Anthropomorphism and the anthropomorphised relationship norm as tool to increase the acceptance of imperfect products Research to support the reduction of product waste

(1)

Anthropomorphism and the anthropomorphised relationship

norm as tool to increase the acceptance of imperfect products

Research to support the reduction of product waste

By

(2)

Anthropomorphism and the anthropomorphised relationship

norm as tool to increase the acceptance of imperfect products

Research to support the reduction of product waste

By

MARIJKE TADEMA

University of Groningen Faculty of Economic and Business

MSc Marketing Management Master Thesis

January 15th, 2018

Supervisor: Dr. Wan Second Supervisor: Dr. Leliveld

Trompsingel 15-21 9724 CX Groningen

(3)

Abstract

With striving for perfection in life and possessions, product waste becomes more and more a problem of today’s world. Products with superficial damages are rejected because humans have the natural tendency to protect themselves for any health or safety concerns, the contamination effect and the inferences consumers make about the connection between the appearance and the functionality. The goal of this study to find out whether the phenomenon anthropomorphism, and an anthropomorphised relationship norm, can contribute to the reduction of product waste by increasing the acceptance of imperfect products. Anthropomorphism is used because it is about humanizing non-human entities and it is psychological healthy for humans to accept that things are not always perfect. An advertisement for a water bottle was presented to 320 respondents in an online survey, whereas this water bottle was manipulated with anthropomorphism, the relationship norm and imperfections. In contrast with existing literature and predictions based on prior research, this study did not find evidence that anthropomorphism is able to increase the acceptance of imperfect products. However, the anthropomorphised relationship norm, whether it is seen as an equal partner or as a servant, shows a significant effect on the acceptance of imperfect water bottle.

Keywords: Anthropomorphism, Imperfect Products, Marketing, Product Waste, Relationship norm, Partner, Servant.

(4)

Acknowledgements

I would like to express my gratitude to Dr. Jing Wan, for taking time to guide me the whole process with constant support and feedback. I am thankful for getting the opportunity to conduct a study about a subject that concerns and interests me.

Second, I would like to thank my family and friends who have always been there for me the past few months. During the ups and downs, they always encouraged and supported me when I needed to make essential decisions and listened to me when I needed someone to. Furthermore, my fellow students from my research group are worthy to be thankful for. Together we made it possible to do this research, and I am grateful that during the process we were able to help and inspire each other.

Last but not least, I would like to thank all respondents who take some time to fill in my questionnaire. Without all these respondents, I would not have succeeded in writing this master thesis. A special thank you to all friends and family that helped me to distribute the questionnaire, to get the required number of respondents.

(5)

Table of content

Abstract ... 3

Acknowledgements ... 4

Table of content ... 5

Table of content: Appendices ... 6

Introduction ... 8

1. Theoretical framework and conceptual model ... 11

1.1 IMPERFECT PRODUCTS ... 11

1.2 ANTHROPOMORPHISM ... 12

1.3 THE RELATIONSHIP WITH THE ‘OTHER’ ... 15

1.4 CONCEPTUAL FRAMEWORK ... 17 2. Methodology ... 20 2.1 STUDY DESIGN ... 20 2.2 MANIPULATIONS ... 21 2.3 MEASUREMENTS ... 22 3. Results ... 24 3.1 MANIPULATION CHECK ... 24 3.2 TESTING HYPOTHESIS 1 ... 25 3.3 TESTING HYPOTHESIS 2 ... 26

3.4 ADDITIONAL TESTING FOR HYPOTHESIS 2 ... 28

(6)

Table of content: Appendices

APPENDIX I: Advertisements ... 43

1. Condition 1 and 2: without any manipulation ... 43

2. Condition 3 and 4: anthropomorphism manipulation ... 43

3. Condition 5 and 6: Partner relationship manipulation ... 44

4. Condition 7 and 8: Servant relationship manipulation ... 44

5. Condition 2, 4, 6 and 8: Imperfection manipulation ... 45

APPENDIX II: Questionnaire ... 46

1. Dependent variable ... 46

2. Anthropomorphism manipulation check ... 46

3. Imperfection manipulation check ... 47

4. Partner/Servant manipulation check ... 47

5. Demographics and other ... 47

APPENDIX III: Descriptive statistics ... 48

1. Number of respondents that finished the questionnaire ... 48

2. Gender ... 48

3. Age ... 48

4. Channels ... 49

5. Familiarity ... 49

APPENDIX IV: Analysis ... 50

1. T-test for anthropomorphism manipulation check ... 50

2. Descriptive statistics for mean of anthropomorphism manipulation check ... 51

3. Cronbach alpha for the dependent variable ... 51

(7)

5. Hypothesis 1a ... 52 6. Hypothesis 1b ... 53 7. Hypothesis 2a ... 54 8. Hypothesis 2b ... 55 9. Hypothesis 2c ... 56 10. Hypothesis 2d ... 57

11. T-test between imperfect partner and imperfect servant ... 58

12. Additional testing 1 ... 58

(8)

Introduction

FLAWSOME:

[flô-sem] adjective

an individual who embraces their ‘flaws’ and knows they are awesome regardless.

In today’s world, humans strive for perfection. We are looking for the perfect job and house, only buy products that perfectly fit our needs and even more important, we want to look perfect, including perfect hair and body weight. Creating a perfect life with possessions tend to be very addictive since consumers tend to repair their feeling of being imperfect with those perfect possessions (Howard, 2011; Bury, 2013). According to Brown, research professor at the University of Houston Graduate College of Social Work, there is one simple reason why people strive for perfection: it will minimise the pain, judgment and shame, and is able to protect us. As we are all busy with living the perfect life, is perfection what we really admire in others? No. We do not expect others to be perfect and even love people more who are real, authentic, messy and imperfect (Brown, 2010). Perfection is subjective and depends on one's social, moral, cultural, personal standards, and world’s view (Martz, 2011). Hence, as Martz (2001) states: ‘as humans are multidimensional concepts and society could never agree upon what perfection is, a person could never be perfect’.

(9)

appearance comes with problems for companies too. The e-commerce business already has alarming numbers of returns. A research by TrueShip shows that 30% of all orders are returned, whereas 20% of these returns is due to product damages (Bussiness2Community). Consumers do not accept these products because of their superficial damages and companies try to deal with large numbers of product waste. Even though these damages, also called imperfections, are superficial and do not harm the functionality, consumers try to avoid these imperfect products from the natural tendency to protect themselves for health and safety concerns (White et al., 2016).

As White et al. (2016) suggest in their implications, there are some ways to diminish the effects of superficial imperfections on products, for example anthropomorphism. Anthropomorphism is a psychological process where people give non-human objects human-like characteristics, motivations, intentions, or emotions (Epley et al., 2007). By applying norms and rules from the social world on products, anthropomorphism is part of the evaluation and judgement process and will help consumers to process the product more efficiently (Long et al., 2012). As humans could never be perfect and we are willing to accept imperfections because it is psychological and mentally healthy, this research will focus on how anthropomorphism is able to reduce the avoidance tendency of imperfect product and make people accept these imperfect products.

(10)

relationship norm of an equal partner of a master-servant connection to see if this will moderate the effect of anthropomorphism on imperfect products. When the consumer sees the product as an equal partner they will co-create the benefits together and in the servant role, the consumer is dominant to the brand and product, whereas it needs to work for him to create benefits. (Aggarwal and McGill, 2012). As materialist use possessions to judge themselves and others, and prefer a dominant servant relationship with a product or brand (Kim and Kramer, 2015), the expectation is that anthropomorphism will increase the acceptance less for imperfect product in a servant role, rather than when there is an equal partnership.

Since distributors, e-commerce and retail are dealing with huge numbers of product waste, this research will focus on finding a way to let consumers accept superficial imperfection on products. By aiming to increase the acceptance of these products; this research will help companies to reduce the number of product returns and waste. Besides the financial benefits for companies to reduce products waste, this paper has some sustainability implications, which has my personal interest.

(11)

1. Theoretical framework and conceptual model

1.1 IMPERFECT PRODUCTS

Products with superficial imperfections getting more and more attention in marketing research since consumers tend to reject these products, even though these imperfections do not affect the functionality (White et al. 2016; Argo et al., 2006). As for the rest of this study, an imperfection is defined by the Business Dictionary: ‘Departure of a quality characteristic from its intended level or state. An imperfection, however, does not affect the conformance of the product or service with its specifications or usability.’

Although the superficial imperfection does not intend any health or safety risk and does not even harm the functionality, White et al (2016) found evidence that consumers try to avoid these products. They state that it derives from humans’ natural tendency to protect themselves from stimuli that have the potential to pose health and safety threats. Another explanation for why consumers avoid imperfections is the so called ‘contamination effect’. This effect shows that consumers perceive the imperfect product as touched by another person, which elicit disgust and makes consumers try to avoid these imperfect products (Argo et al., 2006). Interestingly, marketing tools could diminish the negative effect of imperfections, White at al. (2016) imply to label the product as organic, whereby, the superficial imperfection became acceptable. This could be explained by the inferences people make about the label ‘organic’ (White et al, 2016). Informing consumers that the mistake was made intentionally or that the product is handmade, makes the imperfections more acceptable and even more attractive (Reich et al., 2017; Fuchs et al, 2015). Reich et al. (2017) emphasise in their research that uniqueness is an important driver of accepting or even prefer a product with an imperfection.

(12)

(Kurosu and Kashimura, 1995; Tractinsky et al., 2000). Hoegg, Alba and Dahl (2010) found the importance of the appearance and design in processing the performance of an object and how form and design need to be in line with the performance. Through appearance and design, products are able to communicate a certain meaning, let consumers make inferences about usability and functionality, give information, change the attitude and motivates to buy, and even activate certain emotions (Blijlevens et al, 2009; Kurosu and Kashimura, 1995; Kumar and Garg, 2010; Underwood et al., 2001). Thus, because people make inferences about a product based on the appearance and link superficial imperfections in the appearance with failing functionality, the contamination effect and humans’ natural tendency to avoid imperfections, consumers do not accept products with imperfections and companies are burdened with problems of product waste.

1.2 ANTHROPOMORPHISM

(13)

perceive the functionality of objects that are anthropomorphised. Interestingly, they predict that ‘whereas objects are primarily evaluated based on how well they function, functionality is far less important when evaluating people—and we assume that the same may hold for anthropomorphised objects’. The conclusion from their view is that when an imperfect object is anthropomorphised the link between the superficial imperfection and the functionality becomes far less important and the interpersonal relationship becomes more important.

Anthropomorphism is the ‘tendency to imbue the real or imagined behaviour of non-human agents with non-human-like characteristics, motivations, intentions, or emotions’ (Epley et al, 2007). The Oxford Dictionary describes it as ‘attribution of human characteristics or behaviour to a god, animal, or object’ or even more simple described by MacInnis and Folkes (2017) ‘humanizing non-human entities’. According to MacInnis and Folkes (2017), anthropomorphism is a psychological process, in the mind of the consumer, that goes beyond the observable and descriptive characteristics of a product, and is all about unobservable characteristics using human-like descriptors from the social world. Anthropomorphism stems from the human tendency to make inferences and judgements about the environment and can be found in all kind of situations, ranging from religious belief to effective marketing campaigns (Guthrie, 1993). Examples are religions where gods represent a human form, Disney movies where animals are human characters like Micky Mouse and anthropomorphised representations of brands like The Michelin Man and the talking M&M Mascots.

(14)

interactions and relationships (Guthrie, 1993). Based on the theory of Guthrie (1993), Epley et al. (2007) developed a three-factor model with the motivational triggers to understand why people anthropomorphise and how to increase or decrease the extent in which people anthropomorphise. One factor is Elicited Agent Knowledge, this is when knowledge of humans is accessible and applicable to the non-human entity. To activate the ‘human’ schema and the agent knowledge, and, to be more likely to perceive the agent as human-like, brands could use visual cues, verbal devices and/or rhetorical devices (MacInnis and Folkes, 2017). Besides this cognitive mechanism, there are two other mechanisms which trigger the motivation to anthropomorphise: effectance and sociality. The effectance motivation explains how people want to interact effectively with non-human entities and explain, understand and predict these entities and their behaviour. The sociality motivation explains how people try to satisfy the need and desire for social connections. This factor triggers the way people employ social norms used in social interactions on non-human entities (Wan and Aggarwal, 2015). For example, when people lack social connections, they often compensate this feeling by treating non-human objects as they are human-like (Zlotowski et al., 2014). The quasi-social experience that anthropomorphised brands and products provide are even able to replace person-to-person relationship (Aggarwal and McGill, 2007; Fournier, 1998).

(15)

worthiness of a brand, which all positively influence the performance and acceptance of a brand or product. Furthermore, anthropomorphism has the ability to increase the emotional bonding with the object and consequently affect the evaluation positively (Kim and McGill, 2011).

1.3 THE RELATIONSHIP WITH THE ‘OTHER’

As we have seen above, consumers use norms and processes from the social world in their decision-making process and evaluation of the product that is anthropomorphised. An important factor in the social world is how people tend and intend to form relations with others and how they process information in different kinds of relationships (Kim and Kramer, 2015). Since consumers fully accept the attempts of advertisers to anthropomorphise brands and products, consumers are willing to see them as part of their social culture and the relationships they form in their social culture (Fournier, 1998). A relationship could be defined as series of interactions between two parties, placed in a series of time and with a probability of continuation in the future (Hinde, 1976). These relationships involve a certain degree of hierarchy between the two entities that are involved (Gruenfeld et al., 2008). Over the last few decades, consumer research has focussed on the consequences for brands and products when consumers mirror their social relationships, the rules and norms that go with that kind of relationships, and, how these influences the evaluation process differently. Although a lot of the research has focussed on brands, this could be applied to the products of these brands as well. How brands and products carry out and implement these relations consistently, will influence the consumers’ brand attitude and behaviour (Aggarwal, 2004). Moreover, research has shown that relationships with brands and products depend on if a particular goal could be met by obtaining the object and the overlap between goals and interpersonal motives of the consumer and the brand (Fitzsimons and Shah, 2008; Long et al, 2012).

(16)

relationships with exciting brands develop in short-lived flings (Aaker, Fournier and Brasel, 2004). Fournier (1998) shed light on how consumers tend to form relations with brands and products based on relationships from the social context. Based on six dimensions, for example, voluntary/involuntary or long-term/short-term, they developed different kinds or relationship types like casual acquaintances, close friendships, arranged marriages, committed partnerships, flings, one-night stands, and secret affairs.

(17)

their acquisition lead to happiness and life satisfaction”. Kim and Kramer (2015) conclude that consumers with a low level of materialism, are better in developing equal partnerships that are based on trust. A materialist will prefer servant brands and products, for the reason that they seek for dominance and power over others, even over the other entity in the relationship and their consumption (Bartholomew and Horowitz, 1991; Burroughs and Rindfleisch, 2002; Belk, 1985).

1.4 CONCEPTUAL FRAMEWORK

(18)

H1: An imperfect product will be more accepted by a consumer when it is anthropomorphised than when the imperfect product is not anthropomorphised.

Figure 1: Main mechanism of this research

(19)

H2: When the anthropomorphised imperfect product is seen as a ‘Partner’, the acceptance of the product will be higher than when the product is seen as a ‘Servant’.

(20)

2. Methodology

The goal of this study is to find whether anthropomorphism is able to increase the acceptance of imperfect products. If anthropomorphism is able to increase the acceptance of imperfect products, this will eventually reduce product waste. In the following section, it will be explained what the study design of this research is and how the conceptual model is manipulated and measured.

2.1 STUDY DESIGN

For this study, an online experiment is conducted. The online survey is spread on social media channels (e.g. Facebook and LinkedIn) by myself, friends and family members and QR-cards are handed out on the streets in Groningen. These QR-QR-cards contain a message with the request if people would like to help me with my master thesis research and a QR-code that leads potential respondents to the questionnaire. This is aimed to have a diverse representation of society within an uncontrolled environment, which will increase the external validity of this study. Interestingly, the approach with the QR-cards does not work and half of the respondents (51,9%) received the questionnaire through social media channels (see Appendix III:4). However, 20,3% used the open box and mostly filled in ‘WhatsApp’, which theoretically is a social media channel.

(21)

female and 33,4% male (Appendix III:2). The average age is 35 years (Mage = 35.1 SDage = 17.79) and more than half of the respondents is between 18 and 34 years (41.6% + 19.7%). Of all respondents, 51,9% is not a student and 47,8% do not know the experimenter of this study (see Appendix III:5). In Qualtrics the within-blocks and between-blocks randomization are used to distribute the respondents over the eight different conditions, each participant was assigned to only one of the eight conditions (see table 1). First, participants see the advertisement linked to one of the eight conditions, afterwards, they answer the questions (see Appendix II).

PRODUCT

Perfect Imperfect Total (N)

ANTHRO No C1 (N=46) C2 (N=38) 84

Yes C3 (N=42) C4 (N=37) 79

Yes (Partner) C5 (N=35) C6 (N=41) 76

Yes (Servant) C7 (N=37) C8 (N=43) 80

Total (N) 160 159

Table 1: Distribution of respondents among all conditions (Appendix IV:2)

2.2 MANIPULATIONS

Anthropomorphism: Respondents in the non-anthropomorphism conditions see the advertisement with just the normal water bottle, with a product-like name and a text box about the advantages of using a water bottle. In this study, anthropomorphism is manipulated by giving the water bottle a smiling face and the text box is changed in a speech bubble. The name of the bottle is changed in a human-like name (Billy Bottle) and the advantages are written from a first-person perspective (adapted from Aggarwal and McGill, 2007). The four different advertisements with the anthropomorphism manipulation can be found in Appendix I.

(22)

was drawn to their condition. The imperfect product conditions, also called the experimental group, have an extra announcement before starting to answer the questions. In these conditions, the respondents see a visual representation of the bottle with a scratch and the comment that they should keep in mind that the water bottle has a little scratch on the outside (adapted from White et al., 2016). It is important that the chosen imperfection should not impact the functionality of the product. The visual representation of the water bottle with the scratch on the outside can be found in Appendix I.

Partner vs. Servant: Only the anthropomorphised conditions are used to manipulate the relationship norm. Using manipulations for the distinction between a partner or a servant role have already a certain degree of anthropomorphism built in, hence applying relationships from the social world on products is part of anthropomorphism. This is why it is only used in the experimental group of anthropomorphism. In the partner role, the message in the speech bubble is framed in a ‘works with you’ form, where the emphasis is on how the bottle works together with you to keep you hydrated. The servant role is about how it ‘works for you’ and the text in the speech bubble emphasises how the water bottle works for you to keep you hydrated (adapted from Kim and Kramer, 2015) (see Appendix I).

2.3 MEASUREMENTS

(23)

Control variables: A product-related control question is built-in the questionnaire and asked after the measurements for the dependent variable, to determine the general use of a water bottle for the respondent. This makes a distinction in the results of the analysis with people who do use water bottles frequently and who not. (see Appendix II).

Manipulation check: To check the anthropomorphism manipulation, Epley, Waytz and Cacioppo (2007) set up a manipulation check. Respondents indicate in which extent they think the water bottle has a mind of its own, have intentions, have free will, have consciousness, have the ability to experience emotions, and have a personality (1 = strongly disagree, 7 = strongly agree). To get an impression about the effectiveness of the imperfection manipulation, the experimental group with the comment of the imperfection gets a question about their feelings towards the scratch (annoyed vs not annoyed, not happy vs. happy). To measure the manipulation of the partner versus servant role, the participant gets four statements presented that direct towards a partner or a servant kind of relationship, where they need to indicate in which extent they agree (1= strongly disagree, 7 = strongly agree) (see Appendix II).

(24)

3. Results

3.1 MANIPULATION CHECK

Before starting of the analysis, the manipulations of this research will be checked, to find out whether these manipulations did work out. Adapted from Aggarwal and McGill (2007) the water bottle was manipulated with aspects like a smiling face and talking from a first-person perfective. To conclude if this manipulation worked out, the respondents indicated to what extent they agreed upon quotes about the human schema (adapted from Epley et al., 2007). Table 2 shows a distribution of the mean of this manipulation check per condition. To find out if there is any difference in the impact of the manipulation between anthropomorphism (n=224) and non-anthropomorphism (n=81), an independent sample t-test shows a significance that is close to a managerial significant difference between these two groups (t (303) =-1.608, p = .109) (see Appendix: IV:1). A between-subject ANOVA for the 2 (perfect vs. imperfect) x 4 (not anthropomorphised vs. anthropomorphised vs. anthropomorphised-partner vs. anthropomorphised-servant) model, on 5% confidence level, does not show any significant main effects of anthropomorphism, nor an interaction effect with the imperfection.

PRODUCT

ANTHRO No 1.79 (N=44) 1.54 (N=37) 1.68 (81)

Yes 2.11 (N=39) 1.63 (N=37) 1.87 (76)

Yes (Partner) 1.81 (N=32) 1.89 (N=41) 1.86 (73)

Yes (Servant) 2.18 (N=36) 1.83 (N=39) 1.99 (75)

Total (N) 1.97 (151) 1.73 (154)

Table 2 Mean of anthropomorphism check per condition (see Appendix IV:2)

(25)

the respondents who were assigned to the imperfect conditions. Within this condition, a between-subject ANOVA could not find any significant difference in the impact of the scratch on the feelings towards the water bottle. For the manipulation check of the relationship norm, the respondents indicated to what extent that agreed upon quotes that where framed from a partner or a servant perspective. Measuring the mean of these answers and conducting an independent sample t-test between partner and the rest and servant and the rest, could not confirm any significant differences in the results of the relationship norm manipulation. A one-way ANOVA on the total mean of the relationship norm manipulation on the partner, servant and other condition, could indeed not confirm any significant differences.

3.2 TESTING HYPOTHESIS 1

The first hypothesis says: An imperfect product will be more accepted by a consumer when it is anthropomorphised than when the imperfect product is not anthropomorphised. To test this hypothesis, a 2 (anthropomorphised vs. non-anthropomorphised advertisement) x 2 (perfect vs. imperfect water bottle) between-subject ANOVA was conducted, with the acceptance of the product as dependent variable. To measure the acceptance of the product, respondents answered questions about their overall impression of the water bottle, the degree of liking, the perceived quality and functionality of the product and the willingness to buy. To make one variable that is measuring the acceptance of the product, a Cronbach Alpha reliability test was done with the five items, named previously (5 items; α = .835). The items were combined in one new variable that represents the acceptance of the product (see Appendix IV:3).

(26)

Subsequently, the first hypothesis cannot be confirmed. Furthermore, the general use of the water bottle was included in a between-subject ANCOVA, to see whether this influence the acceptance of the water bottle. General use of the water bottle is used as a covariate because it is possible that consumers who never use a water bottle have low involvement and interest in the way the bottle is positioned. However, this control factor did not influence the main effects, nor the interaction in this analysis (see Appendix IV:6).

3.3 TESTING HYPOTHESIS 2

The second hypothesis says: When the anthropomorphised imperfect product is seen as a ‘Partner’, the acceptance of the product will be higher than when the product is seen as a ‘Servant’. This contains a moderator and to test this hypothesis a 2 (perfect vs. imperfect) x 4 (not anthropomorphised vs. anthropomorphised vs. anthropomorphised-partner vs. anthropomorphised-servant advertisement) between-subject ANOVA was conducted, where the acceptance of the water bottle was used as the dependent variable. A univariate analysis, on a 5% confidence level, shows that there is still a main effect between the perfect and imperfect conditions (F (1, 318) = 22,78, p = .000). Interestingly, there is a significant interaction effect between the imperfection and anthropomorphism (F (3, 318) = 2,49, p = .005) but still no statistically significant main effect of anthropomorphism in the 2 x 4 between-subject ANOVA (F (3,318) = 4,31, p = .061). However, we could say that this main effect is marginally significant (see Appendix IV:7).

(27)

and the interaction effect between perfect and imperfect and the relationship norm (F (1,155) = 4,4, p = .041), are both statistically significant, on 5% confidence level (see Appendix IV:8).

In table 3, the means per condition are classified. The expectation was that imperfect products in the partner role will have a higher acceptance than imperfect products in the servant role but this cannot be confirmed. From an analysis, we see that imperfect product in an

anthropomorphised servant role has a higher acceptance (Mimperfect/servant = 4.144,

SDimperfect/servant = .966) than the one in a partner role (Mimperfect/partner = 4.132, SDimperfect/partner = .962) (see Appendix IV:4). An independent sample t-test could not confirm significant differences between the imperfect partner and the imperfect servant (t (82) =-.059, p =.953) (see Appendix IV:11).

PRODUCT

ANTHRO No 4.596 (N=46) 3.695 (N=38) 4.188 (84)

Yes 4.371 (N=42) 3.638 (N=37) 4.028 (79)

Yes (Partner) 4.023 (N=35) 4.132 (N=41) 4.082 (76)

Yes (Servant) 4.632 (N=37) 4.144 (N=43) 4.370 (80) Total (N) 4.411 (160) 3.916 (159)

Table 3: mean of the acceptance of the water bottle per condition (Appendix IV:4)

(28)

3.4 ADDITIONAL TESTING FOR HYPOTHESIS 2

Diving deeper into the second hypothesis, this analysis contains some additional tests to understand the effects and differences in this study model. Firstly, the anthropomorphism condition is excluded to look more closely at the effect of the anthropomorphised relationship norm compared to the situation where the product is not anthropomorphised. An additional between-subject ANOVA was conducted for a 2 (perfect vs. imperfect) x 3 (non-anthropomorphism vs. partner vs. servant) model. This univariate test, on a 5% confidence level, found a significant main effect of perfect versus imperfect (F (1,239) = 12,33, p = .001) and an a statistically significant interaction effect between perfect versus imperfection and non-anthropomorphism, partner and servant (F (2,239) = 5,78, p = .004). For the comparison between non-anthropomorphism and the partner and servant conditions, this analysis shows a marginally significance (F (2,239) = 2,39, p = .094) (see Appendix IV:12). As prior analysis confirmed, there is no significant difference between the imperfect partner and imperfect servant conditions. However, after the managerial significance of anthropomorphism in the 2 x 3 ANOVA was confirmed, an independent sample t-test between the non-anthropomorphism condition (Mperfect/non-anthropomorphism = 4.59, SD perfect/non-anthropomorphism = .901) and the

imperfect anthropomorphised relationship norm (Mimperfect/relationshipnorm = 4.13, SD

imperfect/relationshipnorm = .958) was conducted. This analysis could confirm statistically significant difference (t (128) =2,66, p = .008) (see Appendix IV:13).

(29)

within the perfect water bottle conditions, on 5% confidence level (F(3, 159) = 2,903, p = .013) (see Appendix IV:14). To understand more about the differences between condition within this perfect water bottle group, a post-hoc test was made (see Appendix IV:11). This confirms significant difference between non-anthropomorphised/perfect condition and the partner condition (p = .023) and the partner and servant condition (p= .020). More interestingly, there is no significant difference between the anthropomorphism condition and the other conditions within the perfect conditions. A between-subject ANOVA, on 5% confidence level, for only the imperfect conditions found significant differences (F (3,158) = 2,957, p = .0.32) (see Appendix IV:15). Again, a post-hoc test was made afterwards. Interestingly, non-anthropomorphism and non-anthropomorphism do not significantly differ from each other (p = .804), and additionally, partner and servant do not too (p =. 954) (see Appendix IV:15). An independent sample t-test confirms significant differences between the relationship norm (M relationshipnorm = 4.13, SD relationshipnorm =.958) and the remaining of the imperfect condition (M rest = 3.66, SD rest = 1.01) (t (157) = 3,01, p = .003) (see Appendix IV:16).

3.5 RESULTS DISCUSSION

(30)

confirm differences in perfect and imperfect but did find a main effect of anthropomorphism and an interaction effect. Nevertheless, the second hypothesis could not be confirmed, since looking more closely to the means per condition, we see that the servant role is higher for the imperfect product, rather than the expected partner role and there is no significant difference between these two.

(31)

4. General discussion

(32)

advertisement contains a human face, but furthermore it was not presented as a family in a product line or did not engage in human activities (Aggarwal and McGill, 2007; Puzakova et al., 2013).

The second expectation was that the relationship norm would impact the acceptance of the anthropomorphised imperfect products. Anthropomorphism is often seen as a necessary antecedent to form relationships with brands and products, and anthropomorphised relationship could have the role of a partner or a servant (Kim and Kramer, 2015). Based on prior research and literature, the expectation was that an equal partner relationship would have higher acceptance, rather than a servant relationship. Despite the fact that this study found evidence that anthropomorphism and imperfections influence each other on the effect on the acceptance of the water bottle and that these conditions significantly differ from each other, the hypothesis could not be confirmed. Research shows that there is no significant difference between the acceptance of a water bottle that is anthropomorphised in a partner role or that is anthropomorphised in a servant role. An explanation for the fact that there is no difference, could be that whether consumers prefer partner relationships or servant relationships depends on the level of materialism. Kim and Kramer (2015) emphasise that which relationship role a consumer prefers with a product or brand depends on certain personal characteristics that determine the level of materialism. These characteristics could not be manipulated in an advertisement, as it could be that the kind of relationship that the respondent has with the water bottle could not be manipulated.

(33)

Trying to find an explanation for why anthropomorphism did not work at all, but when the water bottle with the scratch represented a certain relationship norm it did work, builds upon the conclusion that in the anthropomorphism condition the human schema and agent knowledge where not activated in a degree that it motivated the respondent to anthropomorphise the water bottle. Kim and Kramer (2015) proclaim a close link and overlap between brand roles and anthropomorphism. In this research, only the anthropomorphism conditions where complemented with a certain brand role, as these reinforce each other and create synergy. Whereas anthropomorphised brands and products incite to engage in relationships, in anthropomorphism an important component of partner and servant kind of relationships (MacInnis and Folkes, 2017; Kim and Kramer, 2015).

4.1 LIMITATIONS

(34)

relationship norms, an explanation for failing the manipulation is that it was not humanised enough to see this object as human and anthropomorphism and this kind of relationship norm are interconnected. This was done by purpose because in preparing the questionnaire, the uncanny valley was considered. This effect is the point when characters are as human that it feels creepy (Burleigh et al., 2013). Furthermore, the relationship norm that was applied in this research was chosen because prior research shows the possibilities and effects within anthropomorphism.

The last limitation of this study is the required sample size for representative results. Considering the eight conditions in this study design and a required 50 respondents per condition, this research required over 400 respondents to be representative. Due to time and money limits, this study became an online study that was distributed on social media by the executive researcher and family and friends. However, in time, it became clear that it was really hard to find as many respondents. Other attempts as QR-codes did not succeed. Eventually, only the control group contained around 50 respondents (Fig. 3). Besides the sample size, the samples were not fully representative for the population. Because the questionnaire was distributed and filled in by family and friends, random selection of respondents was not applicable. Not meeting the required number of respondents per condition and no random selection, influences the reliability and representation of this research.

4.2 IMPLICATIONS

(35)

imperfect products. This could be utilized in the marketing strategies, by focus on the aspects that will increase the chance on a meaningful relationship with their products and brands. But also engineers and product designers could use the factors of the human schema to make it easy for consumers to create a relationship with the brand or product. The goal of this marketing strategy will eventually increase a meaningful connection with the brand and/or products, whereas superficial imperfection will be accepted more frequently. When the acceptance increases, product waste and returns for these companies will decrease, which will be an economical and sustainable advantage. Besides, companies that produce products with recycled materials, where materials are not always perfect, could use the insights of this research to increase the acceptance of their products. However, for more elaborated implications of the effect of anthropomorphised relationship norms, future research is required. Suggestions for future research will be explained in the following section.

4.3 FUTURE RESEARCH

(36)

(37)

References

Aaker, J. L. (1997). Dimensions of brand personality. Journal of marketing research, 34, 347- 356.

______, __, Fournier, S. & Brasel, S. A. (2004). When Good Brands Do Bad. Journal of Consumer Research, 31 (1), 1-16.

Adderholdt, M. & Goldberg, J. (1987). Perfectionism: What's Bad about Being Too Good? Minneapolis; Free Spirit Publishing.

Aggarwal, P. (2004). The Effects of Brand Relationship Norms on Consumer Attitudes and Behaviour. Journal of Consumer Research, 31 (1): 87-101.

_______, __. & McGill, A. (2007). Is that car smiling at me? Schema congruity as a basis for evaluating anthropomorphized products. Journal of Consumer Research, 34 (4), 468- 479.

_______, __. & ______, ___. (2012). Partner and Servants: Adopting Traits of Anthropomorphised Brands. Advances in Consumer research, 38.

_______, __. & ______, ___. (2012). When brands seem human, do humans act like brands? Automatic behavioural priming effects of brand anthropomorphism. Journal of Consumer Research, 39 (2), 307-323.

Argo, J.J., Dahl, D.W., & Morales, A.C. (2006). Consumer Contamination: How Consumers React to Products Touched by Others. Journal of Marketing, 70 (2), 81-94.

Bartholomew, K. & Horowitz. L.M. (1991). Attachment Styles among Young Adults: A Test of a Four-Category Model. Journal of Personality and Social Psychology, 61 (2): 226–

(38)

Belk, R. W. (1985). Materialism: Traits Aspects of Living in the Material World. Journal of Consumer Research, 12 (December), 265–80.

Blijlevens, J., Creusen, M. E. H., & Schoormans, J. P. L. (2009). How consumers perceive product appearance: The identification of three product appearance attributes. International Journal of Design, 3(3), 27-35.

Burleigh, T. J., Schoenherr, J.R. & Lacroix, G.L. (2013). Does the uncanny valley exist? An empirical test of the relationship between eeriness and the human likeness of digitally created faces. Computers in Human Behaviour, 29, 759-771.

Burroughs, J. E. & Rindfleisch, A. (2002). Materialism and Well-Being: A Conflicting Values Perspective. Journal of Consumer Research, 29 (December), 348–70.

Castelli, F., Happe, F., Frith, U., & Frith, C. (2000). Movement and mind: a functional imaging study of perception and interpretation of complex intentional movement patterns. NeuroImage, 12(3), 314-325.

Chamberlain, J.M., & Haaga, D, A, F. (2001). Unconditional Self-Acceptance and

Psychological Health. Journal of Rational-Emotive and Cognitive-Behavior Therapy, 19(3), 163-176.

Chandler, J. & Schwarz, N. (2010). Use does not wear ragged the fabric of friendship: Thinking of objects as alive makes people less willing to replace them. Journal of Consumer Psychology, 20 (2), 138-145.

Delbaere, M., McQuarrie, E. F. & Phillips, B. J. (2011). Personification in Advertising. Journal of Advertising, 40(1), 121-130.

Epley, N., Waytz, A. & Cacioppo, J. (2007). On seeing human: a three-factor theory of anthropomorphism. Psychological Review, 114 (4), 864-886.

(39)

Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer research. Journal of consumer research, 24(4), 343-373.

Fitzsimons, G. M., & Shah, J. Y. (2008). How goal instrumentality shapes relationship evaluations. Journal of Personality and Social Psychology, 95(2), 319-337.

Gruenfeld, D. H., Ena, I.M., Magee, J.,C., & Galinsky, A., D. (2008). Power and the

Objectification of Social Target, Journal of Personality and Social Psychology, 95 (1), 111–27.

Guthrie, S. (1993). Faces in the clouds. New York; Oxford University Press.

Hinde, R. A. (1976). On Describing Relationships. Journal of Child Psychology and Psychiatry and Allied Disciplines, 17: 1–19.

Hoegg J, Alba J.W. & Dahl, D. W. (2010). The good, the bad, and the ugly: influence of aesthetics on product feature judgments. Journal of Consumer Psychology, 20, 419 – 430.

Kim, S. & McGill, A.L. (2011). Gaming with Mr. Slot or Gaming the Slot Machine? Power, Anthropomorphism, and Risk Perception. Journal of Consumer Research, 38(1), 94- 107

Kumar, M. & Garg, N. (2010). Aesthetic principles and cognitive emotion appraisals: How much of the beauty lies in the eye of the beholder? Journal of Consumer Psychology, 20, 485-494.

Kurosu, M., & Kashimura, K. (1995). Apparent Usability vs. Inherent Usability: experimental analysis on the determinants of the apparent usability. Conference companion on Human factors in computing systems, 292-293.

(40)

Long, C., Gable, P., Albee, C., & Boerstler, C. (2012). Brand Are Like Friends: Goals and Interpersonal Motives Influence Attitudes Toward Preferred Brands. Advances in Consumer Research, 38: Dahl, D.W., Johar, G.V. Osselaer, S.M.J. van, Duluth, MN : Association for Consumer Research.

Lundh, L.G. (2004). Perfectionism and acceptance. Journal of Rational-Emotive and Cognitve-Behavior Therapy, 22(4): 251-265.

MacInnis, D.J. & Folkes, V.S. (2017). Humanizing brands: When brands seem to be like me, part of me, and in a relationship with me. Journal of Consumer Psychology, 27(3), 355–374.

Martz, E. 2001. Acceptance of imperfection. Society for Disability Studies, 21 (3): 160-165.

Puzakova, M., Kwak, H., & Rocereto, J. F. (2013). When humanizing brands goes wrong: The detrimental effect of brand anthropomorphization amid product wrongdoings. Journal of Marketing, 77(3), 81–100.

Reich, T., Kupor, D. and Smith, R.K. (2017). Made by Mistake: When Mistakes Increase Product Preference. Journal of Consumer Research, 44(5), 1085-1103.

Richins, M. L., & Dawson, S. (1992). A Consumer Values Orientation for Materialism and Its Measurement: Scale Development and Validation. Journal of Consumer Research, 19

(December), 303–16.

_____, __. (2004). The Material Value Scale: Measurement Properties and Development of a Short Form. Journal of Consumer Research, 31 (June), 209–19.

(41)

Tractinsky, N., Katz, A. S., & Ikar, D. (2000). What is Beautiful is Usable. Interacting with Computers, 13, 127-145.

Underwood, R. L., Klein, N. M., & Burke, R. R. (2001). Packaging Communication:

Attentional Effects of Product Imagery. Journal of Product & Brand Management, 10, 403-422

Wan, J. & Aggarwal, P. (2015). Befriending Mr. Clean: The Role of. Anthropomorphism in Consumer Brand Relationships. In Strong Brands, Strong, 119-134.

White, K., Lin, L., Dahl, D. W., & Ritchie, R. J. B. (2016). When do Consumers Avoid Imperfections? Superficial Packaging Damage as a Contamination Cue. Journal of Marketing Research, 53(2), 110-123.

Złotowski J, Proudfoot, D, Yogeeswaran K, Bartneck C. (2014). Anthropomorphism: opportunities and challenges in human–robot interaction. International Journal of Social Robotics, 7(3), 347-360.

INTERNET REFERENCES

Brown, B. (2010). Want to be happy? Stop trying to be perfect. Accessed at October 27, 2017, available at: http://edition.cnn.com/2010/LIVING/11/01/give.up.perfection/index.html

Business2Community. (2016). E-commerce Product Return Statistics and Trends. Accessed at January 6, 2017, available at: https://www.business2community.com/infographics/e- commerce-product-return-statistics-trends-infographic-01505394

Bury, M. (2013). How Our Obsession With Perfections Makes Us So Imperfect. Huffington Post. Accessed October 22, 2017, available at

https://www.huffingtonpost.com/merissa-bury/perfectionism_b_2457599.html

(42)

(43)

Appendices

APPENDIX I: Advertisements

1. Condition 1 and 2: without any manipulation

(44)

3. Condition 5 and 6: Partner relationship manipulation

(45)

(46)

APPENDIX II: Questionnaire

1. Dependent variable

What is your overall impression of this water bottle?

Very negative 0 0 0 0 0 0 0 Very positive

How much do you like this water bottle?

Not at all 0 0 0 0 0 0 0 Very much

What do you think about the quality of this water bottle?

How willing are you to buy this water bottle?

Not willing to 0 0 0 0 0 0 0 Willing to

What do you think about the functionality of the product?

Do you normally use a water bottle? Yes

No

2. Anthropomorphism manipulation check

Based on the advertisement you just saw; please indicate in which extent you agree with the following statements:

The water bottle has… …a mind of its own

Strongly disagree 0 0 0 0 0 0 0 Strongly agree

…intentions

(47)

…consciousness

…the ability to experience emotions

…a personality

3. Imperfection manipulation check

How do you feel about the scratch on the bottle?

Annoyed 0 0 0 0 0 0 0 Not annoyed

Not happy 0 0 0 0 0 0 0 Happy

4. Partner/Servant manipulation check This water bottle works for me to keep hydrated This water bottle works with me to keep hydrated This water bottle is as a friend for me

This water bottle is as a service provider for me

5. Demographics and other • What is your age? • What is your gender?

• By which channel did you end up in this research? • Are you a student?

(48)

APPENDIX III: Descriptive statistics

1. Number of respondents that finished the questionnaire

Finished

Frequency Percent Valid Percent Cumulative Percent

Valid 0 ₆₂ _19.4 _19.4 _19.4

1 258 80.6 80.6 100.0

Total ₃₂₀ _100.0 _100.0

2. Gender

What is your gender? - Selected Choice

Frequency Percent Valid Percent

Cumulative Percent Valid Female 187 58.4 63.2 63.2 Male 107 33.4 36.1 99.3 Other, namely: 2 .6 .7 100.0 Total ₂₉₆ _92.5 _100.0 Missing System ₂₄ _7.5 Total ₃₂₀ _100.0 3. Age Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

What is your age? 296 12 83 35.10 17.790

Valid N (listwise) ₂₉₆

Group the age distribution in groups in new variable:

(49)

What is your age?

Cumulative Percent Valid -18 2 .6 .7 .7 18-24 133 41.6 44.9 45.6 25-34 63 19.7 21.3 66.9 35-44 17 5.3 5.7 72.6 45-54 28 8.8 9.5 82.1 55-64 18 5.6 6.1 88.2 65+ 35 10.9 11.8 100.0 Total ₂₉₆ _92.5 _100.0 Missing System ₂₄ _7.5 Total ₃₂₀ _100.0 4. Channels

By which channel did you end up in this research? - Selected Choice

Cumulative Percent

Valid Personal e-mail 65 20.3 22.0 22.0

Social Media (e.g.

Facebook) 166 51.9 56.1 78.0 Other, namely: 65 20.3 22.0 100.0 Total ₂₉₆ _92.5 _100.0 Missing System ₂₄ _7.5 Total ₃₂₀ _100.0 5. Familiarity

Do you know the experimenter of this study?

(50)

APPENDIX IV: Analysis

1. T-test for anthropomorphism manipulation check

Group Statistics

anthro check of

non-anthro against non-anthro N Mean

Std. Deviation Std. Error Mean Anthropormophism manipulation check - mean Non-anthro 81 1,6790 1,01782 ,11309 anthro 224 1,9129 1,15728 ,07732

Independent Samples Test

Levene's Test for Equality of

Variances t-test for Equality of Means

(51)

2. Descriptive statistics for mean of anthropomorphism manipulation check

Descriptive Statistics

Dependent Variable: Anthropormophism manipulation check - mean Perfect and imperfect

dummy

Nonanthro, anthro, partner,

servant dummy Mean Std. Deviation N

perfect Nonanthro 1,7917 1,09505 44 Anthro 2,1111 1,21415 39 Partner 1,8177 ,82018 32 Servant 2,1759 1,48160 36 Total 1,9713 1,18240 151 imperfect Nonanthro 1,5450 ,91426 37 Anthro 1,6306 ,95350 37 Partner 1,8984 1,11048 41 Servant 1,8333 1,20731 39 Total 1,7327 1,05640 154 Total Nonanthro 1,6790 1,01782 81 Anthro 1,8772 1,11434 76 Partner 1,8630 ,98810 73 Servant 1,9978 1,34775 75 Total 1,8508 1,12507 305

3. Cronbach alpha for the dependent variable

(52)

4. Mean of the dependent variable per condition

Mean of important DV questions

N Mean Std. Deviation Std. Error 95% Confidence Interval for Mean

Minimu m Maxi mum Lower Bound Upper Bound 1 46 4,5957 ,90085 ,13282 4,3281 4,8632 2,60 6,60 2 38 3,6947 1,05599 ,17130 3,3476 4,0418 1,00 5,20 3 42 4,3714 ,90588 ,13978 4,0891 4,6537 2,60 6,20 4 37 3,6378 ,98188 ,16142 3,3105 3,9652 1,40 5,60 5 35 4,0229 ,86197 ,14570 3,7268 4,3190 2,20 5,60 6 41 4,1317 ,96188 ,15022 3,8281 4,4353 1,40 5,80 7 37 4,6324 ,85572 ,14068 4,3471 4,9177 2,00 6,40 8 43 4,1442 ,96641 ,14738 3,8468 4,4416 2,40 6,40 Tot 319 4,1687 ,99028 ,05544 4,0596 4,2777 1,00 6,60 5. Hypothesis 1a

ANOVA – 2 (perfect vs. imperfect) x 2 (non-anthropomorphism vs. anthropomorphism)

Dependent Variable: Mean of important DV questions

Source

Type III Sum

of Squares df Mean Square F Sig. Corrected Model 28,521a ₃ _9,507 _10,344 _,000 Intercept 2686,569 1 2686,569 2923,165 ,000 Perf_Imperf 27,016 1 27,016 29,395 ,000 Nonanthro_Anthro ,799 1 ,799 ,870 ,352 Perf_Imperf * Nonanthro_Anthro ,283 1 ,283 ,308 ,580 Error 146,131 159 ,919 Total 2928,640 163 Corrected Total 174,652 162

(53)

6. Hypothesis 1b

ANOVA – 2 (perfect vs. imperfect) x 2 (non-anthropomorphism vs. anthropomorphism) with general use as covariate

Source

Type III Sum

of Squares df Mean Square F Sig. Corrected Model 28,687a 4 7,172 7,763 ,000 Intercept 312,681 1 312,681 338,462 ,000 General_use ,166 1 ,166 ,180 ,672 Perf_Imperf 26,867 1 26,867 29,082 ,000 Nonanthro_Anthro ,801 1 ,801 ,867 ,353 Perf_Imperf * Nonanthro_Anthro ,301 1 ,301 ,325 ,569 Error 145,965 158 ,924 Total 2928,640 163 Corrected Total 174,652 162

(54)

7. Hypothesis 2a

ANOVA – 2 (perfect vs. imperfect) x 4 (non-anthropomorphism vs anthropomorphism vs. partner vs. servant)

Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 37,858a ₇ _5,408 _6,139 _,000 Intercept 5463,098 1 5463,098 6201,077 ,000 Perf_Imperf 20,065 1 20,065 22,775 ,000 Nonanthro_Anthro_P ar_Ser 6,573 3 2,191 2,487 ,061 Perf_Imperf * Nonanthro_Anthro_P ar_Ser 11,402 3 3,801 4,314 ,005 Error 273,988 311 ,881 Total 5855,320 319 Corrected Total 311,847 318

(55)

8. Hypothesis 2b

ANOVA – 2 (perfect vs. imperfect) x 2 (partner vs. servant)

Tests of Between-Subjects Effects

Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 8,207a 3 2,736 3,252 ,023 Intercept 2776,564 1 2776,564 3300,841 ,000 Perf_Imperf 1,394 1 1,394 1,657 ,200 Part_Serv 3,748 1 3,748 4,456 ,036 Perf_Imperf * Part_Serv 3,453 1 3,453 4,105 ,044 Error 127,858 152 ,841 Total 2926,680 156 Corrected Total 136,064 155

(56)

9. Hypothesis 2c

ANOVA – 2 (perfect vs. imperfect) x 4 (non-anthropomorphism vs anthropomorphism vs. partner vs. servant) with general use as covariate

Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 38,272a 8 4,784 5,421 ,000 Intercept 635,916 1 635,916 720,586 ,000 General_use ,414 1 ,414 ,469 ,494 Perf_Imperf 19,990 1 19,990 22,652 ,000 Nonanthro_Anthro_P ar_Ser 6,649 3 2,216 2,511 ,059 Perf_Imperf * Nonanthro_Anthro_P ar_Ser 11,490 3 3,830 4,340 ,005 Error 273,575 310 ,882 Total 5855,320 319 Corrected Total 311,847 318

(57)

10. Hypothesis 2d

ANOVA – 2 (perfect vs. imperfect) x 2 (partner vs. servant) with general use as covariate

Tests of Between-Subjects Effects

Source

Type III Sum

of Squares df Mean Square F Sig. Corrected Model 8,458a 4 2,115 2,502 ,045 Intercept 323,242 1 323,242 382,502 ,000 General_use ,252 1 ,252 ,298 ,586 Perf_Imperf 1,399 1 1,399 1,656 ,200 Part_Serv 3,721 1 3,721 4,403 ,038 Perf_Imperf * Part_Serv 3,579 1 3,579 4,235 ,041 Error 127,606 151 ,845 Total 2926,680 156 Corrected Total 136,064 155

(58)

11. T-test between imperfect partner and imperfect servant Group Statistics

Comparison partner and

servant imperfection N Mean Std. Deviation

Std. Error Mean Mean of important DV questions imperfect partner 41 4,1317 ,96188 ,15022 imperfect servant 43 4,1442 ,96641 ,14738

F Sig. t df Sig. (2-tailed) Mean Differe nce Std. Error Differe nce 95% Confidence Interval of the Difference Lower Upper Mean of important DV questions Equal variances assumed ,324 ,571 -,059 82 ,953 -,01248 ,21047 -,43116 ,40621 Equal variances not assumed -,059 81,8 45 ,953 -,01248 ,21044 -,43113 ,40617 12. Additional testing 1

ANOVA - 2 (perfect vs. imperfect) x 3 (non-anthropomorphism vs partner vs. servant)

Tests of Between-Subjects Effects

Source

Type III Sum

(59)

13. Additional testing 2

Independent samples t-test between the normal condition and the imperfect relationship norm

Group Statistics

Normal versus the

partner and servant N Mean

Std. Deviation Std. Error Mean Mean of important DV questions normal condition 46 4,5957 ,90085 ,13282

imperfect partner and servant

84 4,1381 ,95840 ,10457

(60)

A between-subject ANOVA and post-hoc test for the perfect conditions

Between-Subjects Factors N Condition 1 46 3 42 5 35 7 37

Tests of Between-Subjects Effects

Dependent Variable: Mean of important DV questions Source

Type III Sum of

Squares df Mean Square F Sig.

Corrected Model 8,708a 3 2,903 3,718 ,013 Intercept 3070,260 1 3070,260 3932,752 ,000 Condition 8,708 3 2,903 3,718 ,013 Error 121,788 156 ,781 Total 3256,320 160 Corrected Total 130,496 159

a. R Squared = ,067 (Adjusted R Squared = ,049)

Multiple Comparisons

Dependent Variable: Mean of important DV questions Tukey HSD

(I) Condition (J) Condition

Mean Difference

(I-J) Std. Error Sig.

95% Confidence Interval

Lower Bound Upper Bound

(61)

A between-subject ANOVA and post-hoc for imperfect conditions Between-Subjects Factors N Condition 2 38 4 37 6 41 8 43

Tests of Between-Subjects Effects

Source

Type III Sum of

Squares df Mean Square F Sig.

Corrected Model 8,870a 3 2,957 3,011 ,032 Intercept 2412,387 1 2412,387 2456,755 ,000 Condition 8,870 3 2,957 3,011 ,032 Error 152,201 155 ,982 Total 2599,000 159 Corrected Total 161,071 158

a. R Squared = ,055 (Adjusted R Squared = ,037)

Multiple Comparisons

Dependent Variable: Mean of important DV questions LSD

(I) Nonanthro, anthro, partner, servant dummy (J) Nonanthro, anthro, partner, servant dummy Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound Nonanthro Anthro ,05690 ,22887 ,804 -,3952 ,5090 Partner -,43697 ,22314 ,052 -,8778 ,0038 Servant -,44945* ,22063 ,043 -,8853 -,0136 Anthro Nonanthro -,05690 ,22887 ,804 -,5090 ,3952 Partner -,49387* _,22470 _,029 _-,9377 _-,0500 Servant -,50635* ,22220 ,024 -,9453 -,0674 Partner Nonanthro ,43697 ,22314 ,052 -,0038 ,8778 Anthro ,49387* ,22470 ,029 ,0500 ,9377 Servant -,01248 ,21630 ,954 -,4398 ,4148 Servant Nonanthro ,44945* ,22063 ,043 ,0136 ,8853 Anthro ,50635* ,22220 ,024 ,0674 ,9453 Partner ,01248 ,21630 ,954 -,4148 ,4398

(62)

T-test to compare condition the relationship norm with the other conditions within the imperfection condition

Group Statistics

Imperfect partner and servant with (non)anthro N Mean Std. Deviation Std. Error Mean Mean of important DV questions (non)/ anthro 75 3,6667 1,01360 ,11704 Partner/servant 84 4,1381 ,95840 ,10457