The effect of advertising on consumers: does it really make a difference?

(1)

The effect of advertising on consumers: does it really

make a difference?

The effect of advertising appeals on purchase intention and the roles of type of

products and need for closure.

(2)

1

The effect of advertising on consumers: does it really make a

difference?

The effect of advertising appeals on purchase intention and the roles of type of products and need for closure.

Master thesis

University of Groningen Faculty of Economics and Business MSc BA specialization Marketing Management

Author: Nora El Ouariachi

(3)

2

Management summary

The main goal of this study is to find out whether there is a difference in effect of a rational and emotional advertisement appeal on the purchase intention of the consumers and whether the type of the advertised products and the need for closure of the consumers influence this effect. According to existing literature, rational advertisement appeals and rational commercials lead to higher purchasing intentions than emotional advertisement appeals and emotional commercials. (Golden & Johnson, 1983; Coulson, 1989). The existing studies on utilitarian and hedonic type of products came to the conclusion that a rational appeal should be used for utilitarian (durable) products and that an emotional appeal should be used for (experimental, value-expressive) hedonic products (Johar & Sirgy, 1991; Fennis & Stroebe, 2010). Studies on the characteristic need for closure has shown us that high-need for closure consumers are more likely to search for price and promotional information. The reason for this is that price and promotional information can provide high-need for closure consumers an easy to use heuristic which helps to make quick and confident decisions (Vermeir & Kenhove, 2001). The method that is used for this study is a 2 (rational vs. emotional ad appeals) x 2 (utilitarian vs. hedonic products) x 2 (high vs. low need for closure) between subjects design. With an online survey platform, four different surveys were set up in order to collect data for this study. This resulted in 180 respondents. The problem statement of this study is:

What is the effect of emotional and rational advertisements on the consumers’ intention to purchase a product and how do the need for closure and the type of products influence this

relationship?

(4)

3

The main results showed that there is no difference in the effect of emotional and rational advertisements on the consumers’ intention to purchase a product. This relationship is also not influenced by the high (vs. low) need for closure of consumers. This personality trait does not change the consumers’ intention to purchase. The type of product on the other hand, does influence this relationship. A rational advertisement containing an utilitarian product has a positive effect on a consumers’ intention to purchase the product. The reason could be that consumers find the information on a rational appeal in combination with an utilitarian product more credible than in combination with a hedonic product.

The findings of this study suggest that a company should make use of rational advertisement appeals when promoting an utilitarian product. Consumers prefer useful information about an utilitarian product. This will lead to a higher purchase intention, compared to depicting an utilitarian product on an advertisement with an emotional appeal. Companies must know that the product type should determine the kind of appeal used for the advertisement.

(5)

4

Preface

The final step in order to receive my master in Business Administration - Marketing Management has come to its end. With certainty I can say that the past couple of years of my study at the University of Groningen has been very instructive for me, more than I could ever have imagined. Although it took me longer than planned to finalize my thesis because of my job and other distractions, I still had a fun and informative time writing this thesis. My interest in the world of advertising has increased and made me curious to find out more about this subject. The time has now come to leave my relaxing student time behind and start to make a contribution to the real world with the knowledge that I have gained in the past couple of years.

I would like to give special thanks to Dr. Liu, who was there for me from the beginning till the end, supporting me and was always ready to advise me and discuss the problems with me that I encountered in writing my thesis. I would also like to thank Mr. Wiebenga for his time to read my thesis and provide it with his feedback. Finally, I would like to thank my family and friends who supported me and talked me trough the hardships I encountered during the writing of this thesis.

(6)

5

1. Introduction

Doctors without borders, an international charity organization uses different campaigns to create awareness under customers. In almost all of their campaigns they make use of emotional appeals. Often a sad little child is shown in order to create affection in the consumers’ minds. This is often a reason for consumers to make donations. The advertisement on this page shows a real-life example of a charity that uses this strategy. This advertisement uses a guilt appeal that should encourage people to make a donation.

There is also another appeal that organizations can make use of in their campaigns to create awareness for their products or services. This appeal is called a rational appeal. Advertisements with rational appeals are based on arguments and information about the advertised products. Both rational and emotional appeals are in there own ways effective, however this also depends on the type of products that are advertised and the level of involvement of the audience.

Previous research has already shown that the overall attitude towards an ad in which a brand is advertised will influence consumers’ brand

(10)

9

conscious processing of web ads, not only the implicit but also het explicit memory performances are increased. In contrast to the previous, if consumers pay minimal or no attention to the web ads, only the implicit memory performance is increased. The study also reveals a positive relationship between the implicit memory performance and the brand attitude. This would mean that advertisements un-consciously have an effect on the attitude and thus behavior of the consumer towards the brand.

In a study about advertising by Defever, Pandelaere & Roe (2011) they study the effect of advertising on value-laden behavior that is unrelated to the advertising context. They explored the possibility that value-laden advertising may activate the ‘’advertised’’ values and influence decisions that transcend the advertising context. The results are that value-laden advertising may affect behavior of the ones who are exposed to it. Participants of this experiment chose the direction option more after being exposed to ads featuring self-direction than after being exposed to ads featuring security. In another experiment, participants persisted longer in a puzzle solving task after being exposed to ads featuring achievement than after being exposed to ads featuring benevolence. However, participants showed more interest in donating blood after being exposed to ads featuring benevolence than after exposure to ads featuring achievement. This indicates again that advertising has a certain effect on consumers, in this case on the behavior of the consumer.

(11)

10

1.1 Problem statement

Although recent research presents lots of studies regarding the effect that advertisements have on brand attitudes or consumers’ behavior, there has been little thorough investigation on the effect of emotional versus rational ads on consumers’ intentions to purchase a product. Thus in this research, I will try to fill this gap by studying this relationship that will be moderated by the need for closure and the type of products, using an unknown brand name. I choose to use an unknown brand name in order to exclude the possibility that a known brand name could moderate this effect. Hence, the problem statement of this study is formulated as follows:

‘‘What is the effect of emotional and rational advertisements on the consumers’ intention to purchase a product and how do the need for closure and the type of products influence this relationship?’’

1.2 Research questions

The problem statement will be translated into three research questions in order to come with a specific answer to the problem statement:

1): How do emotional advertising and rational advertising influence purchase intention? 2): How is this effect moderated by the need for closure

3): How is this effect moderated by the type of products?

(12)

11

1.3 Framework of the conceptual model

Figure 1 illustrates the conceptual model which shows the relationships of the defined problem statement that is about to be investigated.

Figure 1: framework of the conceptual model

(13)

12

2010). In this study I want to know which appeal of advertisements will have a higher effect on the dependent variable. The dependent variable is: purchase intention. Beerli & Santana (1999) explain this as: “the intention to purchase, quantifies the probability that an individual will try out or buy the advertised brand in the near future as a result of exposure to the advertisement”. At last the model shows two moderators. The first moderator is: type of products. This reflects the different type of products that will be used which may have an effect on consumers’ purchase intention. The second and last moderator that is likely to moderate the influence of advertising on the purchase intention is: need for closure. This is a longing for clear knowledge on a certain issue which represents a dimension of stable individual differences and a situational evocable state (Kruglanski & Webster 1996).

1.4 Managerial & academic relevancy

This investigation is a contribution to the academic literature by studying the impact of advertising on purchase intention. The type of products will be moderating the purchase intention as will the need for closure be moderating. In short, this study will give insight into which kind of advertisement will lead to the consumer’s intention to purchase the product. This is in contrast to the existing relevant studies which mainly focus on the relationship of advertisements to attitudes or the purchase behavior of a consumer towards a brand.

This study has a great contribution to managerial implications if the hypotheses are successfully accepted. For marketers this would mean that they have more control on influencing purchase intentions of their products by providing ads with emotional or rational appeals related to their products. The consumers’ purchase intention may also depend on the type of products that are sold and it could also depend the consumers’ characteristic need for closure. By stimulating these two variables the purchase intention could increase

while consumers would not be conscious about this.

(14)

13

1.5 Structure of the paper

(15)

14

2. Theoretical framework

2.1 Introduction

In the following section, the main reasoning underlying the problem statement as described in chapter 1, as well as the description of the conceptual model, will be elaborated. Finally the resulting hypotheses will be described.

2.1.1 Placebo effects by advertisements

Not only in marketing do placebo effects exist but a common known fact is the role of placebo effects in medicine. An interesting question is how this placebo effect takes place in the consumer’s brain. Berns (2005) gives an explanation in his research article regarding this subject. He states that functional magnetic resonance imaging (FMRI) starts to provide a framework into how these processes function in the brain. FMRI depends on the empirical observation that the human blood flow increases to active parts of the brain. Many researchers have focused on the effect that placebo has on lessening painful stimuli through suggestion. Berns (2005) illustrates a study conducted by Wager et al. (2004) wherein participants were given a skin cream, but were told that it was an analgesic cream. The activity that was measured in the same pain regions was less in a group that (after salving the skin cream) received a painful stimulus than in a control group that did not receive the placebo treatment of the skin cream. Berns (2005) then discusses that this has to do with the placebo treatment being associated with increased activity in both the left and the right prefrontal cortex in anticipation of the pain. He states this being an important aspect because the more dorsal and lateral aspects of the prefrontal cortex have consistently been associated with the working memory. The working memory is the part that keeps pieces of information at the forefront of attention (Miller & Cohen 2001). As described previously this placebo effect that operates through conscious memory is mediated by the prefrontal function of the brain (Berns 2005). This is in contrast to the study of Shiv, Carmon & Ariely (2005) who suggest that the placebo effect is also able to operate through non-conscious memory.

(16)

15

Deshpandé 2001; Whittler & Spira 2002). Some of these researchers have studied how values in ads affect self-perceptions. Research of Nairn & Berthon (2003) has shown that certain values lead to certain self-perceptions. In this study, participants were shown different kinds of advertisements. One group of participants was shown a romantic advertisement, another group viewed ads which had a classical character (unemotional and unadorned). The first group rated themselves as more romantic than the group who viewed ads with a classical character. However, this study did not measure the actual behavior. All these different researches regarding placebo effects in marketing show that the phenomenon placebo effect exists. In this research I would like to know whether emotional and rational advertising lead to the consumer’s intention to purchase a product with the need for closure and the type of products as moderators.

2.2 Emotional & rational advertising

Advertising is a medium that uses values to render advertised products or services appealing to certain consumer segments (Defever, Pandelaere & Roe 2011). Advertisements can have different forms of appeals. One appeal is based on arguments while the other appeal is based on feelings and emotions shown in the advertisement to clarify the message (Fennis & Stroebe (2010). These appeals are called argument-based appeals and affect-based appeals. They are also known as informational or rational appeals and emotional appeals. In my study I will use the designations called rational and emotional appeals.

2.2.1 Product types

(17)

16

According to Johar & Sirgy (1991) an emotional appeal should be used for ‘value-expressive’ products and a rational appeal should be used for utilitarian products. Utilitarian products refer to practical, functional consumer needs, as opposed to value-expressive’ products that are directed mainly to psychological aspects of these needs (Bov’ee & Arens, 1992). According to Leonidou & Leonidou (2009) this implies that one need can be approached in two different ways, either with an emotional appeal or with a rational appeal.

2.2.2 Individual types

Different types of individuals are attracted to one of these appeals. it is interesting to know what types of individuals are likely to be influenced with emotional appeals and what types are likely to be influenced with rational appeals. The reasoning of Venkatraman et al. (1990) shows that individuals who enjoy to be engaged in effortful cognitive activities and enjoy the thinking process should be influenced more by rational appeals than by emotional appeals. On the other hand, individuals who rather only think as hard as they need to, might prefer emotional appeals rather than rational appeals. Schiffman & Kanuk (2004) imply that rational appeals will be better accepted by consumers whose purchasing decisions are guided more by information, logic and facts, while emotional appeals attract people who appreciate facts less, but have a higher emotional involvement. When need for cognition of the individuals increases, more prosperous judgments are formed towards rational appeals. In contrast, the need for cognition does not have an effect on judgments of emotional advertising. The message appeal moderates the effects of need for cognition on the judgments that individuals have on messages.

2.2.3 Emotional versus rational advertising appeals

(18)

17

Stafford 1999). Kotler & Armstrong (1994) state that: “Emotional appeals attempt to stir up either negative or positive emotions that can motivate purchase. These include fear, guilt and shame appeals that get people to do things they should, or stop doing things they shouldn’t”. They also state that communicators use positive emotional appeals such as love, humor, pride and joy.

2.2.3.1 Positive effects of both appeals

Scholars of marketing and advertising literature have discussed both approaches and their effectiveness. Albers-Miller & Stafford (1999) summed up some results of different researches which all conclude that rational appeals resulted in positive effects. For example, Golden & Johnson (1983) concluded that rational ads were better liked and resulted in higher purchase intentions. Similarly Aaker & Norris (1982) found that rational appeals showed higher effectiveness ratings than emotional appeals. Also, Coulson (1989) found higher purchase intentions for rational commercials than for emotional commercials. Holbrook (1978) found that rational appeals, or like he named it, factual content is more credible which results in more positive beliefs. Finally, Stafford & Day (1995) found that rational appeals help reduce some uncertainty that is often associated with the purchase of services.

However, Albers-Miller & Stanford (1999) not only summed up researches with positive results of rational appeals, they also came with a list of scholars who came with positive results regarding emotional appeals. For example, Unwin (1975) argues that emotional appeals can help lighten the abstract nature of services. Goldberg & Gom (1987) and Page et al. (1990) found that appeals that generate emotional responses will result in more positive reactions. Finally, according to Choi & Thorson (1983) emotional responses will also result in higher levels of recall.

2.2.4 Differences between emotional and rational advertising

(19)

‘value-18

expressive’ products and a rational appeal should be used for utilitarian products (Johar & Sirgy, 1991) as discussed earlier. The cultural setting of a country also plays an important role when applying a certain appeal in advertisements. Rationality and emotions are not always perceived as opposites in all cultures (Hofstede, 1980). Also, almost every culture has different emotion expressions and they vary in the intensity of the emotions. Research of Leonidou & Leonidou (2009) shows us that there is also a difference in the effects both appeals have on consumer recall. Rational appeals are often described as being uninteresting and dull, although this does not have a negative effect on the recall rates (Chan, 1996). The consumer recall is achieved by the facts and information provided which touches the self interest of the consumers. Emotional appeals on the other hand, attract consumer’s attention by inciting strong feelings. It creates more vivid memories for positive/friendly brands (Moore, Harris & Chen 1995). Emotional appeals are more easily recalled when compared to rational appeals. This is especially the case if the emotional appeal exists of strong, vivid content (Costley & Brucks, 1992). However, having a high consumer recall does not result in the persuasion of buying the product. This applies to both rational and emotional appeals. According to Schiffman & Kanuk (2004), the persuasion to buy a product depends on a number of factors, such as the targeted audience and their degree of involvement with the product. Zinn & Manfredo (2000) come with a general rule that the degree of persuasion of rational appeals depends on the consistency of the arguments with the beliefs that the consumer already has. For emotional appeals, this is determined by the degree of relevancy of the message content to the consumer. Leonidou & Leonidou (2009) add that the persuasiveness of an emotional appeal can also be influenced by the level of emotional intensity it shows. For instance, in the case of fear appeals, the stronger the fear that is created, the lower the effectiveness of the message is. Table 1 on page 19 shows a compact summary of the difference between emotional and rational appeals that is composed by Leodinou & Leonidou (2009).

(20)

19

Table 1: Difference between emotional and rational appeals (Leonidou & Leonidou 2009)

Comparison base Rational advertising appeals Emotional advertising appeals

Stimulating Mechanism

Focusing on product features and/or measurable benefits or reasons for selecting a particular product or brand, based on a rational thinking process.

Stirring up positive (e.g., joy, love, pride) or negative (e.g., fear, anger, shame) emotions that will motivate purchase through an affective response.

Consumer Needs Addressed

Targeting mainly practical, functional needs for a product or service, such as economy in purchase, dependability of quality, and efficiency in operation.

Targeting mainly consumer’s psychological, social, or symbolic needs, as in the case of obtaining sexual attraction, social approval, or pride of possession.

Type of Target Audience

Affecting consumers who better appreciate logic, information, and facts in making their purchasing decisions.

Appealing to consumers who have less appreciation of factual information, and higher emotional involvement.

National Cultural Context

Working better in high cultural contexts, where it is important to use direct speech to highlight clearly, logically, and reasonably the merits of the product.

Working better in low-context cultures, since direct speech is considered argumentative and annoying, compared to the smoother effect of emotional elicitation.

Products Advertised

Applying more to utilitarian products, such as

high-technology consumer products and industrial goods.

Applying more to

“value-expressive” products, such as convenience products and specialty goods.

Information Transmitted

Offering more information, consisting of objective statements that can be verified by consumers independently, thus making the advertisement credible.

Containing less information, that is mainly subjective and thus open to individual consumer

interpretation.

Consumer Recall Attracting consumer attention and recall through the factual

information provided, often described as dull and uninteresting.

Attracting consumer attention by greater appeal and interest, leading to more enhanced brand memory, especially if the content is vivid.

Persuasive Effectiveness

Achieving better persuasiveness when eliciting thoughtful comparison of arguments by appealing to beliefs the recipient already holds.

(21)

20

2.3 Purchase intention

The dependent variable of this research is purchase intention. Purchase intention as the words already say it is the intention to purchase a good or service. Or as explained by Beerli & Santana (1999): “the intention to purchase quantifies the probability that an individual will try out or buy the advertised brand in the near future as a result of exposure to the advertisement”.

2.3.1 Two streams in literature

(22)

21

measured just prior to a purchase (‘memory-based recency’ stream) are more likely to be more diagnostic of the actual actions of a consumer because this seems to reflect a larger amount of information than the ‘preference based fluency’ in which the intentions are measured in advance of the purchase.

2.3.2 Product selections

The product selections of consumers are influenced by different factors. Barber, Kuo, Bishop & Goodman jr (2012) state that the following factors are of influence during a consumers’ brand and product selection: habits, convenience, value, personal health concerns and consumer’s responses to institutional and social norms. Torgler, Garcia-Valinas & Macintyre (2008) argue that nowadays consumers are environmentally conscious and therefore they not only base their purchase decisions on how the products satisfy their needs, but they also want to minimize the negative impact on the natural environment. This is a result of increased concerns regarding the environment. These environmentally conscious consumers value the positive and ethical images that are projected by marketers and even adjust their purchase behavior in such a way that benefits those businesses (Creyer, 1997).

2.3.3 Advertisements leading to purchase intention

(23)

22

Although there is lots of research regarding ads and purchase intentions, I found little relevant information from academic papers that investigate if advertisements with emotional appeals or rational appeals lead to purchase intentions. However, based on the literature review of the independent variables and the dependant variable I can assume that an advertisement with a rational appeal will lead to a higher purchase intention than an advertisement with an emotional appeal. I will therefore investigate if this indeed has a positive effect. Of course this also depends on the moderators which are the type of products and need for closure. This is in contrast to the existing relevant studies which mainly focus on the relationship of advertisements to attitudes or the purchase behavior of a consumer towards a brand. The relationship that will be investigated is presented as hypothesis 1 which will be formulated as follows:

H1) Advertisements with a rational appeal will lead to a higher purchase intention compared to advertisements with an emotional appeal.

2.4 Type of products

Type of products could play a great role in my investigation of rational and emotional appeals’ effect on purchase intention. I therefore chose this variable as a moderator for my research. This moderator is defined as a reflection of different type of products that will be used which may moderate the effect on consumers’ purchase intention.

2.4.1 Type of products as moderators

(24)

23

2.4.2 Search vs. experience products

Moon, Chadee & Tikoo (2008) investigated if the type of products which are divided in to search attributes and experience attributes could have an influence on consumers’ purchase intention in an online retail format. They argue that a great limitation for online shoppers is that they cannot experience a product physically and that therefore search attributes are more appropriate for selling products online. This is also confirmed by Citrin, Stem & Spangenberg (2003) and Klein (1998). Moon, Chadee & Tikoo (2008) concluded that marketers therefore should show more effort to make buyers more comfortable with buying experience goods online.

(25)

24

that hedonic product use, is therefore a subjective symbol rather than an objective substance. A product can have both utility and hedonic characteristics because each individual gives its own weight of importance to a product. This is based on the individual’s subjective judgment regarding the value of a product. Thus, a consumer decides for itself if the product is utilitarian or hedonic (Park & Moon, 2003).

Drolet, Williams & Lau-Gesk (2007) discussed researchers from the past that started analyzing product categories in terms of hedonic vs. utilitarian characteristics (Ratchford, 1987). They suggested that the hedonic vs. utilitarian distinction can make emotions more or less relevant in evaluations (Pham, 1998). According to the multi-component theory of attitudes, consumers in general prefer advertisement appeals that are compatible with the attitude of the object type (Drolet & Aaker, 2002). In other words, consumers prefer rational advertisements for utilitarian products and emotional advertisements for hedonic products.

Based on the above mentioned researches, it is interesting to know if and how the type of products moderate the main effect (H1). There has been little research about this moderating effect on emotional vs. rational advertisement appeals on purchase intention. The division of search attributes vs. experience attributes is not relevant for me since I am not investigating the purchase intention in online stores. However dividing the product type into utilitarian and hedonic products is far more interesting since consumers prefer rational ads for utilitarian products and emotional advertisements for hedonic products. Also, this is the most common division of product types in marketing research. Therefore my hypothesis will be offered as:

(26)

25

2.5 Need for closure

Need for closure is explained as the desire of an individual for a firm answer to a question (Kruglanksi & Webster 1996) rather than uncertainty, confusion or ambiguity (Kruglanski 1990). According to Pervin (1989) the need for cognitive closure is related to the goal a person has. Kruglanksi & Webster (1996) explain that it may encourage activities that are aimed at the realization of closure, prejudice a person’s choices and preferences toward closure-binding pursuits. It may also cause a negative affect when the closure is threatened or undermined, and a positive affect when the closure is facilitated or achieved. The need for closure varies across individuals and situations (Shah, Kruglanksi & Thompson 1998). An example is that need for closure increases under different circumstances that make the process of information difficult or unpleasant such as time pressure (Kruglanski & Webster, 1991) and noise (Kruglanski, Webster & Klem, 1993). According to Shah, Kruglanski & Thompson (1998) the strength of this desire lies in the benefits that are associated with possessing closure and the costs that are associated with a lack of closure.

Understanding the purchase decision-making process of customers who are confronted with different advertisement appeals is important. Need for closure could play an important influencing role in this process. Many studies have already shown that need for closure has a significant impact on the decision-making process of consumers (e.g. Vermeir & Van Kenhove 2001).

2.5.1 High (vs. low) need for closure

Need for closure is a longing for clear knowledge on a certain issue which represents a dimension of stable individual differences and a situational evocable state (Kruglanski & Webster 1996). Individuals with a high (vs. low) need for closure are known for two tendencies: the urgency that inclines them to grab quickly on readily available explanations and to reject alternatives, and the permanence that inclines them to persist on the explanations and stay with them even if it is proven otherwise (Tetlock 1998).

(27)

26

high need for closure (Kruglanski & Webster, 1991) which means that this desire for having confident knowledge would not be sustained without the group members. Individuals with a high need for closure should value their in-group members and should be able to identify with them and will therefore foster a distance with the out-group (Shah, Kruglanski & Thompson, 1998).

Vermeir & Van Kenhove (2001) studied the moderating effect of need for closure on search effort for price and promotional information. They found that high-need for closure consumers are more likely to search for price and promotional information. A reason is that the price and promotional information can provide high-need for closure consumers an easy to use heuristic which helps to make quick and confident decisions (Vermeir & Van Kenhove, 2001).

It is very interesting to have a psychological characteristic as need for closure serving as a moderator. This characteristic helps to predict and understand consumers decisions in a shopping setting. As earlier mentioned, need for closure has shown in many studies that it has a significant impact on the decision-making process of consumers. Taking the above mentioned research into account, it can be assumed that high-need for closure consumers appreciate more factual content which will therefore lead to a higher purchase intention than it will for consumers with low need for closure. My hypothesis will therefore be offered as:

(28)

27

2.6 Final conceptual model

After thoroughly analyzing the existing literature and defining the hypotheses, the conceptual model can be finalized. The final conceptual model and the hypotheses are shown below.

Final conceptual model

H1) Advertisements with a rational appeal will lead to a higher purchase intention

compared to advertisements with an emotional appeal.

H2) The positive relationship between advertisements with a rational appeal and purchase

intention is higher with utilitarian products compared to rational ad appeals with hedonic products.

H3) The positive relationship between a rational ad appeal and purchase intention is

(29)

28

3. Methodology

3.1 Introduction

In this chapter the methodology is explained which describes the study in detail. First, the research design of the study will be explained, then the pre-test will be discussed. In the following paragraphs the participants, the variables, the procedure and the analysis will be explained.

3.2 Research design

The research design of this study is a 2 x 2 x 2 between-subjects design. Table 2 gives a clear overview of the research design. A between subject design means that the participants of the questionnaire are randomly assigned to one of the existing conditions.

High need for closure Low need for closure

Utilitarian product Hedonic product Utilitarian product Hedonic product Rational appeal Condition 1 Condition 2 Condition 3 Condition 4

Emotional appeal Condition 5 Condition 6 Condition 7 Condition 8

Table 2: research design

As shown in table 2 the research design shows that 8 different conditions are present in this study. Each condition requires 20 respondents which makes a total of 160 respondents. However, there are only 4 different questionnaires composed with 4 different advertisements containing either an emotional or rational appeal with either an utility or hedonic product. The moderating variable need for closure entails characteristics of the respondents, which naturally cannot be manipulated. This variable is measured with a list of questions based on the research of Kruglanski, Webster & Klem (1993). The 4 different advertisements will be used in the questionnaire as shown in table 3. Respondents will fill in the questionnaire which contains either advertisement 1, 2, 3 or 4.

Utilitarian product Hedonic product

Rational appeal Ad 1 Ad 3

Emotional appeal Ad 2 Ad 4

(30)

29

3.3 Pre-test

In order to find out what consumers’ reactions are on the advertisements with different appeals combined with different products, four advertisements are created, each representing one of the four conditions. In order to find out which products and which appeals represent the right condition a pre-test is held. This pre-test contains 8 different advertisements. Three different measurements are tested in this study. First, the equality in likability of the ads is measured. It is important that the likability of the ads is equal to be sure that the likability of an ad will not bias the results of the test. Second, the rationality and emotional appeal of the ad is measured to be sure that the ads are seen as either emotional or rational. This measurement is based on Liu & Stout (1987). The ads were manipulated by adding only information and facts to the rational ads, like for instance for the Diet coke ad, a list of daily values of calories, total fat, sodium and protein is listed. The emotional juicy water ad was manipulated by adding ingredient information. A clear and detailed overview of the manipulated texts that are used for the ads can be found in appendix 2.1. The third measurement is to find out whether the depicted products are seen as utilitarian products or as hedonic products. This measurement is based on the study of Vaughn (1986). Table 4 shows a clear overview of the pre-test questions in detail.

Measurements Questions and answering scales Likability ad I think this advertisement is very clear,

I think this advertisement looks beautiful

I think this product has a strong persuasion to buy the product that is shown I like this advertisement

Strongly disagree 0 0 0 0 0 strongly agree Ad appeal I think this advertisement is:

Tangible 0 0 0 0 0 0 0 Intangible Logical 0 0 0 0 0 0 0 Emotional Objective 0 0 0 0 0 0 0 Subjective Factual 0 0 0 0 0 0 0 Non-factual

Product type The buying decision-making of this product is made logically and objectively The buying decision-making of this product is primarily based on functional perspective

The buying decision-making of this product is made primarily based on feeling The buying decision-making of this product reflects my personality

The buying decision-making of this product is made based on appearance, taste, touch, smell or sound

Strongly disagree 0 0 0 0 0 strongly agree

(31)

30

3.3.1 chosen products for ads

In creating the advertisements I chose typically hedonic and utilitarian products with either a non-familiar brand or a brand that was not clearly visible in the advertisement. All four products of my pre-test are from the soft drink category. The reason I chose this category is because there are different kinds of drinks with both hedonic and utilitarian drinks to choose from.

The first two ads contain a drink product called “juicy water”. Juicy water is a soft drink based on water with added fruit extracts and sugar. It does not claim to be good for one’s health nor does it claim that it contains vitamins. Juicy water states to be a tasty and refreshing drink. People most probably buy this drink to make them feel good about themselves. Because the drink contains fruit, they probably think that the drink will be good for their health. As Park & Moon (2003) state in their study, hedonic products are known for self-expression or pleasure. Therefore Juicy water falls under the hedonic product types. Also, utilitarian products are products which are known for their function according to Park and Moon (2003). Since Juicy water does not have nor does it state to have a function, it is a hedonic product type that has become a trend to drink nowadays. The second two ads contain a drink product called “Diet Coke”. Although this is a well known brand amongst others in the United States, the brand is not clearly visible in the advertisement. This bottle of coke is also a typical hedonic product. Two other ads contain a pack of coffee of the unknown brand “Lako”. Coffee is seen as a typical utility product. According to Shavitt (1989) coffee serves primarily utilitarian purposes of providing a pleasant taste and increased alertness. The last two ads depict a bottle of water of an unknown brand called “Voss”. Water is without a doubt an utility product. Of each product, two ads are designed. One ad shows the product consisting of rational texts about the product. The other ad shows the same product consisting of emotional texts or images.

(32)

31

3.3.2 Reliability check

After converting the data into SPSS and renaming the variables in order to get a better overview, the variables measuring the likability of each ad are combined into a new variable after checking the Cronbach’s Alpha. Of each ad there are now two variables measuring the likability. One for the emotional ad and one for the rational ad. Each variable has a sufficient Cronbach’s Alpha of at least 0,6. This is also the case for the variables measuring the ad appeal and the product type. Not all questions measuring the product type could be combined into one variable, so only the questions that could be combined are used.

3.3.3 Results

(33)

32

Table 5: overview results of ad likability

A paired samples t-test was also conducted to compare if the appeals of the ads differed significantly from each other in rationality and emotionality. Table 6 shows an overview of the results. There was a significant difference in the scores for Juicy water rational ad (M=3,0403, SD=1,24849) and Juicy water emotional ad (M=4,8952, SD=1,21411) conditions; t(30)=-4,862, P=0,000. There was also a significant difference in the scores for Diet Coke rational ad (M=2,7258, SD=1,21510) and Diet Coke emotional ad (M=5,2097, SD=1,33401) conditions; t(30)=-6,538, P=0,000. The significant difference in the scores for Lako Coffee were for the rational ad (M=2,8629, SD=1,13255) and for the emotional ad (M=5,0403, SD=1,41583) conditions; t(30)=-6,048, P=0,000. There was also a significant difference in the scores for Voss Water rational ad (M=2,6935, SD=1,15778) and Voss Water emotional ad (M=5,1613, SD=1,43694) conditions; t(30)=-5,779, P=0,000. All the ads differed significantly from each other in appeal. This means that the ads that were originally made in an emotional setting, were seen as emotional ads and the ads that were made in a rational setting were seen as rational ads. Appendix 2 shows the results in detail.

(34)

33 Paired samples t-test Measurement ad appeal Mean Std. deviation Sig. (2-tailed) Ads Juicy water -1,85484 2,12404 0,000 Ads Diet coke -2,48387 2,11536 0,000 Ads Lako coffee -2,17742 2,00436 0,000 Ads Voss water -2,46774 2,37762 0,000

Table 6: overview of results of the ad appeal

Table 7 shows the results of the paired samples t-test that is conducted to compare if the products are seen as either hedonic or utilitarian. There was not a significant difference between Lako coffee seen as a hedonic product (M=3,03, SD=1,110) and Lako Coffee seen as an utilitarian product (M=3,29, SD=0,824) conditions; t(30)=1,114, P=0,274. This means that Lako Coffee is seen as both an utilitarian product and a hedonic product. The other product that I chose as a typical utility product is Voss water. This product is in the eyes of the respondents more an utility product than a hedonic product. The difference was significant for Voss Water seen as a hedonic product (M=3,03, SD=1,080) and Voss Water seen as a utility product (M=3,55, SD=0,810) conditions; t(30)=2,380, P=0,024. The other two typical hedonic products are also seen more as hedonic products than as utility products. There was a significant difference between Juicy Water seen as a hedonic product (M=3,52,

SD=0,851) and Juicy Water seen as a utility product (M=2,68, SD=0,909) conditions; t(30)=-3,763, P=0,001. And there was also a significant difference between Diet

(35)

34

Paired samples t-test

Measurement product type

Mean Std. deviation Sig. (2-tailed) Ads Juicy water -0,839 1,241 0,001 Ads Diet coke -0,581 1,501 0,039 Ads Lako coffee 0,258 1,290 0,274 Ads Voss water 0,516 1,208 0,024

Table 7: overview results of difference in product types

3.3.4 Choice of advertisements

(36)

35

3.4 Participants

Nowadays advertisements can be seen and found anywhere. People are always confronted with campaigns and promotions no matter where they are. My selection for participants is therefore not very hard. Anyone of at least 18 years or older who can either read Dutch or English can participate in the study. In order to reach at least a sample of 160 respondents, the questionnaire will be placed on an online platform called Amazon Mechanical Turk. This will get more variety in the demographics of respondents and will get me more results faster. The questionnaire will also be send around to friends, family and other contacts, who are asked to forward the questionnaire to their own contacts in order to make my sample more diversified. The sample should consist of at least 160 respondents because the study consists of eight conditions of which each conditions should have a sample group of 20 participants. This is needed in order to have a sufficient amount of data that is reliable.

3.5 Variables

The measurement of the variables of this study are described in this paragraph. The independent variable, dependent variable and moderators are explained on how they are measured in the questionnaire.

3.5.1 Independent variable: rational and emotional appeals

(37)

36

rational advertisements and that the emotional advertisements are recognized as emotional advertisements.

3.5.2 Dependent variable: purchase intention

The purchase intention of the products shown in the advertisements will be measured in order to find out which advertisement appeal, which type of product and which need for closure (high or low) will lead to a higher purchase intention. The measurement of purchase intention consists of 4 questions and is based on the study of Barber et al (2012). The participants will first get to see the advertisement and will be asked to take a good look at the advertisement before answering the following questions. The purchase intention measurement will immediately be asked after participants have seen the advertisement. The questions are: “I would consider purchasing this product”, “I intend to try this product, “I plan on buying this product” and “I am interested in tasting this product”. The 7 point likert answering scale ranges from “strongly disagree” to “strongly agree”.

3.5.3 Moderator 1: Type of products

The type of products used in this study are divided in two different products: hedonic and utilitarian products. In the pre-test several products were tested on hedonistic and utilitarian aspects. The products based on this measurement are already chosen in the pre-test results. A bottle of water is used as an utility product and a bottle of diet-coke is used as a hedonic product in the study. This measurement will be used as a manipulation check in the questionnaire. This manipulation check consists of 5 question items and is based on the study of Vaughn (1986). First the respondents will be told that the objective of the following statements is to determine the criteria used in their decision-making regarding buying the product that is shown in the advertisement. The sentence: “The buying decision-making of this product” will be stated and after this sentence five statements will be given (see table 5). A 7 point Likert answering scale will be ranging from “strongly disagree” to “strongly agree”.

3.5.4 Moderator 2: Need for closure

(38)

37

Kruglanski, Webster & Klem (1993) consists of 42 questions, for this study I will use a valid short version of the NFC scale in order to keep a clear and plain questionnaire. The short version of this scale is based on the study of Roets & Van Hiel (2011). A list of 15 statements will be answered on a 7 point Likert scale which ranges from “strongly disagree” to “strongly agree”. Statements like: “I don’t like situations that are uncertain” and “I dislike questions which could be answered in many different ways” are stated. An overview of the shortened NFC scale measurement can be found in appendix 2.3.

3.5.5 Brand familiarity

In order to check if the brands of the advertised products that are used in this study will or will not have an effect on the results, a brand familiarity measurement based on the study of Kent & Allen (1994) will be included in the survey as a control variable. Respondents will have to fill in a 3 item scale: “Regarding the product Voss Water/Diet Coke, are you:” the answering scale is a 7 point bipolar format with the following statements: familiar/unfamiliar, experienced/inexperienced and knowledgeable/not knowledgeable.

3.6 Procedure

(39)

38

check will be asked. When the survey is finished, the participants will be thanked for their participation.

3.7 Analysis

(40)

39

4. Results

This chapter shows the gathered data of the survey and its results. First, the demographic variables will be analyzed which will describe the total sample of the survey. Second, The manipulation checks will be analyzed. Also a reliability analysis will be performed on the survey questions. Then the hypotheses will be tested and the results will be discussed. Finally, the control variable will be elaborated upon.

4.1 Sample characteristics

As mentioned earlier, the site qualtrics.com is used to create the online survey. Next to using the snowball technique for distribution, the survey was also placed on Amazon Mechanical Turk. About half of the sample exists of Amazon Mechanical Turk users, the other half exists of my network. A total of 180 respondents are counted. This is more than the minimum required participants of 160. Since the participants were forced to answer all the questions of the survey, all surveys were complete. This section describes the demographic variables of the total sample and next to this, the demographic variables per condition are also described.

4.1.1 Total sample demographics

(41)

40

Graphic 1: distribution males and females Graphic 2: distribution age

Chart 1: distribution level of education Chart 2: distribution spending income

4.1.2 Demographics per condition

Next to describing the demographics of the total sample, the demographics per condition will also be described. Participants filled in one of the four available surveys. Each survey had the same questions, only the advertisements differed. The demographics should be about equal to the demographics of the total sample. Appendix 3.3 shows a table which gives a detailed overview of all demographics per condition. None of the demographics show remarkable differences. The demographics do not seem to differ across the conditions.

(42)

41

4.2 Manipulation checks

The survey consists of 2 manipulation checks. The first manipulation check is for the independent variable ‘ad appeal’. Participants got to see one of the four manipulated ads, of which two ads have an emotional appeal and two ads have a rational appeal. This manipulation check was to find out whether the manipulated ads were indeed seen as rational and emotional appeals. The second manipulation check is for the moderator ‘type of products’. Two different products, one hedonic and one utilitarian are used for the ads in the survey. The manipulation check will show us if these products are recognized as hedonic and utilitarian by the participants.

4.2.1 Manipulation check for ad appeal

The manipulation check for ad appeal is performed by asking the participants to rate one of the four ads (a rational ad and an emotional ad for each product) on four measurements in order to find out whether the manipulated rational ads were seen as rational ads and whether the manipulated emotional ads were seen as emotional ads by the participants.

After checking the Cronbach’s Alpha of the ad appeal measurement, the questions are combined into a new variable which is named SOMadappealTOTALSAMPLE. The questions have a Cronbach’s Alpha of 0,83 (see appendix 4.1). Hair, Black, Babin & Anderson (2007) state that a Cronbach’s Alpha of 0,6 is sufficient for summed scales to be reliable. An independent samples t-test is performed, and the results show a significant difference between the two appeals t(178)=7,30, P=0,000. This says that the mean of the emotional ads (M=4,6, SD=1,29) differs with the mean of the rational ads (M=3,24, SD=1,24) on the emotional dimension (see also appendix 4.2). The conclusion is therefore that all four ads communicate the intended type of ads.

4.2.2 Manipulation check for moderator type of products

The manipulation check for type of products is performed by asking participants to rate on one of the four ads (with a hedonic product or a rational product) on five items in order to find out whether the hedonic products were seen as hedonic products and whether the rational products were seen as rational products.

(43)

42

this measurements does not have a sufficient Cronbach’s Alpha (0,30). However, if we removed one item, the Cronbach’s Alpha would be 0,52. The item that is removed is the statement: “the buying decision making of this product reflects my personality”. The reason that this item completely did not fit in could possibly be because the statement was too personal for respondents. Where the other items had neutral statements, this item directly addressed the respondent. This could have affected their answers. Although the cronbach’s Alpha is now very close to being sufficient, the decision is made to remove another item in order to be more sure that the questions will measure the same underlying concept. The cronbach’s Alpha with the remaining 3 questions is now 0,64. The questions are combined into a new variable which is called TOPtotalsample (see appendix 4.3). An independent samples T-test is performed and shows that there is no significant difference between the two type of products, t(178)=-1,57, p=0,119. The mean of the utilitarian product (M=3,8, SD=1,26) and the mean of the hedonic product (M=4,2, SD=1,3) do not differ significantly on the hedonic dimension (see appendix 4.3 for detailed results). This means that the utilitarian product (a bottle of water) is not seen as a typical utilitarian product, but that the hedonic product (a bottle of Diet Coke) is seen as a hedonic product. Although the difference is not significant, the mean of the utility product is lower than the mean of the hedonic product. This could be because the bottle of water is not perceived as being a very hedonic product but rather in between.

4.3 Reliability

Just like for the above mentioned measurements, a reliability analysis is performed for the dependent variable purchase intention and for the moderator need for closure. These measurements have a sufficient Cronbach’s Alpha. Table 8 shows the Cronbach’s Alphas of both variables. At the end of the survey, the control question measuring the brand familiarity of the products is asked. The Cronbach’s Alpha for these variables is also sufficient and can be found in table 8 (appendix 4.4 shows all results). Since all measurements have very high Cronbach’s Alphas, none of the question items were deleted.

Measurements Nr of items Cronbach’s Alpha

Purchase intention 4 0,95

Need for closure 15 0,86

Brand familiarity 3 0,97

(44)

43

The variables of purchase intention are computed into a new variable which is called SOM_PI. The new variable of need for closure is called SOM_NFC and the new variable for brand familiarity is called SOM_BF. For all new variables that are computed, the mean of the variables is used. This is done by dividing the summated variables by the amount of variables used in the new variable.

4.4 Hypotheses tests

The main subject of this section is whether the hypotheses can be confirmed or rejected. In order to give a simple overview of the purchase intention outcomes, a table is presented which shows the mean purchase intentions per condition. In Appendix 5.1 a more detailed overview can be found. The simple overview can be viewed in table 9.

Purchase intention Condition 1 Emotional & utility Condition 2 Rational & utility Condition 3 Emotional & hedonic Condition 4 Rational & hedonic Mean 2,78 4,54 3,65 2,92 SD 1,196 1,388 2,274 1,863

Table 9: means purchase intention per condition

The condition with the highest purchase intention mean is condition 2, a rational ad appeal with a utilitarian product. The lowest mean is condition 1 which is an emotional ad appeal with a utilitarian product. Based on just the means, not much can be said about the hypotheses. In order to see if the hypotheses are confirmed or rejected, a multiple regression analysis is performed. In order to check for significance a confidence interval of 95% will be chosen. The p-value should therefore be <0,05 for the hypotheses to be supported.

(45)

44

in order to see if this could have influenced the results. Brand familiarity could possibly have an effect the purchase intention if the brand was known to them.

The equation that is used for this analysis which shows the analyzed interactions is as follows:

Yi = α 01 + β 1X1i + β 2X2i + β 3X3i + β 4X4i + β 5X5i + β6X6i + e1

Yi = Purchase intention α 01 = intercept

X1i = Ad appeal X2i = Type of products X3i = Need for closure

X4i = Interaction ad appeal and type of products X5i = Interaction ad appeal and need for closure X6i = Brand familiarity

e1 = Error term i = Respondent i

(46)

45

Table 10: results hypotheses and control variable

4.4.1 Hypothesis 1: Ad appeal

Hypothesis 1 states that advertisements with a rational appeal will lead to a higher purchase intention compared to advertisements with an emotional appeal. From table 10 can be derived that H1 (B=-0,609, t(173)=-1,723, p >0,01) has a negative Beta weight and not a significant p-value, meaning that the ad appeal does not have a significant effect on purchase intention. This means that advertisements with a rational appeal do not necessarily lead to a higher purchase intention compared to advertisements with an emotional appeal. Hypothesis 1 should therefore be rejected. A possible reason for this, is that rational appeals probably do not elicit the same effect with hedonic products as it does for utilitarian products. This could lead to the reason that rational appeals do not significantly differ from emotional appeals. Looking at the means of the previous mentioned conditions in table 9, we can see that rational appeals combined with utilitarian products have a higher purchase intention mean than rational appeals depicted with hedonic products. However, the results of H2 could tell us more about this.

4.4.2 Hypothesis 2: Ad appeal and type of products

Hypothesis 2 states that the positive relationship between advertisements with a rational appeal and purchase intention is higher with utilitarian products compared to rational ad appeals with hedonic products. The results (B=1,999, t(173)=3,888, p <0,01) show a positive Beta weight and a significant p-value. This means that an increase in type of products, Model Unstandardized Coefficients Standardized

Coefficients

B Std. Error Beta t. Sig.

Constant 4,055 0,305 13,296 0,000 H1: Ad appeal -0,609 0,353 -0,165 -1,723 0,087 H2: Ad appeal * type of products 1,999 0,514 0,468 3,888 0,000 Type of products 0,230 0,482 0,062 0,477 0,634 H3: Ad appeal * need for closure -0,084 0,272 -0,030 -0,309 0,758

Need for closure 0,485 0,196 0,244 2,474 0,014

(47)

46

increases the purchase intention. An increase in type of products means we are talking about an utilitarian product, so this means that rational appeals with utilitarian products lead to a higher purchase intention than do rational appeals with hedonic products. H2 can therefore be accepted.

The results also show the effect of rational appeals on purchase intention, without taking the appeals into account. These results (B=0,230), t(173)=0,477, p >0,05) show that hedonic products or utilitarian products (without taking the appeals into account) do not significantly lead to a higher purchase intention. This is probably very logical, because people buy both products in their lives, people probably do not make a distinction between both type of products because they buy both kinds of products.

Another interesting finding about these results is that although consumers did not see the bottle of water as a typical utility product, they did have a higher purchase intention for this utility product in combination with a rational ad appeal. A possible explanation for this is that the ad appeal might have influenced the perceived type of product. An emotional ad appeal which states pleasant and unimportant information about the bottle of water could therefore make the consumer believe that the product is only used for pleasure and ideals that consumers pursue. A rational appeal could lead to the opposite effect. This could be an interesting finding for further research.

4.4.3 Hypothesis 3: Ad appeal and need for closure

The 3rd and last hypothesis states that the positive relationship between a rational ad appeal and purchase intention is stronger for high-need for closure consumers compared to low-need for closure consumers. Results (B=-0,084, t(173)=-0,309, p >0,05) show that the effect is not significant. Next to this, the negative Beta also shows us that the direction is not correctly described in H3. This could possibly be because high-need for closure consumers are only affected if the rational ad appeals are combined with utilitarian products. Since the hypothesis talks about rational appeals in general, hypothesis 3 cannot be accepted.

(48)

47

purchase intention if the type of product is utilitarian. Another possibility is that high-need for closure consumers might have a higher purchase intention than low-need for closure consumers because of the available information on the ads, regardless of the fact that the information was emotional or rational. This might have something to do with the two tendencies that high-need for closure consumers have that are stated by Tetlock (1998). These tendencies are: the urgency that inclines them to grab quickly on readily available explanations and to reject alternatives, and the permanence that inclines them to persist on the explanations and stay with them even if it is proven otherwise. The information on the ads (either rational or emotional) might elicit these tendencies and they might therefore have higher purchase intentions.

4.4.4 Control variable brand familiarity

Brand familiarity was added to the questionnaire to find out if respondents would answer differently on purchase intention if the brand was known to them. At first, the impression was that respondents would not recognize both of the brands in the surveys since Voss is a Norwegian brand that sells its products mostly trough hotels and restaurants. Diet Coke is of the well-known brand Coca Cola. However, the name Coca Cola is not mentioned in the Diet Coke ads. The impression was therefore that Diet Coke would not be recognized. When we look at the results (B=-0,196, t(171)=-2,503, p <0,05) we can see that brand familiarity has a significant influence on purchase intention. This means that if the brand is recognized, the purchase intention decreases. To find out if there is a brand that is more familiar than the other, the means per brand on brand familiarity are checked. Table 11 shows that the brand Diet Coke has a lower mean than Voss. This shows that Diet Coke is more familiar than Voss.

Brands Means

Voss 5,76

Diet Coke 2,35

Table 11: means between brands