clemons, 2008). Resonance marketing means managing and profiting from hyperdifferentiation, by ‘creating’ loyal customers. Clemons (2008) describes hyperdifferentiation as the ability of a brand to offer numerous products that fit many different needs and wants. This would not make sense if people do not know what products a brand offers, what these products are, where these products come from and how these are produced. In other words, consumers need product transparency to evaluate –and optimize choices, which will result in brand love –or customer loyalty (Clemons, 2008). This view on transparency and customer loyalty resulted in the third hypothesis. H3a: Transparency positively influences customer loyalty.
  • 14. [14] Involvement After mentioning the three dependent variables; information need, purchase intent and customer loyalty the moderating variables risk aversion and involvement are being discussed now. An example of a study in which involvement is conceptualized is Olsen’s study (2003): This research is speaking of health involvement mediating the consumption (i.e. preferences) of sea food. In order to conceptualize involvement in this thesis being a parent or not was chosen, because parents can be considered to be a completely different audience than other consumers (Carey et al., 2008). Furthermore, reducing risks of purchases for your baby or child might be even more important to parents than to others. Risk and involvement seem to be interrelated, and risk has a strong connection to transparency. Namely, transparency is able to reduce risks (Hultman & Axelsson, 2007). Involvement facilitates the need for information processing (Mittal, 1988). The higher a person’s involvement the more important cognitive processing becomes for buying a product opposed to affective based motives (Mittal, 1988). In other words, high involvement increases peoples’ information need (Mittal, 1988) and therefore moderates the impact of transparency on the dependent variable information need, resulting in the following hypothesis: H1b: Transparency decreases information need, but this relationship is moderated by involvement, in the sense that the impact of transparency on information need is stronger under high involvement conditions. Ethical issues are very important to consumers willing to buy products, and transparency plays an important role in verifying ‘good business’ and eventually consumers’ purchase intents (Bhaduri & Ha-Brookshire, 2011). However, this grown interest for ‘good practices’ cannot be explained by a grown awareness of ethics alone, but is for a large part
  • 15. [15] facilitated by the arrival of a new born child (Coughlan, 2002; Foster, 2004; Hickman, 2004; in Carey et al., 2008). Parenthood as a determinant of involvement makes ethical decision making even more important (Carey et al., 2008) and in that sense moderates the relationship of hypothesis 2. Ethical decision making gets facilitated more by involvement, because the birth of a child ‘awakes’ peoples’ sense of ethics, and changes peoples’ view of the world. In other words, parents want to contribute to a better world (Carey et al., 2008). H2b: Transparency positively influences purchase intent, but this relationship is moderated by involvement, in the sense that the positive impact of transparency on purchase intent is stronger when involvement is high. Involvement also plays an important role in customer loyalty, because being able to evaluate choices is more important when involvement is high (Carey et al., 2008). In this sense, the importance of transparency for hyperdifferentiation grows, thus making Clemons’ (2008) ‘resonance marketing’ more important under high involvement conditions. Being able to make ‘good’ consumer choices, and being loyal towards a selection of these choices means parents can make a trade off and balance their less ethical consumer decisions, which is an important driver for their consumer choices (Carey et al., 2008). Transparency helps making such trade offs (Clemons, 2008). H3b: Transparency positively influences customer loyalty, but this relationship is moderated by involvement, in the sense that the positive impact of transparency on customer loyalty is stronger under high involvement conditions. Risk aversion Risk, an old –and well known concept in research, is what consumers perceive as a possible negative outcome of a purchase. In other words, it is the possibility a purchase will not meet
  • 16. [16] the expectations or anticipated benefits (Bauer, 1960). The way a company communicates information about their products or services determines the risk aversion of a consumer (Shimp & Bearden, 1982). An important driver of this second concept, (perceived risk) reduction, is the verifiability of given information, because transparency without the possibility of tightening the given information is useless (Loewenstein et al., 2013). Different studies have talked about pictures (of negative impacts of smoking) increasing the awareness of risk (Borland et al., 2009; Hammond et al., 2006; Thrasher et al., 2011). Information need has much to do with perceived uncertainty of consumers; and increased transparency, as a source of information, can lower this uncertainty (Shimp & Bearden, 1982). The same view is shared by Hultman & Axelsson (2007). Pardo (2013) researched the moderating role of risk aversion in entrepreneurial investment decisions. His study found that the independent variable information asymmetry, determined by transparency, influences entrepreneurial investment decisions and is moderated by risk aversion. Risk aversion as a moderator has also been researched in consumption settings. For example, Chu & Chintagunta (2011) claim risk aversion is an important moderator for the relationship of warranties and information asymmetry. Information asymmetry is determined by transparency (Shroff et al., 2005), and the higher information asymmetry is the more important a warranty is, because warranties provide a sense of protection against knowing less than the seller (Chu & Chintagunta, 2011). Risk aversion moderates this relationship, because the sense of feeling protected becomes more important when risk aversion is higher. So, transparency lowers information need and lowers the importance of warranties. However, risk aversion moderates the impact on this, because a higher risk aversion demands more transparency to lower information asymmetry and subsequently lower information need, in order to lower the importance of warranties.
  • 17. [17] H1c: Transparency decreases information need, but this effect is moderated by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree of transparency in order to decrease information need. Shimp & Bearden (1982) claim that if consumers are able to assess products, they will have a lowered risk aversion: This has mostly to do with increased trust according to Kim et al. (2008). Subsequently, this lowered risk aversion results in an increased purchase intent (Shimp & Bearden, 1982; Clemons, 2008; Kim et al., 2008). Transparency plays an important role in offering consumers the ability to assess products (Clemons, 2008). So, risk aversion has a moderating role on the relationship between transparency and purchase intent, because a higher degree of risk aversion demands a higher degree of transparency to increase purchase intent (Kim et al., 2008). H2c: Transparency positively influences purchase intent, but this relationship is moderated by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree of transparency in order to increase purchase intent. In case of the dependent variable customer loyalty, risk aversion even plays a more important moderating role opposed to purchase intent: A product towards which a person is loyal to can be considered to be an optimal choice (Clemons, 2008). Risk aversion strongly influences this optimal choice, because in case of a high degree of risk aversion, a person will not bond to a certain product (Clemons, 2008). So, transparency increases customer loyalty, but only when a person’s risk aversion is low (Clemons, 2008). H3c: Transparency positively influences customer loyalty, but this relationship is moderated by risk aversion, in the sense that it has positive influence under low risk aversion conditions only.
  • 18. [18] This research will be conducted in the Netherlands, which has an impact on the importance of risk aversion for consumers, because risk aversion tends to be different per country. A study conducted by Weber & Hsee (1998) finds that respondents from China, the U.S.A., Germany and Poland are different in their risk perceptions towards different financial options. Another study conducted by Pennings et al. (2002) finds that consumers from different countries react differently to food crises like the mad cow disease. The study also finds that Dutch consumers are significantly less risk averse compared to German consumers and slightly less risk averse compared to North American consumers. On top of this, the study finds that involvement is negatively related to risk aversion: the more you consume a product the less risk averse you will be in case of a crisis. In general Dutch people score relatively low on risk aversion (Hofstede, 1983). In order to find out if product transparency has negative or positive consequences for business practice and to conclude if consumers really value this construct; the following research question was answered: ‘How does transparency impact 1) information need, 2) purchase intent and 3) customer loyalty; and how do involvement and risk aversion moderate these effects?’ Scheme 1: Relations of the research question Transparency Involvement Risk aversion Information need Purchase intent Customer Loyalty
  • 19. [19] Data and Method Method The method used for this thesis is the experimental design (in a controlled setting). An experiment is a “research strategy that involves the definition of a theoretical hypothesis; selection of samples of individuals from known populations; the allocation of samples to different experimental conditions; the introduction of planned change on one or more variables; and measurement on a small number of variables and control of other variables” (Saunders & Lewis, 2012, pp 114). A major disadvantage of conducting an experiment is the lack of external validity, which is the possibility to generalize findings beyond the research setting. However, experiments tend to be high on internal validity (MCDaniel & Gates, 2012). Internal validity is important to the research question of this thesis, because this study wants to proof the impact of product transparency (disclosure of companies about their products) on information need, purchase intent and customer loyalty, moderated by risk aversion and involvement. A major benefit of experiments is that different variables can easily be adjusted in this controlled setting (MCDaniel & Gates, 2012). On top of this, the experimental design is ideal to test the theoretical hypotheses (Saunders & Lewis, 2012). The lack of external validity could be overcome by conducting a field experiment (MCDaniel & Gates, 2012). This could have been an interesting possibility, but it would have been too difficult to adjust information in a real life setting.
  • 20. [20] Respondents The sample consisted out of students, family, friends and employees of different Albert Heijn stores. Despite, the fact these people where easily accessible the experiment did not suffer from common biases related to convenience sampling. This was due to the fact that opposed to regular convenience sampling the people selected where selected randomly (Saunders & Lewis, 2012). Sample size The experiment was based on a 2 x 2 between subjects design. Based on a general rule this means 25 participants per condition where needed. By using high (product) transparency, low (product) transparency, being a parent, and not being a parent as conditions this experiment needed a minimum of 100 participants. Independent variable (IV) The ‘degree of product transparency’, based on the two extremes of transparency of Lamming et al. (2001), was the independent variable used for this experiment. This experiment will distinguish 4 groups, of which 2 consist out of a group to which products with low transparency were shown and a group to which products with a high transparency were shown. Each respondent was shown 4 products and each of these products contained another type (manipulation) of transparency. In the low transparency condition hardly any info was given about the product, opposed to a high degree of disclosure in the high transparency condition. Each respondent in both the high –and low transparency condition was shown the same 4 products in order to really measure the impact of transparency on their information need, purchase intent and customer loyalty. In the meanwhile, other causes than transparency could be excluded, because every respondent was confronted with the same scenario except
  • 21. [21] from the degree of transparency. The manipulations used were based on three out of the four types of transparency of Hultman & Axelsson (2007). Each manipulation was separated and assigned to one question so that the differences in impact could be seen as well. The four types of transparency Hultman & Axelsson (2007) mention were technological transparency, supply (chain) transparency, cost/price transparency and organizational transparency, of which 3 were used in this research. Each of these types of transparency were slightly adapted to fit a Business to Consumer setting, opposed to the Business to Business setting Hultman & Axelsson (2007) published about. The fourth type of transparency used was health based, which was especially important to research in relation to involvement (Olsen, 2003). Stimuli The product used in the experiment was ‘Unox Soup in pack’. This is a fairly new product and was therefore unlikely to have associations that could interfere the results of the experiment. Having tried a product before might influence peoples’ purchase intent and customer loyalty beforehand. The choice for this product tries to account for this. Each of the stimuli was about disclosure of product information by the Unox brand, making these stimuli a form of product transparency. This type of transparency was mentioned in a study by Cohn & Wolfe (2012).
  • 22. [22] Pictures 1-4: Unox ‘Soup in pack’ pictures used in the experiment (without manipulations). Health transparency (“Erwtensoup” flavour) In the health transparency manipulation the salt intake per serving was shown on the package to the respondents in the high transparency condition. Each serving of 250 ml contains 0,88 grams of salt, which corresponds with 37% of the daily need of salt of an adult. Technological transparency (“Tomatensoep” flavour) In the technological transparency manipulation the package of Unox ‘Soup in pack’ was shown in combination with which technology the product was manufactured. The following text was communicated to the respondents of the experiment, which was derived from the Unox website: “Soup in bag and canned is pasteurized and sterilized longer. The time in the
  • 23. [23] production of soup in suit is significantly shorter. In addition, the soups are canned and bag snatched heated, while the soup in a pack is first heated before it is taken away. Due to the short and different manner of heating, the flavor and the color of the ingredients remains optimally preserved. The shelf life is the same (unox.nl, 2014).” Supply (chain) transparency (“Chinese Tomatensoep” flavour) In the supply chain manipulation in the high transparency condition was communicated were the product was produced, namely in the Netherlands. Despite, Unox is a Dutch company it offers Chinese tomato soup, not being produced in China at all or having a reference to the country in relation to the ingredients used. Knowing this could influence people’s purchase intent. However, Dutch people could prefer Dutch made products as well. Cost/price transparency (“Pompoensoep” flavour) This experiment has used price transparency in the high transparency situation by showing how many percent of the profits goes to farmers delivering the pumpkins for the Unox ‘Soup in pack’ product. Dependent variables (DV) Information need, purchase intent and customer loyalty were the dependent variables of the research question, because these were tested on the influence of (product) transparency, moderated by involvement and risk aversion. In order to measure the information needs of the respondents an information needs scale consisting out of 4 items was constructed based on the research of Rijswijk & Frewer (2012), which focusses on information traceability of products. This research mentions different categories of information traceability of products, of which 3 were incorporated in a self-constructed scale to measure information needs in this research. The three chosen categories were related to gathering information from authorities, relatives
  • 24. [24] and points of sale. The fourth item was a general statement about looking for information. It was not chosen to use an existing information needs scale, because each of these are based on a very different type of information need than used in this thesis. This scale was important to measure what the impact of transparency on information needs is. Increased transparency might facilitate a higher information need and therefore decrease the purchase intent (Bhaduri & Ha-Brookshire, 2011). The opposite could work as well. See appendix C. ‘Information need scale’ for the items used in the experiment. Purchase intent was measured by using a 5-point Likert scale. Respondents were confronted with each of the four earlier mentioned stimuli in either a high –or low transparency condition. From an economist perspective product transparency should increase purchase intent, because it benefits consumers by reducing search costs for information (Koslow, 2000). Secondly, Bhaduri & Ha-Brookshire (2011) studied transparency in the apparel industry and concluded transparency positively influences people’s purchase intent. If the same accounts for the food industry was researched by this thesis’ survey experiment. The third and final dependent variable customer loyalty was measured by a 5 – point Likert scale as well. Abougomaah, Naeim, Schlater, and Gaidis (1987) argue for price as being very important in terms of purchase intent towards a brand or product, and therefore brand loyalty as well. Cohn & Wolfe (2012) argue transparency is close to the importance of price for brand loyalty. Thus, making the relationship between customer loyalty and transparency relevant according to previous research, but still covering an existing research gap as well. Moderators In order to conceptualize involvement in this thesis being a parent or not was chosen, because the real life example of the Baby milk drama has clearly illustrated the impact of involvement
  • 25. [25] on the different relationships between transparency and the DV’s. Furthermore, different research supports this view, such as the studies of Mittal (1988) and Carey et al. (2008). In order to see if parenthood, transparency and the DV’s relate to each other the Pearson correlations for these variables were analyzed. Risk aversion was the second moderating variable. It is important to find out how the respondents rate their risk aversion in order to use this as a moderating variable. The chosen questions to asses respondents’ risk aversion were based on a 5-point Likert scale and were derived from the research of Weber et al. (2002) on domain specific risk-attitude scales. The research of Weber et al. (2002) distinguishes 5 types of risk-attitude domains, which are ethical, financial, health/safety, recreational and social. The health/safety scale was chosen for this experiment, because the transparency of the products in this research was mainly health/safety based. This was the case, because people were confronted, with more or less, information regarding a consumption/eatable product. See appendix B. ‘Risk aversion scale’ for the items used in the experiment. Treatments This experiment distinguished 4 groups, of which 2 consisted out of a group to which products with low transparency were shown and a group to which products with high transparency were shown. The third and fourth group are respectively a group consisting out of parents, and a group consisting out of non-parents. The treatments as such were based on involvement (being a parent or not) and on product transparency (low –and high transparency), resulting in 4 treatments. The treatment group was confronted with a high transparency condition and the control group with a low transparency condition to find out the differences.
  • 26. [26] Tool The experiment was conducted in a survey format online using the program Qualtrics. Qualtrics is a program which enables researchers and students to make online surveys. This program is suited for this research, because it facilitates the opportunity for creating a survey based experiment. Procedure The following procedure was used to conduct the experiment. This procedure resulted in answering the research questions by testing the mentioned hypotheses. Various people were contacted by dropping a link to the experiment on Facebook. The respondents were not selected, but reacted to the link themselves to keep the sample free from biases related to selecting the sample. Before participating the experiment each respondent had the opportunity to read about the research and the goal of the experiment. As being a good practice in conducting an experiment these respondents were told they were not obliged to cooperate and were ensured that their responds would be operated in trust. The experiment was divided in two parts; the ‘Low transparency condition’ and the ‘High transparency condition’: Qualtrics assigned each participant randomly to one of these conditions. Before each experiment respondents were asked if they are a parent or not, determining the ‘route’ of the experiment they took. This resulted in the division of a parent and a non-parent group.
  • 27. [27] Data analyses David Hayes In order to analyze the data the David Hayes (2014) paper was used primarily. This paper facilitates many models to choose from in order to make ‘prefab’ regression analyses accounting for, for example moderators and mediators. Scheme 2. Overall scheme of the different relations The scheme as shown above was split in 3 models, because analysis of the dependent variables showed these DV’s were not correlating. The first model ran the effect of transparency on consumers’ information need, moderated by involvement and risk aversion. The second model ran the effect of transparency on purchase intentions, moderated by involvement and risk aversion. The third model ran the effect of transparency on customer loyalty, moderated by involvement and risk aversion. Transparency (nominal/ IV) Involvement (Moderator) Risk aversion (Moderator) Information need (continuous/ DV) H1 / Research Question Purchase intent (continuous/ DV) H2 / Research Question Customer Loyalty (continuous/ DV) H3/ Research Question
  • 28. [28] Cronbach’s Alpha Two scales were used in the experiment. The first was the risk aversion scale and the second one was the scale measuring information need. For each of these scales the Cronbach’s Alpha was calculated showing if the scales were internally consistent, thus showing if each of the items used in the scales measured the same concept (Dalen & Leede, 2009). Correlation calculation The last type of statistical analyses conducted was calculating the correlations in order to find out if certain variables correlated strongly, and had to be combined; and in order to analyze positive and negative correlations between variables possibly contributing to answering the hypotheses. Finally, if applicable, the Pearson correlations were indicated as being significant in the correlations matrix, which shows if strong conclusions could be drawn based on the correlations.
  • 29. [29] Results Data cleaning and missing values In total 102 people responded to the survey experiment. At first, the missing data values were replaced by using Hotdeck variables. Hotdeck variables in a sense use donor values of a similar respondent in the dataset to replace missing data values (Myers, 2011). Hotdeck variables are typically useable in case of discrete values, rather than continuous ones (Myers, 2011). Fortunately, all data in this survey experiment were of discrete nature. The most important advantage of Hotdeck variables is that they are both valid (under most conditions) and easy to use (Myers, 2011). However, before being able to run the HotdeckMacro it should be checked if the missing data is < 10% (Myers, 2011). After analyzing the ‘missing system’ percentages in the various Frequency tables only one statement out of 25 had a higher missing data score than 10%. This was for the following statement ‘Q7: I would buy Unox Soup in bag repeatedly in the future considering my children will consume this product as well’. This was due to the fact respondents only were confronted with this question in case they are a parent, which was one of the conditions of the experiment. Because of the high missing data score, and the fact non-parents could not answer this question in the first place, Q7 was not used in the Hotdeck treatment. Recoding counter-indicative items All scales were analyzed to see if reverse coding was needed. This survey experiment used 2 different scales. None of the items of the information needs scale had to be reverse coded. However, items 6, 7 and 8 of the risk aversion scale did need reverse coding, because these
  • 30. [30] items used the word ‘never’ in their statements. Thus, resulting in an opposite answering of the question by respondents compared to the other items. Item 6: Never using sunscreen when you sunbathe. Item 7: Never wearing a seatbelt. Item 8: Not having a smoke alarm in or outside of your bedroom. Computing reliability As mentioned earlier the experiment used 2 scales. The Cronbach’s Alpha of the information needs scale was 0,814 and for the risk aversion scale 0.392. A rule of thumb is that a Cronbach’s Alpha of 0.7 is enough to show if items within a scale correlate (Dalen & Leede, 2009). Cronbach’s Alpha’s of scales sometimes can be improved by deleting certain items. However, researchers have to be careful in doing that, because this means deleting possible valuable information/data as well (Dalen & Leede, 2009). In case of the information needs scale the Cronbach’s Alpha if deleted can be 0.831 at most. However, it was argued that in this case the increase was too little to justify deleting data. Furthermore, the Cronbach’s Alpha was above 0,7 in first place after all. In case of the risk aversion scale the Cronbach’s Alpha could become 0.511 in case items 8 ‘Not having a smoke alarm in or outside of your bedroom’ and 9 ‘Regularly riding your bicycle without a helmet’ were deleted. Deleting these items resulted in a significant and justifiable (in terms of data loss) increase of the Cronbach’s Alpha of >0.10. Despite the rule of thumb, which states a Cronbach’s Alpha of 0.7 is needed to justify the use of a scale, the analysis was continued, because a low Cronbach’s Alpha is not always a problem (Boyle, 1991).
  • 31. [31] Correlation matrix In total three significant correlations were found. When looking at the Pearson correlations it could be seen that when a consumer knows how a product is produced this positively impacts customer loyalty at the 0.05 level with a Pearson correlation of 0.515*, thus making technological transparency very important. No significant correlations were found for transparency and one of the moderators or dependent variables. When looking at the correlations of the dependent variables it could be seen that almost every correlation was insignificant. The only significant effect could be found for customer loyalty and purchase intent correlating positively with a Pearson correlation of 0.753** at the 0.01 significance level. Other than this, the dependent variable purchase intent correlated negatively at the 0.05 level with the moderator involvement (-0.236*). Looking at the manipulation checks we could see respondents did not know about the disclosed product information in advance meaning this research yielded significant outcomes. The final results can be found in a table on the subsequent page:
  • 32. [32]
  • 33. [33] Regression analyses In total 3 models were run, because the dependent variables were not correlated. Otherwise, all variables would have been combined into one model. All models were matched with models shown in the ‘Process Guide’ of Andrew Hayes (2014). The models were based on a 95% confidence interval. By observing the models offered in this Process Guide eventually model number 2 was chosen, which is an additive moderation showing the conditional effect of X on Y (Hayes, 2014). This conditional effect consists out of the following formula: Y = c'1 + b7M + c'4W (Hayes 2014). All variables in the model were tested, which means a F- distribution was used to test the hypotheses (Dalen & Leede, 2009). The regression analyses of the different models resulted in various coefficients mentioned in schemes 3, 4 and 5: These were unstandardized coefficients (Hayes, 2014). The first model had transparency as IV and purchase intent as DV, with both involvement and risk aversion as moderators. The R square of model 1 was 0.3365. The adjusted R square was 0.1132 (<0.3365), which essentially means the model could not be improved (Dalen & Leede, 2009). The critical value was 2.33. Because of the lower F-value H0 is accepted, which means the model did not have a significant effect on purchase intent. The positive unstandardized coefficient of 5.815 as shown in scheme 3 had a p-value of 0.4244. Because of this p-value it could not be confirmed there was a positive direct impact of transparency on purchase intent. The moderators play a small role, making the model insignificant (p=0.9445). Next to this, the moderators resulted in a very small increase in the R squared of 0.0013, showing the minor role of them. On top of this, the impact of the moderators was insignificant as well with a p-value of 0.9554. Model 2 had transparency as IV and information need as DV, with involvement and risk aversion as moderators. This model had an R square of 0.1763. The adjusted R square
  • 34. [34] was 0.0311 (<0.1763). The F-value in the output was 0.5066. The critical value was 2.33 and because of the lower F-value H0 was accepted, which means the model did not have a significant effect on information need. The moderators played a small role, making the model insignificant (p=0.9216). Besides, the moderators resulted in a very small increase in the R squared of 0.0020, showing the minor role of these moderators. No significant direct effect of transparency on information need could be confirmed, because the (positive) unstandardized coefficient of 0.2780 had a p-value of 0.9698. The final model was about the impact of transparency (IV) on customer loyalty (IV), with involvement and risk aversion as moderators. The model had an R squared of 0.1840. The adjusted R square was 0.0338. The F-value in the output was 0.5538 and the critical value was 2.33. Because of the significantly lower F-value H0 was accepted, which means the model did not have a significant effect on customer loyalty. Again, the moderators played an insignificant role making the whole model’s impact negligible (0.0032 increase in R squared and p value of 0.8786). The positive unstandardized coefficient of 1.5294 did not show transparency increases customer loyalty, because of the p-value of 0.4918 making this direct effect insignificant. In other words, a (positive) effect of transparency on customer loyalty could not be confirmed. Because of the very wide LLCI and ULCI’s of the different regression models no conclusions could be drawn on these (Hayes, 2014). Concluding, it could be said hypothesis 1, 2 and 3 could not be confirmed, because the direct effects between transparency and the DV’s were not significant. On top of this, the moderators did not play a significant role.
  • 35. [35] Table 2 Interaction effects transparency on purchase intent, moderated by involvement and risk aversion R2-chng F df1 df2 p-value LLCI ULCI Int_1 Transparency x Risk aversion 0.0003 0.0294 1.000 79.000 0.8642 -0.9315 0.7837 Int_2 Transparency x Involvement 0.0006 0.0508 1.000 79.000 0.8223 -6.9705 5.5525 Both interactions 0.0013 0.0571 2.000 79.000 0.9445 R R-sq F df1 df2 p-value Model complete 0.3365 0.01132 2.0173 5.000 79.000 0.0852 Table 3 Interaction effects transparency on information need, moderated by involvement and risk aversion R2-chng F df1 df2 p-value LLCI ULCI Int_1 Transparency x Risk aversion 0.0019 0.1537 1.000 79.000 0.6960 -1.0518 0.7056 Int_2 Transparency x Involvement 0.0007 0.0577 1.000 79.000 0.8109 -5.7539 7.3326 Both interactions 0.0020 0.0817 2.000 79.000 0.9216 R R-sq F df1 df2 p-value Model complete 0.1763 0.0311 0.5066 5.000 79.000 0.7704 Table 4 Interaction effects transparency on customer loyalty, moderated by involvement and risk aversion R2-chng F df1 df2 p-value LLCI ULCI Int_1 Transparency x Risk aversion 0.0007 0.0561 1.000 79.000 0.8133 -0.2973 0.2341 Int_2 Transparency x Involvement 0.0013 0.1066 1.000 79.000 0.7449 -2.3030 1.6538 Both interactions 0.0032 0.1296 2.000 79.000 0.8786 R R-sq F df1 df2 p-value Model complete 0.1840 0.0338 0.5535 5.000 79.000 0.7352
  • 36. [36] Scheme 3. Model 1, H1. Scheme 4. Model 2, H2. Scheme 5. Model 3, H3. Involvement (W) Risk Aversion (M) Purchase Intent (Y) Transparency (X) 5.815 9 -0.0470-0.1756 Involvement (W) Risk Aversion (M) Information Need (Y) Transparency (X) 0.2780 0.1125-0.1075 Involvement (W) Risk Aversion (M) Customer Loyalty (Y) Transparency (X) 1.5294 -0.02900.0064
  • 37. [37] Discussion Let us first come back to the research question as stated in the Literature Review section of this thesis: ‘How does transparency impact 1) information need, 2) purchase intent and 3) customer loyalty; and how do involvement and risk aversion moderate these effects?´ The aim of this research was to find out if transparency of companies about their offerings/products results in a decreased information need, and an increased purchase intent and customer loyalty. Based on different scientific articles, and the researchers’ opinions in these articles, it was argued that involvement and risk aversion could play a moderating role in this interesting relationship. In order to find an answer to the research question 9 different hypotheses were formulated, which provided an interesting view on transparency in marketing practice. H1a: Transparency decreases information need. H1b: Transparency decreases information need, but this relationship is moderated by involvement, in the sense that the impact of transparency on information need is stronger under high involvement conditions. H1c: Transparency decreases information need, but this effect is moderated by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree of transparency in order to decrease information need. H2a: Transparency positively influences purchase intent. H2b: Transparency positively influences purchase intent, but this relationship is moderated by involvement, in the sense that the positive impact of transparency on purchase intent is stronger when involvement is high.
  • 38. [38] H2c: Transparency positively influences purchase intent, but this relationship is moderated by risk aversion, in the sense that a higher degree of risk aversion demands a higher degree of transparency in order to increase purchase intent. H3a: Transparency positively influences customer loyalty. H3b: Transparency positively influences customer loyalty, but this relationship is moderated by involvement, in the sense that the positive impact of transparency on customer loyalty is stronger under high involvement conditions. H3c: Transparency positively influences customer loyalty, but this relationship is moderated by risk aversion, in the sense that it has positive influence under low risk aversion conditions only. However, first of all is has to be mentioned no relationships between (product) transparency and the dependent variables could be confirmed. This essentially means none of the hypotheses could be attested. Let us start off with information need. The regression analyses based on model number 2 from the David Hayes Process Guide (2014) showed that the direct relationship between transparency and information need had a positive unstandardized coefficient of 0.2780. This coefficient was coupled with a p-value of 0.9698. Because of this insignificant p-value a direct relationship could not be confirmed. Previous studies have shown information need is determined by information asymmetry. Lowering information asymmetry by transparency (Shroff et al., 2005) reduces information need (Xiang, J. Zhou, X. Zhou, & Ye, 2012). In other words, increased transparency should lower information need. Nevertheless, the positive standardized coefficient found (0.2780) suggested the opposite, but the insignificant p-value (0.9698) hinders decisive claims on this. Respondents were not able to verify the disclosed information from the Unox soup in pack manipulations which probably resulted in the case no effect of transparency on information need could be found. Different
  • 39. [39] researchers have argued for the importance of verifiable disclosed information in order to actually lower information need (Loewenstein et al., 2013; Rijwijk & Frewer, 2012). Transparency is ought to reduce search costs and therefore benefits consumers from an economic perspective (Koslow, 2000). Next to this, consumers actively search for information and will use this information in order to select products (McNeal, 1987). Transparency was expected to provide this needed information and therefore enhance purchase intent. Bhaduri & Ha-Brookshire (2011) claim transparency creates trust as a basis for a good buyer – seller relationship and helps consumers evaluate ‘good business’, which makes transparency key in increasing consumers’ purchase intents. On top of this, consumers are even willing to pay significantly more for transparent products (Zhang et al., 2012). This research has not found additional support to these claims of previous researchers. Because of a p-value of 0.4244, the positive unstandardized coefficient of 5.815 found with a regression analyses, could not confirm a direct relationship between transparency and purchase intent. The third and final hypothesis about the relationship between transparency and customer loyalty was initially substantiated by the research of Cohn & Wolfe (2012). Subsequent research of these researchers in 2013 showed transparency is increasingly becoming more important. Furthermore, Clemons (2008) found transparency is important for resonance marketing and hyperdifferentiation leading to customer loyalty. The regression analyses did not find an effect. The positive unstandardized coefficient of 1.5294 was coupled with a p-value of 0.4918, making the anticipated relationship insignificant. When looking at the separate Pearson correlations it could be seen that only two dependent variables correlated: Customer loyalty and purchase intent correlated positively with a Pearson correlation of 0.753** at the 0.01 significance level. In fact this was to be expected, because purchase intent is the first step to at least the spurious loyalty Backman & Crompton (1991) write about. There were no other significant relations between any of the
  • 40. [40] dependent variables. The moderating variables risk aversion and involvement surprisingly did not have a significant moderating role. Involvement, being a parent or not, was argued to be of influence because parents are a completely different audience than other consumers (Carey et al., 2008). The results of this thesis argue against this view based on this research showing that involvement does not play a significant role in the relationship of transparency on information need, purchase intent and customer loyalty. However, the dependent variable purchase intent correlated negatively at the 0.05 level with the moderator involvement (-0.236*) showing being a parent or not influences peoples’ purchase intent. Risk aversion was expected to be of great importance, because studies conducted by Shimp & Bearden (1982), Clemons (2008) and Kim et al. (2008) find that risk aversion as a moderator positively influences consumers’ purchase intentions and via the intentions the actual purchase. Next to this, the study of Kim et al. (2008) finds that risk is a more important determinant of purchasing behavior than trust and benefit, which underlines the importance of this concept (for business practice). Risk aversion can play en even more important moderating role in the relationship with transparency and customer loyalty (Clemons, 2008). In this research risk aversion did not play a significant moderating role. The negligible impact of the moderators was substantiated by the different changes in the R’s square: In the relationship of transparency on information need the moderators increased the R square with 0.0020. In the relationship of transparency on purchase intent the moderators increased the R square with 0.0013, and in the final relationship between transparency and customer loyalty the moderators increased to R square with only 0.0032. Finally, all F-values of the complete regression models were below the critical values showing the unimportant role of the moderators as well.
  • 41. [41] Despite, the Cronbach’s Alpha of the risk aversion scale was below 0.7 the analysis was still continued, because low Cronbach’s Alpha scores are not always a problem (Boyle, 1991). Business practice now knows involvement based on family composition does not have to be taken into account, because being a parent or not does not impact any relationship between transparency and the dependent variables. Because no effects of transparency on the dependent variables could be confirmed additional research is needed. Many researchers think transparency is a good practice, but this research cannot support them. On the other hand, opposing views from Hultman & Axelsson (2007) and Koslow (2000) cannot be endorsed either. The one thing that can be said based on this research is that transparency at least does not seem to harm business practice, so why would business not be more transparent about their products and reach for the more transparent and sustainable world Bhaduri & Ha- Brookshire (2011) write about. Just do good business and let transparency help consumers evaluate good business (Bhaduri & Ha Brookshire, 2011). In order to avoid skepticism as a result of transparency (Koslow, 2000), make sure information is verifiable (Loewenstein et al., 2013; Rijwijk & Frewer, 2012) in order to avoid possible disadvantages as a business. Business practice should for now hold on to positivism around the construct transparency, because nothing indicated us we should be anxious about it. Future research can possibly help transparency rise as a ‘must’.
  • 42. [42] Future research Further research should be conducted on the four types of transparency as mentioned in this thesis. These were derived from Hultman & Axelsson (2007) and transformed from a B-to-B setting to a B-to-C setting. Because of the lack of research on transparency dimensions in a B- to-C setting such a research could be worthwhile. Cohn & Wolfe (2012) did develop dimensions of transparency for a B-to-C setting, but these are too extensive and not operational. Secondly, more extensive research should be done on the moderator involvement. Except for being a parent –or not, involvement is supposed to know many other forms possibly influencing or moderating transparency. Finally, more research should be conducted on the relationship between (product) transparency and the three dependent variables information need, purchase intent and customer loyalty. Just now this research has not been able to confirm a relationship either being positive or negative. To start, a study about the importance of verifiable disclosed (product) information could be worthwhile, because respondents were not able to verify information in this research’ survey experiment resulting in a too wide perceived information asymmetry: This could be a reason why no relationship between transparency and information need could be confirmed. Because of the growing attention in the current world for transparency (Bhaduri & Ha-Brookshire, 2011; Cohn & Wolfe, 2012; Cohn & Wolfe, 2013) we must find decisive answers: A challenge lays ahead.
  • 43. [43] Conclusions The research question ‘How does transparency impact 1) information need, 2) purchase intent and 3) customer loyalty; and how do involvement and risk aversion moderate these effects?´ was not answered decisively by this research. At first, all relevant literature related to this research question was reviewed resulting in an operational theoretical framework and substantiation of the hypotheses. Furthermore, the appropriate research method was chosen. A survey experiment was created in Qualtrics and distributed amongst respondents by using Facebook. All surveys were analyzed by using SPSS 20 in order to find out about the correlations and to test the hypotheses. When looking at the correlations and direct effects of transparency on the dependent variables information need, purchase intent and customer loyalty; no relationship could be confirmed based on this research. This is essentially meaning none of the hypotheses could be confirmed or rejected. The belief of different researchers that transparency has a positive impact on all three mentioned dependent variables do not have to be ‘thrown away’, but additional research is needed. The different analyses showed the moderating variables risk aversion and involvement had no significant impact on the direct effect of transparency on purchase intent, information need and customer loyalty. In the first place, because of the small contribution to the R squared and the insignificant p-values. Because of this, all F-values of the three regression models (Hayes, 2014) were below the critical values resulting in rejecting H0. This meant the models did not have an impact on the dependent variables purchase intent, information need and customer loyalty. Altogether, this research offered some contributions to business practice and
  • 44. [44] marketing science. Firstly, a significant gap in marketing science was closed concerning the relationship between transparency, purchase intent, information need, and the moderating variables involvement and risk aversion. Namely, because the anticipated relationship of transparency on the dependent variables has not been researched much in the past. Secondly, business practice now knows involvement does not play a significant moderating role. No negative consequences of transparency were found, so business practice should listen to the fast majority of researchers whom are claiming transparency is a good practice, beneficial for both businesses and consumers. Not being able to confirm the hypotheses should not hinder businesses in becoming more transparent as long as they take the possible important role of verifiability of the disclosed information into account. This could be important to avoid skepticism (Koslow, 2000) and increase information need (Loewenstein et al., 2013; Rijwijk & Frewer, 2012). In the end, three interesting possibilities for future research were provided. The first opportunity is doing more extensive research on the relationship between (product) transparency, and the dependent variables information need, purchase intent and customer loyalty. Since, this research has not confirmed a relationship yet a challenge lays ahead. The first breakthrough could be found in researching verifiable product information. Secondly, research should be conducted on transparency dimensions in a Business to Consumer setting. Unfortunately, existing literature does not provide good and operational dimensions of transparency in a B-to-C setting yet. Finally, the moderating variable involvement should be researched more thoroughly, because this variable possibly knows many interesting forms influencing transparency, and its impact on other variables.
  • 45. [45] Limitations This research had one possible limitation related to the respondents selected, which can be named ‘subject error’. Subject error is a research error which derives from chosen respondents, which may give different responses as normal: This is due to the fact they are for example night shift workers (Saunders & Lewis, 2012). Many respondents selected for this research were familiar with retail stores, because they work at Albert Heijn. This could have resulted in the case that these people were possibly familiar with the products used in the experiment. For this reason these people could have had product knowledge without the products used in the experiment even disclosing any information. To account for this a fairly new product was chosen: Unox Soup in pack. A second possibility is that these people developed preferences for certain products while handling these during their work, thus possibly influencing their choices in the experiment. Secondly, the low Cronbach’s Alpha as discussed in the Results –and Discussion part of this thesis was a limitation of this research. Thirdly, this research was not built upon transparency types designed for B-to-C relationships. Due to lack of research on transparency in B-to-C settings, types scientifically built for B-to-B settings had to be altered. Therefore, a strong scientific base for these dimensions lacked. However, this means interesting possibilities for future research as well. Final and most important limitation is that this research could not confirm a relationship between (product) transparency, information need, purchase intent and customer loyalty, moderated by involvement and risk aversion. Despite, many previous research claiming transparency has a positive impact on information need, purchase intent and customer loyalty more additional research is needed to irrefutably confirm this.
  • 46. [46] References Abougomaah, Naeim H., John L. Schlater, and W. Gaidis. (1987). Elimination and choice phases in evoked set formation. The Journal of Consumer Marketing, 4 (4), 67-73. Arndt, J. (1967). Perceived Risk, Sociometric Integration, and Word-of-Mouth in the Adoption of a new Food Product. Risk Taking and Information Handling in Consumer Behavior. Harvard University, 289-316. Backman, S.J., & Crompton, J.L. (1991). Differentiating between high, spurious, latent and low loyalty participants in two leisure activities. Journal of Park and Recreation Administration, 9(2), 1-17. Bannister, E.M., & Hogg, M.K. (2004). Negative Symbolic Consumption and Consumers’ Drive For Self-Esteem. European Journal of Marketing, 38(7), 68-850. Bauer, R.A. (1960). Consumer Behavior as Risk Taking. American Marketing Association, 389-398. Bhaduri, G., & Ha-Brookshire, J. E. (2011). Do Transparent Business Practices Pay? Exploration of Transparency and Consumer Purchase Intention. Clothing and Textiles Research Journal, 29(2), 135-149. Boyle, G.J. (1991). Does Item Homogeneity Indicate Internal Consistency or Item Redundancy in Psychometric Scales?. Personality and Individual Differences, 12(3), 291-294. Borland R., Wilson N., Fong G.T., Hammond D, Cummings, K.M., et al. (2009). Impact of graphic and text warnings on cigarette packs: findings from four countries over five years. Tobacco Control, 18(5), 64-358.
  • 47. [47] Bruce, D. (2001). Me and My Brands: How the Stories of the Brands We Use Contribute to Our Own Personal Stories. Marketing Magazine, 106(8), 30. Carey, L., Shaw, D., & Shiu, E. (2008). The impact of ethical concerns on family consumer decision-making. International Journal of Consumer Studies, 32, 553-560. Carter, C. R., & Rogers, D. S. (2008). A framework of sustainable supply chain management: Moving toward new theory. International Journal of Physical Distribution & Logistics Management, 38, 360-387. Chu, J. &, Chintagunta, P.K. (2011). An Emperical Test of Warranty Theroies in the U.S. Computer Server and Automobile Markets. Journal of Marketing, 75, 75-92. Clemons, E.K. (2008). How Information Changes Consumer Behavior and How Consumer Behavior Determines Corporate Strategy. Journal of Management Information Systems, 25(2), 13-40. Cohn & Wolfe. (2012). Transparency and Authenticity: A New Communications Landscape?. Corporate Affairs, 1-15. Cohn & Wolfe (2013). From Transparency to Full Disclosure?. Corporate affairs, 1-23. Coughlan, S. (2002). Easy money. The Guardian, 23 November. Dalen van, J., & Leede de, E. (2009). Statistisch onderzoek met SPSS voor Windows. Den Haag: Lemma Ford, G.T., Smith, D.B., & Swasey, J.L. (1990). Consumer Skepticism of Advertising Claims: Testing Hypotheses from Economics of Information. Journal of Consumer Research, 16(4), 433- 441.
  • 48. [48] Foster, L. (2004). Socially stylish. Drapers Record & Menswear, 22, 34-36. Golan, E., Krissof, B., Kuchler, F., Calvin, L., Nelson, K., & Price, G. (2004). Traceability in the U.S. food supply: Economic theory and industry studies. Agricultural Economic Report. Washington DC: U.S. Department of Agriculture, ERS. Hammond, D., Fong, G.T., McNeill, A., Borland, R., & Cummings, K.M. (2006). Effectiveness of cigarette warning labels in informing smokers about the risks of smoking: findings from the International Tobacco Control (ITC) Four Country Survey. Tobacco Control, 15(3), 19-25 Hayes, D. (2014). Process Documentation. May 2nd , 2014, Afhayes.com: www.afhayes.com Hayes, D. (2014). Frequently asked questions. June 12th , 2014, Afhayes.com: www.afhayes.com. Hickman, L. (2004). How green is your house? The Guardian, G2. 29 January, 2-3. Hofstede, G. (1983). The cultural relativity of organizational practices and theories. Journal of International Business Studies, 14, 75-89. Hultman, J., & Axelsson. B. (2007). Towards a typology of transparency for marketing management research. Industrial Marketing Management, 36(5), 627-635. Kim, D.J., Ferrin, D.L., & Rao, H.R. (2008). A trust-based consumer decision-making model in electric commerce: The role of trust, perceived risk, and their antecedents. Decision Support Systems, 44(2), 544-564.
  • 49. [49] Koslow, S. (2000). Can the Truth Hurt? How Honest and Persuasive Advertising Can Unintentionally Lead to Increased Consumer Skepticism. The Journal of Consumer Affairs, 34(2), 245-265. Lamming, R. C., Caldwell, N. D., Harrison, D. A., & Phillips, W. (2001). Transparency in supply relationships: Concept and practice. Journal of Supply Chain Management, 37(4), 4−10. Lee, L., Frederick, S., & Ariely, D. (2006). Try It, You'll Like It: The Influence of Expectation, Consumption, and Revelation on Preferences for Beer. Psychological Science, 17(12), 1054‐1058. Lee, M.S.W., Conroy, D., Motion, J. (2009). Brand Avoidance: A Negative Promises Perspective. Advances in Consumer Research, 36(1), 421-429. Loewenstein, G., Sunstein C.R. & Golman, R. (2013). Disclosure: Psychology Changes Everything. Annual Review of Economics, 1-33. MCDaniel, C., & Gates, R. (2012). Marketing Research. John Wiley & Sons, inc McNeal, J.U. (1987). Children as Consumers: Insights and Implications. Lexington, MA: Lexington Books. Mittal, B. (1988). The Role of Affective Choice Mode in the Consumer Purchase of Expressive products. Journal of Economic Psychology, 9, 499-524. Myers, T.A. (2011). HOTDECK: An SPSS Tool for Handling Missing data 1. Communication Methods and Measures, 5(4), 297-310.
  • 50. [50] Olsen, S.O. (2003). Understanding the relationship between age and seafood consumption: the mediating role of attitude, health involvement and convenience. Food Quality and Preference, 14(3), 199-209. Pardo, C. (2013). Entrepeneurial Risk Aversion, Net Worth Effects and Real Fluctuations. Review of Financial Economics, 22, 158-168. Pennings, J.M.E., Wansink, B., & Meulenberg, M.T.G. (2002). A note on modelling consumer reactions to a crisis: The case of the mad cow disease. International Journal of Research in Marketing, 19(1), 91-100. Reynolds, F. D., & W. R. Darden (1972). Why the Midi Failed. Journal of Advertising Research. 12(8), 39-46 Rijswijk, W. & Frewer, L.J. (2012). Consumer needs and requirements for food and ingredient traceability information. International Journal of Consumer Studies, 36, 282-290. Rust, R.T., Lemon, K.N., & Zeithaml, V.A. (2004). Return on Marketing: Using Customer Equity to Focus Marketing Strategy. Journal of Marketing, 68, 109-127. Saunders, M., & Lewis, P. (2012). Doing Research in Business & Management: An Essential Guide in Planning Your Project. Prentice Hall, Pearson Shimp, T.A., & Bearden, W.O. (1982). Warranty and Other Extrincic Cue Effects on Consumers’ Risk Perceptions. Journal of Consumer Research, 9(6), 38-46. Slavin, M. (2009). Commentary: Transparency increases firms’ credibility. Daily Journal of Commerce, 5.
  • 51. [51] Shroff, N., Sun, A.X., White, H.D., Zhang, W. (2005). Voluntary Disclosure and Information Asymmetry: Evidence from the 2005 Securities Offerings Reform. Journal of Accounting Research, 51(5), 1299-1345. Thrasher, J.F., Rousu, M.C., Hammond, D., Navarro, A., & Corrigan, J.R. (2011). Estimating the impact of pictorial health warnings and “plain” cigarette packaging: evidence from experimental auctions among adult smokers in the United States. Health Policy, 102(1), 8-41. Trouw. (2013). ‘Voor mijn baby alleen buitenlandse melk’. January 10th, 2013, Trouw.nl: http://www.trouw.nl/tr/nl/4496/Buitenland/article/detail/3411121/2013/03/18/Voor-mijn- baby-alleen-buitenlandse-melk.dhtml Unox. (2014). ‘Vragen over soep in pak’. April 10th, 2014. Unox.nl: http://www.unox.nl/nl/verder-op-unoxnl/vragen-over-soep-pak Weber, U., & Hsee, C. (1998). Cross-cultural Differences in Risk Perception, but Cross- cultural Similarities in Attitudes Towards Perceived Risks. Management Science, 44(9), 1205-1217. Weber, E.U., Blais, A-R., & Betz N.E. (2002). A Domain-specific Risk-attitude Scale: Measuring Risk Perceptions and Risk Behaviors. Journal of Behavioral Decision Making, 15, 263-290. Xiang, P., Zhou, J., Zhou, X., & Ye, K. (2012). Construction Project Risk Management Based on the View of Asymmetric Information. Journal of Construction Engineering & Management, 138(11), 1303-1311. Zhang, C., Bai, J., & Wahl, T.I. (2012). Consumer’s willingness to pay for traceable pork, milk, and cooking oil in Nanjing, China. Food Control 27, 21-28.
  • 53. [53] A. The actual experiment (treatment group) (*in reality this experiment was conducted in Dutch) Q1: I am a parent Yes / No *Dear respondent, thank you in advance for participating. Your participation is voluntary and you can opt out whenever you like. Your answers will be treated trustworthy. This research wants to find out if transparency of products matters in your purchasing behavior. I want to stress there are no bad answers, just be honest. This research will take 5 to 10 minutes. Q2: I would buy this product (5-point Likert scale) Manipulation 1 (high transparency) Q3: I would buy this product (5-point Likert scale) Manipulation 2 (high transparency) Q2: I would buy this product (5-point Likert scale) Manipulation1 (high transparency) Q3: I would buy this product (5-point Likert scale) Manipulation 2 (high transparency) Q7: I would continue buying the product repeatedly in the future (considering my children will consume them as well). (5-point Likert scale) Q5: I would buy this product (5-point Likert scale) Manipulation 3 (high transparency) Q5: I would buy this product (5-point Likert scale) Manipulation 3 (high transparency) Q8: Risk aversion scale. – See appendix B – Q4: Information needs scale. – See appendix C – Q6: I would buy this product (5-point Likert scale) Manipulation 4 (high transparency) Q4: Information needs scale. – See appendix C – Q6: I would buy this product (5-point Likert scale) Manipulation 4 (high transparency) Q7: I would continue buying the product repeatedly in the future. (5-point Likert scale) Q9 – Q12: Did you know about manipulation 1 to 4? Yes/No – See appendix C – Q9 – Q12: Did you know about manipulation 1 to 4? Yes/No – See appendix C – Q8: Risk aversion scale. – See appendix B – Read ‘low transparency’ instead of high transparency for control condition. Each 5-point Likert scale consists out of the following options: Firmly disagree / disagree / neutral / agree / firmly agree
  • 54. [54] B. Risk aversion Scale The risk aversion scale is derived from the research of Weber et al. (2002) on domain specific scales for risk aversion. The following Likert scale based questions were used for this research’ questions: (*in reality this experiment was conducted in Dutch) *Indicate the likelihood of engaging in each activity. Provide a rating from 1 to 5 using the following scale: 1. Extremely unlikely 2. Unlikely 3. Unsure 4. Likely 5. Extremely likely Eating ‘expired’ food products that still look okay. Frequent binge drinking. Ignoring some persistent physical pain by not going to the doctor. Taking a medical drug that has a high likelihood of negative side effects. Engaging in unprotected sex. Never using sunscreen when you sunbathe. Never wearing a seatbelt. Not having a smoke alarm in or outside of your bedroom. Regularly riding your bicycle without a helmet. Smoking a pack of cigarettes a day.
  • 55. [55] C. Information needs scale The information needs scale is derived from the research of Rijswijk & Frewer (2012) by modifying the categories of searching for information by consumers into a scale. The following Likert scale based questions were used for this research’ questions: (*in reality this experiment was conducted in Dutch) *Indicate the likelihood of engaging in each activity. Provide a rating from 1 to 5 using the following scale: 1. Extremely unlikely 2. Unlikely 3. Unsure 4. Likely 5. Extremely likely From now on I will search as much information as possible about products I buy (general statement). From now on I will contact points of sale to gain information about products (based on Rijswijk & Frewer). From now on I will contact the authorities to gain information about products (based on Rijswijk & Frewer). From now on I will ask people I know to provide me information about products I want to buy (based on Rijswijk & Frewer).
  • 56. [56] D. Revision letter 1 June 23, 2014 Joris Demmers, first supervisor Master thesis Amsterdam Business School – UVA Dear Joris Demmers, Hereby I send you my revised master thesis in its final version. I believe we had a good understanding, and I value your often quick and detailed feedback. Especially, the face-to-face sessions were of great help in completing this final manuscript. Below, you can find de headlines of your feedback on which I commented as well. Thanks for your support. 1. Your fist comment was about the writing style being to personal. - After your explanation in person I understood what you meant by this. I have replaced my personal writing style for an academic writing style, and removed my personal experiences from the thesis. 2. Secondly, you commented on the literature review. It was not coherent enough and lacked a certain “flow”. Other than this, you would appreciate stronger argumentation to the hypotheses and a more in dept analysis of the construct transparency. - I have revised the whole literature review and started off with a ‘writing plan’ as you suggested. I tried to combine all lose pieces of information by finding connections and used more compound sentences. I went back into the literature as well and found interesting additional argumentation for the dependent variables information need and purchase intent. Finally, I provided a more in dept view on transparency, were this
  • 57. [57] independent variable comes from, and how I operate this construct in my thesis. ‘Product transparency’ is the way in which the broad construct of transparency was operated. 3. The contributions in the thesis’ introduction lacked clear indications. - I have restructured the introduction and clearly mentioned the different contributions. 4. The methods section had to be revised in terms of building. - I have restructured the methods sections as you indicated in the detailed feedback. 5. The Data section had to be matched with an existing study, because it was not structured well and you did not like the lay out. Next to this, I initially found separate correlations for high – and low transparency which cannot be. - I have looked for an article with a comparable method and based my thesis’ lay out and structure on it. I have made a mistake in SPSS which resulted in separate correlations, but have revised this and created a single correlation matrix. 6. You do not believe involvement plays a moderating role and thought proper argumentation lacked for using this variable. - However, research by Carey et al. (2008) suggests parenthood does influence consumers’ choices. The Pearson correlations showed involvement had a negative influence on purchase intent, so choosing this variables was not all to strange after all. Eventually, involvement did not turn out to play an important role when looking at the regression analyses. Yours sincerely, Patrick Heeremans 10660623
  • 58. [58] E. Revision letter 2 August 12, 2014 Joris Demmers, first supervisor Master thesis Amsterdam Business School – UVA Dear Joris Demmers, Hereby I send you my revised master thesis in its final version. After receiving my grade I felt disappointed, but your feedback helped me revise my thesis in a good way. On top of this, I believe my thesis became a better whole. Thanks again for your support, invested time and effort. 1. Your fist comments were about the introduction. The proposed contributions did not exceed ‘never been researched before’. The relation between the different constructs was not clear. - I have explained the relationships between the constructs more clearly and used previous research to support my claims about a construct being an independent, dependent –or moderating variable. Furthermore, I have explained the contributions better and stressed the ‘never been researched before’ factor. 2. Secondly, you commented on the absence of argumentation and hypotheses for the moderator relations. - I have revised the literature review and looked for argumentation for the moderator relationships in the literature. Next to this, I have produced hypotheses for all separate moderator relationships.
  • 59. [59] 3. The argumentation of hypothesis 1 was not strong enough and only made intuitively sense. - I have followed your advice to look at the connection between information asymmetry and information need. 4. Hypothesis 2 lacked nuance. - I have added literature about the negative impact of negative transparency on purchase intent. 5. Argumentation for hypothesis 3 had to be sharpened. - I have read the article of Clemons (2008) again and rewrote my argumentation for hypothesis 3 based on this article. Now, the thesis also explains why transparency leads to customer loyalty and not only that it does. 6. The Research Question (RQ) had to be revised. - The RQ was revised in order to show the three dependent variables were separate relations. 7. There was no need for limitations in the methods section. - The limitations were deleted from the methods section and are only to be found in the limitations section of the thesis. 8. Argumentation for low Alpha scores had to be revised. - Argumentation was revised. 9. It was written ANOVA was used, but this was untrue. - I have mistaken to believe I did use ANOVA. After your explanation that my statistical analyses actually did not use ANOVA I deleted the announcement for using these.
  • 60. [60] 10. It was unclear why the results section used 3 separate models. - After testing the correlations of the dependent variables (DV) it was shown these were not correlated. Because of this it was right to use separate models instead of 1 model including all DV’s. An explanation of this was also added to the thesis itself. Yours sincerely, Patrick Heeremans 10660623