UNIVERSITY OF GRONINGEN FACULTY OF ECONOMICS AND BUSINESS

(1)

UNIVERSITY OF GRONINGEN

FACULTY OF ECONOMICS AND BUSINESS

Consumer Decision-Making in Selling of

Anthropomorphized Product Bundles

MASTER THESIS

June 2018

BY

Simona Sopčáková

(s3255816)

First supervisor:

dr. Jing Wan

(2)

ACKNOWLEDGMENTS

(3)

3

ABSTRACT

(4)

4

INTRODUCTION

In today’s world everyone is surrounded by enormous number of products. One of the ways people tend to declutter their surroundings is by selling their unused or old possessions. This usually happens between friends or people we know, in yard sales or lately even via internet on websites such as craigslist, E-bay, Marksplaats, or through Facebook groups. When looking through these web pages, one can notice that people not only sell products on its own but in a combination with others. Dinnerware sets, aquarium with cleaning systems, bed and bedside tables – all of these packages of various items can also be called product bundles. This strategy of product bundling (i.e. various products put into one package) has been successfully used by many companies in order to increase their sales (Guiltinan, 1987). However, research has been considerably sparse in focus on this different type of marketplace involving product bundles sold by one individual to another.

Typically, when consumers decide to sell their possessions, they do so with the goal of profit maximization, otherwise they could have just donated them. When products are put together into one package, consumers first evaluate the value of the package as a whole and decide whether its combined value is worth it, or whether for the customer the value of products is higher when they are separated (Gaeth, Levin, Chakraborty and Levin, 1990; Yadav, 1994). The aim of current research is to look at consumers’ behavior towards selling products in bundles and how it can be affected by other factors, one being anthropomorphism. Anthropomorphism is part of people’s lives even without them being necessarily aware of it. An easy example of the anthropomorphism can be seen in an episode of ‘Friends’ where the character Joey has ‘Stevie the TV’ and ‘Rosita the recliner’. He perceives these products as someone with a personality of its own and not as something inanimate with no feelings whatsoever. This perception changes his attitude and behavior towards these items (e.g. being sad about Rosita’s ‘death’). Anthropomorphism therefore involves imbuing imagined as well as real objects with mental and physical features of humans (Epley, Waytz and Cacioppo, 2007). Since we perceive anthropomorphized products to look like us or be like us humans are, the treatment of the anthropomorphized products has been found to reflect the social interaction with humans (Ahn, Kim and Aggarwal, 2014; Luczak, Roetting and Schmidt, 2003).

(6)

6 2006). This unity is stressed and respected not only because of legal but also social reasons – one just cannot suddenly decide that someone does not belong to a family because such decision can cause social outrage and negative emotional response within the family but also within the person separating the family (Broderick, 1993). By looking at these findings, it creates a question of whether anthropomorphization of product bundle makes consumers see it as one family, therefore changes how they value their product bundles and its willingness to separate them according to value.

Exchanging value through selling and buying has always part of human existence. Yet, one thing that still might hold people against putting a price tag on own or other people’s life are sacred values (e.g. honor, family). While bearing in mind that consumers still want to sell the products, would they be able to omit potential increased financial benefit for the sake of personal values? Or choose to overlook their personal values and separate the product bundle in order to gain more money? This research predicts that the effect of anthropomorphism makes people to see product bundle as an inseparable entity (i.e. family) which is not worth the damage of this personal value put into a trade-off with increase in profit from sale where the products are separated.

Since products are not all the same in the role they present for customer’s needs, consumers have different perceptions and attitudes towards them. Certainly, by looking at a package of purely utilitarian products such as gardening products (e.g. saw, leaf blower, lawn mower), it might by easier to imagine selling these items separately for more money than humanized package of products that provided fun or enjoyable experiences (e.g. Christmas ornaments). All in all, this research proposes the idea that anthropomorphized product bundle discourages people from gaining more profit by selling each product within the bundle separately while the effect is to be moderated by the role of the product bundle (hedonic or utilitarian).

(7)

7

LITERATURE REVIEW

Product bundling

For many people, selling and buying certain products entails lots of decision-making and thinking about how this can be accomplished while satisfying one’s wants and needs as well. In order to simplify decision-making for customers and increase sales, sellers introduced offers which comprise of different combinations of products. This strategy is called product bundling – “the practice of marketing two or more products or/and services together in a single

‘package’ for a special price” (Guiltinan, 1987: 74). By product bundling, sellers are able to

satisfy diverse needs of their customers in one offer (Gaeth, Levin, Chakraborty and Levin, 1990). Going camping? – then a product bundle including a tent, sleeping bag and an air mattress might come in handy! Product bundles can be comprised in two ways: as a (a) pure

bundle - the goods are sold only as a package and cannot be purchased separately (e.g. laptop

with pre-installed software), or as (b) mixed bundle, where certain products can be added or removed from the bundle and are available to be purchased separately (e.g. laptop with a docking station, a mouse etc.; Adams and Yellen, 1976).

Consumers’ decision of whether they will buy the products separately or in already set up product bundle is based on the impression of product bundle. To make this impression, customers use averaging model where each product’s value in the bundle is evaluated separately and then combined into one average value of the whole bundle (Gaeth, Levin, Chakraborty and Levin, 1990; Yadav, 1994). Additional to product characteristics, this value includes perception of additional financial savings from buying the bundle or the product separately (Yadav and Monroe, 1993). For example, the impression of a package comprised of a table and chairs is created from first contemplating the attributes and total price of the chairs and table separately, and then putting these evaluations together to form one whole impression of advantages and disadvantages of buying the whole package. Yet, the customers also adjust their impression by the product which they perceive as more important and thus the evaluations of other products in the bundle are based on the evaluation of this one, more important product out of the bundle (Yadav, 1994). This suggests that people base their impressions of product bundles first on the product which is stated as a first (leading) one in the bundle but inherently, the decision to purchase a bundle or each product separately depends on the costs and savings.

(8)

8 largely depends on these economic factors - the level of costs and the reservation price space (i.e. price limit) of both the seller and a buyer (Adams and Yellen, 1976). This means that whether the products put into a bundle are indeed sold as a bundle depends on the fit between the lowest/highest possible prices set for the product bundle. The fit needs to exist between a price for which the seller’s costs are at (still profitable) maximum and the highest possible price the customer has in mind for buying each individual product separately. To put this situation into an example; if a customer has a limit of 500 euros and wants to buy a laptop with a MS Office, then the package consisting of a laptop and two programs instead of one (MS Office and Adobe Acrobat) for 530 euros might be an attractive offer. However, the difference in willingness to buy this product bundle depends on the knowledge of what value such program and laptop has (Gaeth, Levin, Chakraborty and Levin, 1990). If separately each program costs around 80 euros and the laptop 400 euros (and the customer knows this) then the product bundle presents a bargain. On the other hand, if the seller has an option to sell these items for more than 530 euros (which is seller’s lowest possible price) to someone who is aware of the value of each item, they would most probably decide to sell these items separately and thus increase their profit.

The already discussed judging of product bundles from buyers’ and sellers’ perspectives is however, able to be put together and occur within one individual – a buyer who now acts as a seller. Product bundling not only occurs on the company level but also when individuals are trying to re-sell their possessions. Many people have encountered a situation where they wanted to dispose of various products which they do not use anymore. Certainly, selling a list of products, item by item is a long exhausting process which may end up without success. In this situation, selling in bundles might come in handy. Existing literature on product bundling implies that the decision of choosing products in a bundle or separately largely depends on the value and costs the product bundle creates both to the seller and the buyer. However, the literature largely consists of seller’s point of view only at the company level, whereas the research about individuals as sellers is scarce. Therefore, it might be useful to look into how this process of selling products in bundles by an individual can be influenced by other than monetary factors (i.e. profits and costs).

Anthropomorphism

(9)

9 anthropomorphism. After thorough analysis of philosophical and scientific views, Guthrie (1995) stated that anthropomorphism is perceived as a cognitive bias where humans process stimuli in anthropomorphized versions until it can be proven that the object is in fact just an object, not a human-like form. However, according to Epley, Waytz and Cacioppo (2007: 865),

“anthropomorphism entails attributing humanlike properties, characteristics, or mental states to real or imagined non-human agents and objects”.

There are many forms of anthropomorphism, ranging from literal (e.g. mistaking inanimate non-human object or event to be a person) to metaphoric (e.g. earthquake as a punishment form from the ‘higher power’, talking to flowers). In a marketing context, Reavey, Puzakova and Kwak (2011: 657) categorized anthropomorphism as (a) explicit – “imbuing a

nonhuman entity with observable and apparent human characteristics” (e.g. nose, ears) and (b) implicit – “imbuing a nonhuman entity with indirect and subtle human characteristics” (e.g.

windows positioned to form a face). Nevertheless, one underlying concept which connects anthropomorphism is familiarity – by familiarizing unknown object it becomes more understandable to people (Guthrie, 1995).

Anthropomorphism puts physical (e.g. body shape, limbs, face, eyes) and mental attributes (e.g. emotions, thinking capacity, intentions) unique to humans, to objects (Epley, Waytz and Cacioppo, 2007). Simply by looking around one’s surroundings, people might see objects put into such compositions, and being designed in such way that they appear human or human-like. People always look for faces in their environment first since it comes naturally, very quickly and with ease (e.g. faucet with two knobs, electrical outlets as having eyes and mouth; Purcell and Stewart, 1986). In addition, inanimate objects may appear alive therefore humanlike if their movements resemble motions of a person (Heider and Simmel, 1944; Morewedge, Preston and Wegner, 2007). Due to survival, quick detection of one’s environment is essential. Rationally inexplainable justification for the anthropomorphism comes second (Atran and Norenzayan, 2004). This explains why when walking around alone in the dark, a pole and a trashcan may appear as a man waiting to cause harm.

(10)

10 increases the ability to see them as something that deserves higher levels of respect, as if it was another person (Epley, Waytz and Cacioppo, 2007; Ahn, Kim and Aggarwal, 2014). Since the non-human anthropomorphized objects tend to be treated as human, it can be said that people perceive two-way interaction - the human has feelings towards the humanized object and the object perceived as human acts the same way and possess feelings towards humans (Epley, Waytz and Cacioppo, 2007; Waytz, Epley and Cacioppo, 2010). This implies that anthropomorphized objects deserve to be treated morally and their actions should be taken into consideration (Gray, Gray and Wegner, 2007).

Treatment of anthropomorphized products

Earlier results of a study by Aggarwal and McGill (2007) posit that the evaluation of anthropomorphized objects changes due to combination of already built social knowledge of an individual and his/hers processes of social cognition. This implies that people base their attitude towards anthropomorphized products on their knowledge of how to behave in various situations towards various people, and on how the people think and feel about the others. However, it needs to be further discussed whether the anthropomorphic versions of products indeed elicit comparable attitudes and reactions of people as if the products are like a human.

One of few examples of this behavior was presented by study of Chandler and Schwarz (2010) where people were found to less likely replace their cars, if they thought of them as having a personality rather than just them being thought of as an inanimate object. The car having a personality (i.e. being more like person), made them less likely to replace it because people usually do not replace people who are already close to them for new people. This aligns with a previous finding of Waytz, Cacioppo, and Epley (2010) where thinking about a product as having a mind of its own activates a treatment where people care and are more concerned about the anthropomorphized product. To support this line of reasoning even further, a car rental company which encouraged their customers to name and create personalities for their rentals, claimed that the customers’ treatment of the anthropomorphized cars significantly improved when compared to times when the cars had no human-like attributes (Levine, 2009). Therefore, it can be said that when people perceive objects to have a personality, they treat these objects more as they would have treated a person.

(11)

11 lightbulb). The participants were found more likely to help the anthropomorphic versions; however, this compliance was mediated by anticipatory guilt from not helping the entity. Therefore, it is clear that the guilt of not helping anthropomorphized products elicited a feeling of need to help, which can indeed be experienced in a situation involving a human being (e.g. homeless person asking for food).

Treating anthropomorphized products as humans might also occur in a negative way since people treat each other negatively too. Normally, it is socially unacceptable to physically harm a person in a day to day life, however this does not mean it is not happening (e.g. calling names, violence, abuse). This behavior towards anthropomorphized entities was described in a study of Luczak, Roetting and Schmidt (2003) where the respondents admitted to negative treatment of products they also believed to have a mind of its own, thus are humanlike. In case the product failed to complete the human’s request, people cursed and scolded their computers and sometimes even hit it. Similar situations appear in a human to human interactions where one might get yelled at for disobeying one’s orders such as not cleaning the dishes or failing to meet one’s expectations of getting a good grade at school. Therefore, treating an anthropomorphized product in a (socially proclaimed) negative way, occurs as well.

Wan and Aggarwal (2015) suggest that people who perceive inanimate objects as human or human-like, treat them in interactions accordingly to social norms as they would treat people. This fairly recent view is supported by previous findings where a social norm of not simply replacing a close person for another new one, made people less likely to replace their anthropomorphized car (Chandler and Schwartz, 2010) and the norm of helping people in need, made individuals more likely to help save environment by helping anthropomorphic products with their problems and needs (Ahn, Kim and Aggarwal, 2014). Further scientific explanation for this association might come from the findings of Castelli, Happé, Frith and Frith (2000). Here, the same neural systems which are involved in judging humans were found to be activated when people made anthropomorphic judgements about the non-human objects. This would imply that social knowledge is being activated when people come in contact with anthropomorphized product. All in all, it can be said that anthropomorphized products tend to be treated differently than its non-anthropomorphic versions, meaning treated more like a human.

‘Bundle’ of humans - family

(12)

12 people (products) are put together, it is generally called a group (bundle). However, when these individuals within the group also interact and have some sort of relationship with each other, it is perceived to be a social group (Macionis, 2011).

The creation of social groups primarily stems from human need to belong - to create and preserve positively impactful social relationships with others (Baumeister and Leary, 1995). People are social creatures seeking human interaction (Fiske, 2010) and combined with the previous finding, it implies that people have a natural drive of wanting to create a long-lasting unit of people which is essentially unbreakable. As a matter of fact, the preservation itself even leads to avoidance of breaking the bonds within the group (i.e. the unit) since the tendency of humans to react to ending of a relationship (thus separating the unit) is to be distressed and protest against it (Hazan and Shaver, 1994). One of the most prominent reactions to involuntary separation (e.g. physical distance with no possible interaction, divorce, death) is the presence of general state of upset (e.g. negative emotions such as crying, anxiety or sadness; Northen and Kurland, 2001). Therefore, it can be said that due to human need to belong, people do not want to be separated from social groups; and in case this unity of group is indeed dissolved, it mainly induces negative feelings. This notion aligns with Baumeister and Leary’s (1995) proposition that people act strongly against dissolution of social bonds, even if it goes beyond rational or practical thinking (e.g. leaving abusive relationship). However, it needs to be noted that dissolution of some groups creates much stronger emotions and protest than some others (Northen and Kurland, 2001). Indeed, primary social group of a family produces more detrimental effects if its unity is affected, especially among younger members (Derluyn and Broekaert, 2008).

Despite multitude of different definitions, in many cultures the most important social group is still a family (Starr and Brilmayer, 2003), nowadays also defined as a “group of

intimates who generate a sense of home and group identity and who experience shared history and shared future” (Koerner and Fitzpatrick, 2002: 71). Contrary to this scholar definition, the

(13)

13 own within the family (Turner and West, 2006). Therefore, when people think of families, they think of their members and how these members interact and follow certain roles (e.g. mother, father, child). As suggested by previous findings, in case of separation of one member from the family, the family is no longer what it used to be, and the previously held roles and responsibilities change as well.

Keeping family together (i.e. preserving its unity) is generally stressed and supported social norm across all nations, but the strength of this norm varies accordingly to cultures (Starr and Brilmayer, 2003). Deviation from this norm of ‘family should stick together, no matter what’ by separating family members against their will, generates not only potential legal prosecution (depends on the country) but informal social sanctions (e.g. judgment, alienation) and in majority of the cases, personal guilt (Broderick, 1993). Indeed, even a suggestion of separating family members from each other in a simple setting such as seating in the restaurant might make us uneasy even though rationally, it should not matter since everyone gets a seat somewhere for them to eat anyway. Be the cause of a divorce or sibling estrangement? - majority of the people do not want to be the reason which breaks families apart. Nevertheless, families are being separated despite protests of some or all members due to people not following this norm in their decision making, or the norm being outweighed by other factors (e.g. illegal immigrant families).

Taboo Trade-Offs

Fershtman, Gneezy and Hoffman (2011) posit that every society has three incentives which influence and monitor people’s behavior: (a) private rewards i.e. money, (b) social

incentives i.e. norms, taboos and (c) legal incentives i.e. laws and regulations. Sometimes

(14)

14 Values can be categorized as sacred and secular (e.g. Tetlock, Mellers and Scoblic, 2017; Hanselmann and Tanner, 2008; Tetlock, Kristel, Elson, Green and Lerner, 2000). The most valued – sacred (e.g. love, honor, human right) are those values to which people are fully committed to hold on to, they are incomparable and should not be put into a trade-off together (tragic trade-off; Tetlock, McGraw and Kristel, 2004). A classic example of this trade-off might be putting a love for a child against murdering someone - indeed, hardly imaginable situation to compare such options. The opposite of such sacred values are secular values (e.g. money) which when put against each other, a list of ‘pros and cons’ is a rational and normal thing to do (Tetlock, Kristel, Elson, Green and Lerner, 2000).

A situation in which a secular value is put to trade-off with a sacred value is considered to be taboo (Tetlock, Kristel, Elson, Green and Lerner, 2000). Hence, this type of trade is called

taboo trade-off, where “[taboo trade-off] calculus subverts or undercuts the other value”

(Tetlock, McGraw and Kristel, 2004: 249). Important note of Tetlock and Belkin (1996) states that what is considered as a taboo trade-off differs in various cultures and across the timeline of the history. Imagine a scenario where choosing a secular value of money over saving someone else from being physically or mentally harmed (i.e. sacred value) is presented as options. These comparisons and even simply thinking about them (not actually pursuing them) are perceived to be taboo because they contaminate and corrode morals, therefore are considered as ‘unthinkable’ (Fershtman, Gneezy and Hoffman, 2011; Tetlock, McGraw and Kristel, 2004; Tetlock, Kristel, Elson, Green and Lerner, 2000). If one was to choose or even think about selling own or any child to get one million euros, they would certainly be condemned by the society. Exchanging sacred values against money is therefore perceived as inexplicable and outright taboo (Tetlock, Mellers and Scoblic, 2017).

(15)

15 situations which implies that people’s morals and norms are not that set in stone thus cannot ever change.

Despite the evidence of some sacred values being only pseudo-sacred thus making the trade-off easier on decision-making, it does not mean the decision-making is less heavy on creation of negative emotions. For example, when people were asked choose between improving flood protection of a village (sacred value of potentially saving lives) or renovating village’s square (secular value of improving quality of the village), this scenario left them feeling more upset as when the trade-off was between better annual salary compared to more vacation days in a year (both secular values; Hanselmann and Tanner, 2008). As described by the setting used from the study of these authors, a trade-off which involves a sacred value is found to be accompanied with more negative emotions than those trade-offs which do not include a sacred value (ibid). Additionally, these authors found that relationship between negative emotions and perceived decision-making difficulty was discovered to be non-linear (Hanselmann and Tanner, 2008). This means that taboo trade-offs tend to elicit negative emotions but are solvable with less difficulty than routine trade-offs (i.e. situations without sacred values; e.g. better salary versus shorter travelling distance).

CONCEPTUAL FRAMEWORK AND HYPOTHESES

The review of literature revealed certain gaps and created certain questions which lead to the current study’s research question of whether anthropomorphized product bundle discourages people from gaining more profit by selling each product within the bundle separately.

When customers think about choosing to purchase products which are in a bundle versus opting for products sold individually, they largely base this decision on what costs and savings each of the options provide (Yadav and Monroe, 1993). For the sellers, the decision to put products into bundles or sell them separately depends on costs and financial benefits product bundling creates (Adams and Yellen, 1976). Therefore, it is suggested that in case a buyer becomes the seller, the seller should look for the most profitable option of selling their possessions which would be the option involving maximization of profits while the costs are at the minimum.

(16)

16 mental attributes (e.g. emotions, thinking capacity) unique to humans to objects, people no longer see the products as just objects but as something which resembles a human (Epley, Waytz and Cacioppo, 2007). Further, these anthropomorphic versions of products are believed to be treated in similar manner than the humans themselves treat each other – in positive (e.g. helping social cause) as well as negative ways (e.g. physical abuse; Ahn, Kim and Aggarwal, 2014; Luczak, Roetting and Schmidt, 2003). This rationale stems from Aggarwal and McGill (2007) who posit that people base their attitude towards anthropomorphized products on combination of already built social knowledge of how to behave in various situations towards various people and their processes of social cognition (i.e. thinking and feeling about people). If by anthropomorphizing, people indeed perceive products in a bundle to be something ‘human’, then they can also see this bundle of anthropomorphized products as a human family - integrated emotional unit of personalities, each having a role of its own within the family (Turner and West, 2006). Contrary to the buyers and sellers’ thinking about product bundles, people do not base the decision to keep unity of the family (not separating the ‘bundle’) on financial factors. Due to human need to belong, people generally avoid and act against breaking already established bonds (Baumeister and Leary, 1995). In case these relations are involuntarily disrupted, it usually creates long-lasting negative emotions for the people involved (Hazan and Shaver, 1994). Involuntary separation of members of human family is not only illegal (in majority of societies) but by failing to adhere to a norm of ‘keep family together’ (Starr and Brilmayer, 2003), people can experience feelings of personal guilt and face informal social sanctions (Broderick, 1993) from breaking what’s generally considered to be a cornerstone of society.

People’s behavior can be influenced by private incentives such as money, social incentives (e.g. social norms) and legal incentives (Fershtman, Gneezy and Hoffman, 2011). Many decision-making situations deal with these incentives as misaligned or even against each other. This type of situation, where one has to choose between following one incentive over the other is called a trade-off (Hanselmann and Tanner, 2008). In this trade-off of incentives, people essentially make a decision on what they value more. As proposed by Tetlock, Kristel, Elson, Green and Lerner (2000) a trade-off in which one should not hesitate to choose one of the most valued value - sacred value (e.g. family unity) over secular value of money can also be called a taboo trade-off.

(17)

17 believes that when anthropomorphized product bundle’s unity is put into trade-off with a secular value of potential personal financial benefit, consumers are less likely to corrode their morals and choose the secular value over sacred. Hence, this taboo trade-off of values makes consumers more likely to omit higher financial benefits and not separate the ‘members’ of anthropomorphized family. This line of reasoning formally transforms into the following hypothesis:

Hypothesis 1: Consumers are less willing to separate the products in an anthropomorphized product bundle in exchange for higher profits derived from selling the products separately.

This research further predicts that this effect is to be moderated by the role of the product bundle. Product role is considered to be an important aspect in processing of available information and subsequent decision making about the product (Huettl and Gierl, 2012). Generally, products can be characterized as having either utilitarian or hedonic role. Utilitarian products are bought for their practicality and functionality since they appeal more to the rational thinking (Woods, 1960). They take on the role to accomplish a task or fulfill a basic need, with consumers rarely having any pleasure during its usage or consumption (e.g. lawn mower, toilet paper, road salt; Strahilevitz, 1999). On the other hand, consumers buy hedonic products for their own entertainment and pleasure since its role is to evoke emotions and provide multisensory experience. Therefore, they are considered to be purchased (for the most part) for their symbolic value (e.g. TV, jewelry, cookies; Hirschman and Holbrook, 1982). Nevertheless, it is important to mention that substantial number of products do not simply belong to purely hedonic or utilitarian category (Strahilevitz, 1999). The categorization of these products depends on the perception of value they represent to each individual (Park and Moon, 2003). For example, some people have a bicycle purely to get from point A to point B, hence the bike has a utilitarian role for them. On the other hand, some individuals might treat the bike as a way to relax thus representing the hedonic role.

(18)

18 positively to a potential interaction with an anthropomorphized camera which provides only its different functions. By looking at these results, this study implies that in case a product has a utilitarian role, the presence of anthropomorphism is less likely to change in its effect on dependent variable compared to the effect when the product has a hedonic role.

People are social creatures who seek human interaction (Fiske, 2010). Anthropomorphized products can provide people with social interaction (Fournier, 1998). As previously mentioned that hedonic products affect our emotions and provide plausible experience (Hirschman and Holbrook, 1982) while utilitarian are used to accomplish tasks (Strahilevitz, 1999), it can be suggested that interaction with anthropomorphized hedonic products is more valuable due to increased potential social impact. Certainly, when one goes on a trip full of adventures around a world with their long-time friend Billy the bike, the interaction is more valued than when Billy the bike is just a companion who is used for moving from one place to another. Combined with the finding of Dhar and Wertenbroch (2000) that people have stronger attachments to hedonic rather than utilitarian products, it can be said that anthropomorphized hedonic products are more likely to be perceived as some close friends/family members, someone one can have stronger bonds and enjoyable times with, and not as an occasional stranger with whom one interacts due to their utility to accomplish something for us. In this line of reasoning, when anthropomorphized hedonic products are in a bundle, they should be perceived as a group of family members who lived together in one household with the consumers. Whereas bundle of anthropomorphized utilitarian products is to be perceived as a group of strangers who shared one common place to accomplish tasks within the household. To put this in the purely ‘human’ perspective, one would be more likely to separate a group of people for their own benefit when the members provide only functional use (e.g. plumber, cleaner and carpenter) rather than people who share aspects of life together and share one symbolic value as a whole (e.g. childhood friends, family members). Hence, it is suggested that people are more likely to sell separately the anthropomorphized utilitarian products which provide only functional usage. The reason for it is that this separation would most likely not corrode people’s morals as it would, if they would be to choose personal financial benefit over unity of the anthropomorphized product bundle with the hedonic role (i.e. separating the symbolic value of unity of group of close friends/family for money).

(19)

19

derived from selling the products separately than for an anthropomorphized hedonic product bundle.

METHODOLOGY

Pre-test

This pre-test study was conducted with a purpose of developing a manipulation for the independent variable of a product role in the main study. The goal was to test and identify a product bundle of which the characteristics are rather ambiguous in terms of its product role and therefore can be manipulated to appear as more utilitarian or hedonic. The ambiguity would have ensured that any variation in answers between conditions in the main study are strictly due to manipulations of the product role and not any other possible differences such as liking of the different products (e.g. comparison of bundles which are different in its nature - fireplace accessories vs. video game console and games).

Pre-test study consisted of an online survey completed by sample of mostly graduate students (N=30, 20 women, 10 men; Mage= 23,3). Participants were randomly assigned to only one of two conditions - one in which product bundles were framed as having utilitarian product role and the other condition framed as having hedonic product role. Both conditions included three short descriptions of various product bundles (matching teapot set, bookcase set, armchair set) with the only difference between conditions being in the descriptions which were manipulated to accentuate either only the utilitarian features in one condition or hedonic features of the product bundles in the other condition. The descriptions in utilitarian conditions focused on listing only functional aspects of the products (e.g. sturdiness, ease of cleaning) whereas in the hedonic condition, the descriptions included only experiential and design aspects

(20)

20 (e.g. ability to uplift mood, material with modern patterns). Both full surveys can be found in

Appendix A.

Participants had to evaluate the aforementioned product bundles using a set of six 7-point Likert scales adapted from Voss, Spangenberg And Grohmann (2003). The set of scales measured degree of utilitarianism/hedonism of the product bundles on six dimensions (three measuring only utilitarianism: 1=helpful/7=unhelpful, necessary/unnecessary, practical/impractical; the other three only hedonism: 1=fun/7=not fun, exciting/dull, enjoyable/unenjoyable). To examine whether the description manipulations of the product bundles were successful within the assigned condition thus appeared as either only utilitarian or hedonic, first the average utilitarian score (measures of utilitarianism combined) and average hedonic score (measures of hedonism combined) were created for each product bundle respectively. Subsequently, these average scores were compared within the condition to find whether there is a significant difference between average utilitarian score and average hedonic score for the product bundle. Composite score of product role was created for this product bundle in utilitarian framed condition and hedonic framed condition only if the aforementioned differences between the average utilitarian and hedonic scores were significant in both conditions. The composite measure of the product role (-7 = purely utilitarian, 0 = equally utilitarian and hedonic, +7 = purely hedonic) was created by subtracting average hedonic score from the average utilitarian score of the product bundles. It served as an identifier of whether the tested product bundle is significantly seen as more utilitarian in utilitarian framing and as more hedonic in hedonic framing when put into comparison. Pre-test also included basic demographic questions about age, gender and annual income.

Results of pre-test

Paired samples test in the utilitarian condition showed all the mean differences between average utilitarian score and average hedonic score of all the product bundles as significant

(p<0,05; Table 1). In the hedonic condition, only the armchair set’s average utilitarian score

(Marmchair_uti_scores=3,467, SD=1,045) was significantly different from its average hedonic

(21)

21

Table 1: Paired Samples Test - utilitarian condition

Paired Samples Statistics Paired Differences (utilitarian - hedonic score)

product bundle scores Mean N SD

SE Mean Mean* t df Sig. (2-tailed) teapot_uti_scores_U 2,711 15 0,975 0,252 -1,511 -3,473 14 0,004 teapot_hed_scores_U 4,222 15 1,407 0,363 bookcase_uti_scores_U 2,622 15 0,881 0,227 -1,733 -5,195 14 0,000 bookcase_hed_scores_U 4,356 15 0,972 0,251 armchair_uti_scores_U 3,022 15 1,185 0,306 -1,000 -2,696 14 0,017 armchair_hed_scores_U 4,022 15 0,859 0,222 *composite score

Table 2: Paired Samples Test - hedonic condition

Paired Samples Statistics Paired Differences (utilitarian - hedonic score)

product bundle scores Mean N SD

SE Mean Mean* t df Sig. (2-tailed) teapot_uti_scores_H 3,533 15 0,990 0,256 0,222 0,632 14 0,537 teapot_hed_scores_H 3,311 15 1,324 0,342 bookcase_uti_scores_H 3,356 15 0,811 0,209 0,289 0,698 14 0,497 bookcase_hed_scores_H 3,067 15 1,508 0,389 armchair_uti_scores_H 3,467 15 1,045 0,270 0,778 2,263 14 0,040 armchair_hed_scores_H 2,689 15 1,050 0,271 *composite score

Since only the armchair set’s average utilitarian and average hedonic scores were significantly different from each other in both utilitarian and hedonic framing, the independent samples test was made only for this product bundle. The test confirmed that the description manipulations in utilitarian and hedonic condition were significantly different from each other since the composite score from utilitarian condition (Marmchair_comps_uti=-1,000, SD=1,436) was

significantly different from the composite score in the hedonic condition (Marmchair_comps_hed=0,778, SD=1,331; t(28)=-3,515, p=0,002). Therefore, the pre-test for this

(22)

22 However, after a comparison of average utilitarian score from utilitarian framing (Marmchair_uti_scores_U=3,022, SD=1,185) with average utilitarian score in the hedonic condition

(Marmchair_uti_scores_H=3,467, SD=1,045), the difference was found to be insignificant

(t(28)=-1,089, p=0,285). From the practical point of view, this difference did not necessarily have to be significant since the armchair set has certain functional (i.e. utilitarian) representation which cannot be fully eliminated. Nevertheless, this observation was taken into consideration for the main study as a potentially needed adjustment to the description.

Main study

The main study of this paper was conducted partly via randomized in-person physical distribution in the general public of Slovakia and various company offices (N=109, 87 women, 22 men; Mage=42,72, SDage=11,94). The rest of the respondents (N=122, 85 women, 37 men;

Mage=40,43, SDage=14,64) was acquired via online version of the survey form through market research company’s database which has access to online panel of respondents registered at Slovak National Panel. The data was collected through Qualtrics. No monetary means were used to acquire respondents in the in-person nor the online setting. The survey had two language versions, Slovak and English but the sample predominantly consisted of people with Slovak as their first language (92,6 % Slovak speakers). The respondents were mainly from average to lower income classes (68,6%) and only 14,6% of the respondents had annual income higher than 13 200 EUR. The design of this study was 2 (anthropomorphism vs. control) x 2 (utilitarian vs. hedonic product role) manipulated between-participants. Each condition included product bundle descriptions manipulated to either appear to have (or not) human mental characteristics or actions, and as having either hedonic or utilitarian product role.

(23)

23 hedonic characteristics (e.g. description of design, comfort) in the other pair. As suggested in the pre-test results, the descriptions as used in pre-test were changed for the main study. This occurred due to a perceived need (detected from the results) of addition of more hedonic factors to increase perception of product bundle as more hedonic in the hedonic framing, for the composite score to appear more to hedonic side of the scale. The descriptions were also shortened in length to accommodate the total length of the main study survey. Due to these changes, additional manipulation check was added to test and confirm the degree of consumers’ perceived utilitarianism/hedonism of the product bundle which was located towards the end of the survey.

With the first question “Assuming that the effort to sell to these buyers (one buyer vs.

three buyers) is similar, how would you prefer to sell your Drommar furniture set (control

condition)/Drommars family?” the respondents had to decide on seven-point Likert scale their preferred transaction type: 7=selling to one or 1=multiple (three) buyers, which served as the measure of the dependent variable. The option of selling the items as one whole set is potentially more convenient than selling to three different buyers; in order to make the three-buyer option appealing, the total sum of the selling price of the three products added up to an amount 25 Euros more than selling to only one buyer. In case the respondent chose to sell the products as one set, they could get 175 Euros. However, if they decided to sell each product separately to three buyers for its respective price (armchair for 70 Eur, ottoman 30 Eur and loveseat for 100 Euros), in total they would receive 200 Euros. Additionally, in the anthropomorphism framing, each of these products had its own human first name (e.g. Armie) while sharing one family name as a product bundle. The names were translated for the Slovak surveys to fit the culture. The question measuring dependent variable of ‘product bundle selling decision’ was followed by three questions measuring three variables intended for the exploratory analysis to assess the differences between the anthropomorphism versus control conditions. First question concerned the perceived difficulty of respondent’s choice made which served to explore whether a sacred value was present in the trade-off. Second question, their confidence in this choice and the last explored whether they cared about the products going to caring users to provide further evidence of the presence of anthropomorphism. The outcomes of this exploratory analysis served as an extra insight into the effects of anthropomorphism discussed in this study.

(24)

24 put into one average score of anthropomorphism for the analysis of results. Extra manipulation check was added for the product role due to the change of product bundle descriptions. It followed the same methodology mentioned in the pre-test (i.e. comparison of average scores and composite scores). The last questions in the survey served to check respondents’ understandability of the survey and demographic questions about gender, age, annual income and first language.

The number of respondents stated earlier was modified by cleaning the data from unfinished survey forms (N=1) and respondents who did not understand the survey (i.e. respondents who selected 1=’not clear at all’ as an answer to a question “Was this survey clear

enough and easy to understand?”; N=4). The elimination process of respondents who did not

understand the survey was based on the analysis of outliers via frequency tables and a box plot

(see Appendix B). Therefore, the final sample used for subsequent analyses included 106

respondents from in-person setting (86 women, 20 men; Mage=42,55, SDage=11,81) and 120 respondents acquired online (84 women, 36 men; Mage=40,28, SDage=14,71). Due to set up of this study using two ways of data collection, a potential difference in attention paid to the survey may have transcribed into adjustment of the results, thus the same analyses were done for each sample individually. This difference might have showed in success of manipulations and therefore the results of dependent variable. For further interest, the whole survey with its different versions is provided in Appendix C1 and Appendix C2.

ANALYSIS AND RESULTS

Manipulation check

Combined sample

Anthropomorphism: To test the manipulation success of anthropomorphism, a one-way

ANOVA test was realized. Despite the differences being significant between all four conditions (F(3, 222)=4,441; p=0,005), only one of the two differences (i.e. anthropomorphism vs. control) between conditions which were needed to be different were significant. Post-hoc multiple comparison test revealed the anthropomorphism check to be significant only for the conditions in which the product bundle was framed to have a utilitarian role (Manthro_uti=2,637, SD=1,669

vs. Mcontrol_uti=1,710, SD=1,784; p=0,002; Manthro_hed=2,293, SD=1,807 vs. Mcontrol_hed=2,692,

(25)

25

Product role: The product role manipulation success for the armchair set was tested as well.

Here, the ANOVA analyses revealed no significant differences between composite scores for all conditions (F(3, 222)=1,959; p=0,121). However, only one, marginally significant difference between utilitarian and hedonic framing was found between composite scores of control conditions (Mcompsc_control_uti=-1,707, SD=1,852 vs. Mcompsc_control_hed=-1,128,

SD=1,247; p=0,074; Mcompsc_anthro_uti=-1,005, SD=1,674 vs. Mcompsc_anthro_hed=-1,345,

SD=1,660; p=0,247).

Online and in-person sample

Anthropomorphism: The results of one-way ANOVA analysis showed that the difference

between average anthropomorphism scores were marginally significant for the in-person sample of respondents (F(3, 102)=2,239; p=0,088). However, the comparisons of needed differences between anthropomorphism and control conditions in utilitarian and hedonic framing resulted as having insignificant differences (Minperson_anthro_uti=1,804, SD=0,997 vs.

Minperson_control_uti=1,604, SD=1,132; p=0,624; Minperson_anthro_hed=2,059, SD=1,687 vs.

Minperson_control_hed=2,580, SD=1,498; p=0,160).

Average anthropomorphism scores were found to differ between all conditions in the online sample (F(3, 116)=3,409; p=0,020). Yet, after deeper analysis of applicable pairs, this difference was significant only between samples with utilitarian framing (Monline_anthro_uti=3,081, SD=1,793 vs. Monline_control_uti=1,808, SD=1,234; p=0,002;

Monline_anthro_hed=2,625, SD=1,952 vs. Monline_control_hed=2,796, SD=1,332; p= 0,708).

Product role: No significant differences were found across conditions both in in-person and

online sample (Finperson(3, 102)=1,128; p=0,341; Fonline(3, 116)=1,850; p=0,142). Pairwise

comparisons revealed only one, marginally significant difference between composite product role scores of conditions manipulated as anthropomorphic (Manthro_uti=-0,829, SD=1,489 vs.

(26)

26

Table 3: Pairwise comparisons of composite product role scores - in-person and online sample

In-person

sample Mean SD Sig.

Online

anthro_uti -1,333 1,967 0,756 anthro_uti -0,829 1,489 0,055 anthro_hed -1,186 1,862 anthro_hed -1,569 1,328 control_uti -2,014 1,994 0,222 control_uti -1,423 1,701 0,187 control_hed -1,400 0,943 control_hed -0,877 1,448

-7=purely utilitarian, 0=equally utilitarian and hedonic, +7=purely hedonic

Further details from manipulation check results can be found in Appendix D.

Main study

In the following analyses of results, it was expected to find support for the hypothesis that anthropomorphism causes the effect of selling product bundles to go in the direction of selling products more likely as one package. Additionally, product role was expected to serve as a moderator of this relationship where the utilitarian role of the product bundle would cause the relationship to go more likely towards the option of selling the product in the bundle separately and go in opposite direction for the hedonic product bundle (i.e. direction of selling the product as a bundle).

Factorial 2 (anthropomorphism vs. control) x 2 (utilitarian vs. hedonic product role) ANOVA test revealed anthropomorphism to have no main effect on the product bundle selling decision (i.e. sell as one set vs sell separately) since the difference between anthropomorphism framed conditions (Manthro=5,202, SD=2,379) and control conditions (Mcontrol=4,980,

SD=2,338) was insignificant (F(1, 222)=0,604, p=0,438). The difference between conditions in which the product bundle was framed to have utilitarian (Muti=4,922, SD=2,382) or hedonic

product role (Mhed=5,291, SD=2,328) was found to be insignificant as well, thus no main effect

of the product role on product bundle selling decision was detected (F(1, 222)=1,156, p=0,284). The moderation effect of product role on relationship between anthropomorphism and product bundle selling decision was found to be without significance (F(1, 222)=1,747,

p=0,188). Further decomposition of this interaction effect to simple effects revealed the

(27)

27 effect of moderation was detected (see Figure 2) albeit the differences between conditions were still not statistically different from each other to provide any support for the predicted effect (Manthro_uti=4,848, SD=2,457 vs. Manthro_hed=5,603, SD=2,240).

Figure 2: Product bundle selling decision based on manipulation of product role and

anthropomorphism – combined sample

Further details of the main study results for the combined sample can be found in

Appendix E.

Since the success of the manipulation checks differed for in-person and online data samples, ‘data collection type’ was added as additional variable to previous 2x2 ANOVA to form 2x2x2(in-person/online) ANOVA. This tested whether different ways of data collection had any influence on the dependent variable ‘product bundle selling decision’. The results of this analysis showed no main effect of data collection type (F(1, 218)=0,037, p=0,848). However, a significant interaction was discovered between anthropomorphism, product role and the data collection type on the product selling decision (F(1, 218)=10,140, p=0,002). Therefore, a subsequent analysis of anthropomorphism and product role effects needed to be constructed for both in-person and online sample.

Table 4: Between subjects of anthropomorphism and product role effects on product

bundle selling decision - in-person and online sample

In-person sample df Mean square F Sig.

(28)

28

Online sample df Error F Sig.

anthro_cnd 1 1,183 0,203 0,653

role_cnd 1 0,292 0,050 0,823

anthro_cnd*role_cnd 1 9,387 1,611 0,207

Error 116 5,826

No main effects of anthropomorphism nor product role were found for the product bundle selling decision in either of the samples (see Table 4). A significant interaction effect was found only for the in-person sample (F(1, 102)=10,945 , p=0,001). While observing simple effects in in-person sample, a significant difference between means of product selling decision was found was for utilitarian and hedonic product role in anthropomorphism condition (F(1, 102)=10,965 , p=0,001) but not in the control condition (F(1, 102)=2,025, p=0,158). Therefore, consumers were significantly less likely to separate anthropomorphized product bundles which had hedonic role (Manthro_uti=4,304, SD=2,530 vs. Manthro_hed=6,265, SD=1,781; see Figure

3).

anthropomorphism - in-person sample

Simple effects of product role manipulation in online sample were insignificant in both anthropomorphism (F(1, 116)=0,591, p=0,444) and control condition (F(1, 116) =1,037, p=0,311). Despite the insignificant difference between product roles in anthropomorphism condition, the effect appeared as reversed (Manthro_uti=5,140, SD=2,396 vs. Manthro_hed=4,667,

(29)

29

anthropomorphism - online sample

Further details from main study results of in-person and online sample can be found in

Appendix F.

Overview of main study results

Overall success of manipulation checks of anthropomorphism and product role for the combined sample, also for the in-person and online sample separately, was found to be insignificant except for some conditions (i.e. utilitarian framed conditions). However, this significance found in these particular conditions is not useful to conclude that any support found for the predicted effects (in any sample) derived from the manipulation of the conditions.

The analysis of predicted main and interaction effects for the combined sample (in-person sample combined with online sample) yielded results supporting the direction of predicted effect in Hypothesis 1 and Hypothesis 2 albeit the effects were not significant to provide any support for these hypotheses.

After splitting the sample by its data collection type, in-person and online samples were found to differ in some results. The predicted main effect of anthropomorphism on product bundle selling decision was found to be insignificant (Hypothesis 1 not supported) as well as the main effect of the product role in the in-person sample. However, the interaction of anthropomorphism and product role was found to have a significant effect on the product bundle selling decision. With the interaction effect being significant and in the predicted direction, the Hypothesis 2 was fully supported for the in-person sample. Therefore, only Hypothesis 2 can be supported for in-person sample.

For the online sample, the results of all effects (main and interaction) were found to be insignificant to support either of hypotheses. Contrary to the predicted effects of Hypothesis 2

(30)

30 confirmed in the in-person sample, the direction of the predicted interaction effect was reversed yet still statistically insignificant to provide support for this finding. Therefore, Hypothesis 1 and 2 were not supported for online sample.

All in all, consumers were not found to be less willing to separate the products in an anthropomorphized product bundle in exchange for higher profits derived from selling the products separately. However, product role was found to influence this relationship but only for the in-person sample. Here, the consumers were more likely to separate the products in an anthropomorphized utilitarian product bundle than for the anthropomorphized hedonic product bundle when the option for getting higher profits derived from selling the products separately was presented.

Exploratory analysis

Difficulty of choice

As proposed by finding of Hanselmann and Tanner (2008), the difficulty of choice is harder when two secular values (wholeness of product bundle vs. money) are put against each other rather than secular and sacred (wholeness of family of humanized products vs. money). For the in-person and online sample, the differences between anthropomorphism and control conditions were found to be significant only between hedonically framed groups; however, the direction was reversed for the online sample (see Table 5).

Table 5: Mean differences in respondents’ perceived difficulty of choice between anthropomorphism

and control conditions - in-person and online sample In-person

Online

anthro_uti 5,609 1,530 0,463 anthro_uti 5,837 1,290 0,401 control_uti 5,917 1,442 control_uti 5,538 1,476 anthro_hed 6,441 1,050 0,036 anthro_hed 4,958 1,681 0,043 control_hed 5,640 1,753 control_hed 5,778 1,340

1=very difficult, 7=very easy

(31)

31 more difficult in the anthropomorphism condition where according to the aforementioned authors, should be less difficult.

Confidence in choice

No significant difference was found between respondents’ confidence in choice to either sell product bundle as one package or sell it separately between anthropomorphized and control conditions in either of the samples (p>0,1; see Appendix G). However, respondents’ confidence in choice was found to be significantly correlated with the perceived difficulty of choice in both in-person (r=0,620 , p<0,01) and online sample (r=0,549 , p<0,01). Considering the scale being 1= ‘not confident at all’/ 7=’very confident’, an evidence was found that the less difficult the respondents perceived the choice to be, the more confident they were in their choice regardless of the anthropomorphism.

Caring about the product bundle

ANOVA analysis revealed respondents’ answers to the question ‘Does it matter to you

if your Drommars family/furniture set goes to a good home or caring user(s)?’ to significantly

differ between anthropomorphism and control conditions only in the in-person sample. However, the significance was only marginal for the utilitarian-framed group (p=0,008); see

Table 6). For the in-person sample, the results align with the finding of Waytz, Cacioppo, and

Epley (2010) who found that people tend to care about anthropomorphized products more than for those which are not humanized. Therefore, further evidence for the presence of anthropomorphism in this study can be provided only for the in-person sample.

Table 6: Mean differences in respondents’ care about product bundles going to good home or caring

user(s) between anthropomorphism and control conditions - in-person and online sample In-person

Online

anthro_uti 3,696 2,010 0,054 anthro_uti 4,419 1,967 0,112 control_uti 2,500 2,043 control_uti 3,615 2,137 anthro_hed 4,941 2,059 0,008 anthro_hed 3,458 2,126 0,202 control_hed 3,440 2,059 control_hed 4,185 1,882

1=does not matter at all, 7=matters very much

(32)

32

GENERAL DISCUSSION

People often sell things they do not use because they want to maximize their profit. In order to decrease their effort in doing so, they put the products in bundles. Therefore, when someone is presented with an option to sell the no longer wanted items separately (initially listed for sale as one bundle) for the same effort but increased monetary value, it should be a no brainer for the seller who follows the mindset of profit maximization. However, this study presents two factors, namely anthropomorphism and product role, which were hypothesized and later confirmed to have an impact on the consumer decision-making process in a way that profit maximization seems to no longer serve as a priority during product selling decision.

Previous studies of anthropomorphism revealed consumers to act towards anthropomorphized products as if they would react to humans in social interactions (e.g. Ahn, Kim and Aggarwal, 2014; Chandler and Schwartz, 2010; Luczak, Roetting and Schmidt, 2003). This reaction also includes showing more care and concern for these entities (Waytz, Cacioppo, and Epley, 2010) which was also discovered to occur in the current study. Respondents in this study were found to care about whether their products are going to good home or caring user after the sale. This research used a product bundle manipulated to appear as a humanized family which in the view of majority of societies, should not be separated or else, in most cases, would be met with emotionally and socially damaging consequences (e.g. Derluyn and Broekaert, 2008). Following this reasoning, the value of wholeness of humanized family was predicted to be treated as a sacred value which by the social standards should always have bigger weight over the money when put into trade-off (Tetlock, McGraw and Kristel, 2004) and thus this decision should not be difficult (Hanselmann and Tanner, 2008). However, current research is not able to provide significant evidence that it indeed is the value people take as sacred. The participants’ reports of level of difficulty of the trade-off differ not only between observed samples but between product roles as well.

(33)

33 rather than anthropomorphized hedonic product bundle which is less likely to be separated for a prospect of higher profits. Findings of this study suggest that in case consumers perceive the product bundle to have only practical functions and characteristics (e.g. durability, easy maintenance), it makes the anthropomorphized product bundle to appear as some group of humanized products which exists to fulfill some task only, therefore its value would not diminish if they were to be sold separately. Putting this into a human example, one would be more likely to separate already established group of maintenance guys, if it meant more savings/profits for the consumer since they are already interacted with for their utility anyway. On the other hand, consumers are more reluctant to separate anthropomorphized product bundles which at some point made them feel happy or satisfied. This might be explained by people not wanting to separate some human entity which brought them these feelings after they already decided on what is its combined value, and getting less or more money for separating it would feel wrong. Certainly, one would not separate family for money because it already has its own combined value of its members which feels wrong to decompose and change for own benefit.

Limitations

(34)

34 Secondly, it cannot be safely concluded that the predicted effects which were found to be supported in this study occurred due to the manipulations of different conditions since both anthropomorphism and product role manipulation checks failed. The scales used to test the manipulation of anthropomorphism were previously applied only in studies where the products tested were manipulated visually (on a picture) to have human physical characteristics (e.g. Kim and McGill, 2011). In hindsight, these measures might not be as useful for detecting anthropomorphism induced only by textual descriptions. Using speech bubbles appointed to the picture could have been a potential factor which increases the significance of anthropomorphism manipulation as it points the speech coming directly from the object.

Lastly, despite the success of manipulation check of product role in the pre-test, the check for the main study might have failed due to different sample consisting of older participants or due to the changes in length of the text already used in the pre-test study. However, coming from the results, the main issue might have lain in inappropriate choice of scales measuring hedonism since ‘fun/not fun’ and ‘exciting/dull’ are more less applicable for experiences rather than opinions which one might have about furniture. Out of the three scales, the one measuring enjoyability of the product seemed to perform the best.

Implications

This study provides an interesting insight and addition to existing literature of psychology of consumers and their decision-making behavior. From the utility-maximizing, economic view of people trying to maximize their limited resources, the behavior found in the current study might seem irrational. However, decision-making can be influenced by both economical and psychological reasons. In the current research, the sellers who perceive the product bundle they are selling as a humanized family providing hedonic satisfaction, are less willing to separate the product bundle for higher monetary profits. This implies that people are willing to sacrifice money for keeping their conscious clear of potential damage to personal values (disrupting happy family). Consumers act in such manner despite the fact that this belief runs against rationally explainable behavior of products indeed being just inanimate items without any consciousness, memories, social bonds or feelings. However, this ‘irrationality’ presents a positive societal implication that people are still generally more inclined to value family (or humans) wellbeing over monetary benefits, even if the family is indeed represented by monetizable products and not living humans.

(35)

35 that humanized family of utilitarian products is more likely to be separated by the seller, the buyers can take advantage of this during negotiations. By stressing the practical functions of the products and how the bundle can indeed be separated without causing deterioration of functionality of each product in the bundle, buyers can potentially change or at least have an influence on the seller’s decision. For the sellers on the other hand, the only way in which they can get potentially full financial benefit out of the sale, is to assign the selling decision to someone with rational thinking, someone who does not see the products as something resembling humans.

Finally, this study extends the already existing but fairly new field of research of anthropomorphism and its influence on consumer decision-making by observing potential reasons why some sellers might not strive for maximization of their profits. Additionally, its finding on preferred selling option for products to potential new users adds on to the limited and neglected research in consumer behavior during product disposal stage.

Further research

It would be beneficial to replicate this study by testing different product bundles and potentially apply the study to different study design (e.g. laboratory setting) to reflect more the real selling situation where the sample would consist of people actually selling their possessions in bundles. If the study design of this study was to be replicated, it might be reasonable to apply the attention check and modification to the study design mentioned in the limitations section.

(36)

36 on product selling decision, this study might be extended in a new direction. It might bring new insight into existing influences on consumer decision-making and to what extent social norms and individual values reflect in interactions with anthropomorphized product bundles of different product roles.

Conclusion

UNIVERSITY OF GRONINGEN FACULTY OF ECONOMICS AND BUSINESS