Restaurant tipping:
the impact of the acknowledgment
of gratitude on customer’s future
intentions regarding revisit and
word of mouth.
Marion Holtkamp
Student number: 10475443
January 31
st, 2015
MSc Business Studies, Marketing track
Supervisor: Dr. K. Venetis
2
Acknowledgements
This Master thesis is the final assignment for my Master Business Studies at the University of Amsterdam. It explores the impact of the acknowledgment of gratitude on customer’s future intentions in the context of restaurant tipping.
Completing my study and this thesis would not have been possible without the help of a number of people. True to the subject of this thesis, I would like to express my feelings of gratitude for their help.
I would first like to thank my thesis supervisor, Dr. Karin Venetis. Her guidance and critical view gave me confidence and additional insight to write this thesis.
Furthermore I like to thank my study-buddy Tabitha Oostra-Brusse for her feedback, support and talent to put things into perspective. Also, I like to say thanks to Nick Melis, my family and Carien Verweij for their support and understanding.
Last but not least, I like to thank all my respondents for participating in this study!
Marion Holtkamp
3
Abstract
Restaurant tipping: the impact of the acknowledgment of gratitude on customer’s future
intentions regarding revisit and word of mouth.
Keywords: tipping; relationship marketing; consumer behavior; (expression and acknowledgment of) gratitude; reciprocity; social norm and compliance
Tipping is a fascinating, dynamic and widely-adopted phenomenon, not only is it strongly culturally determined, it also is an illustration of irrational consumer behavior as consumer voluntarily increase the price of their restaurant visit. The majority of research into tipping has been limited to the motivations and antecedents which drive tipping behavior; it neglects the effect the practice has on the buyer-seller relationship.
This study uses the concept of gratitude in relationship marketing to explore the impact of acknowledgement by the service provider, as a response to expressed gratitude (by means of a tip), on future intentions by the customer, being word of mouth and revisit intentions.
As survey among 304 people in the Netherlands confirms the significant role gratitude plays in relationship marketing. Our results suggest that respondents who expect and/or perceive to be acknowledged for the expression of gratitude consider acknowledgment to be a hygiene factor. This quantitative study contributes to earlier qualitative research by Raggio et al. (2014) and confirms that a completed gratitude cycle enchances future intentions; as failing to do so has a negative effect on revisit and word of mouth intnetions.We also identify that when consumers make a voluntary financial contribution to express their gratitude, no threshold for overvaluing this contribution exist.
Practically, this research provides hospitality service providers valuable insight in consumer behavior related to tipping in the Netherlands. It stresses the importance of relationship marketing investment, e.g. extra service and adaptations in policies, to make customers feel grateful. As a result service providers benefit from an increased tip. To strengthen the relationship-quality this study suggests that service providers should explicitly thank customers for their tip. Consequently, people are more likely to speak positively about the restaurant and revisit. The latter is of extra importance for service staff as this study shows that people who visit the restaurant more frequently tip significantly higher.
4
Table of Content
1. Introduction ... 6
2. Literature review... 9
2.1 Tipping ... 9
2.1.1 Tipping and social norm ... 10
2.1.2 Tipping and gratitude ... 11
2.1.3 Tipping – across countries ... 12
2.1.4 Effects of tipping ... 13
2.2 Relationship marketing ... 14
2.2.1 Relationship benefits ... 15
2.2.2 Relationship quality ... 16
2.2.3 Relationship marketing and gratitude ... 16
2.3 Conceptual model ... 21
3. Research method ... 22
3.1 Data collection procedure ... 22
3.2 Sample ... 22
3.3 Measure development ... 23
4. Results ... 24
4.1 Descriptive data of sample ... 24
4.2 Variables and measurements ... 26
4.2.1 Reliability of scales ... 26
4.2.2 Skewness and kurtosis ... 26
4.2.3 Correlation Matrix ... 27
5
4.2.5 Tip percentage vs. perceived height of tip ... 32
4.3 Hypothesis testing ... 33
5. Discussion and conclusions... 44
5.1 Summery of the results ... 44
5.2 Discussion and conclusions of the results ... 47
5.3 Theoretical implications ... 49
5.4 Managerial implications ... 50
6. Limitations and further research ... 50
Bibliography ... 53
Appendixes ... 57
I Hardcopy invitation to participate in survey ... 57
II Relevant Marketing Scales: ... 57
III Questionnaire ... 58
IV Multiple regression analyses (hypotheses 1, 2, and 3) ... 63
V Regression analyses (hypothesis 4) ... 67
VI Output moderation results (hypothesis 5) ... 73
VII Moderator values defining Johnson-Neyman significance region (hypothesis 5)... 75
6
1. Introduction
Tipping is a fascinating, dynamic and widely-adopted phenomenon, not only is it strongly culturally determined it also challenges the standard assumption of selfishness in the economic exchange theory (Lynn, Zinkhan, & Harri, 1993). A tip, is paid after the service has been received; the customer therefore intentionally increases the price he or she has to pay. This irrational behavior stresses the importance of psychological reasons and social norms in motivating economic behavior (Conlin, Lynn, & O'Donoghue, 2003; Azar, 2007) and presents interesting insight into consumer behavior and the buyer-seller relationship.
Since 1977, in which Ben-Zion and Karni offered the first economic model for tipping, the majority of research into tipping has been limited to the motivations and antecedents which drive tipping behavior; it neglects the effect the practice has on the buyer-seller relationship. The little interest of the effects on the buyer-seller relationship is remarkable as relationship marketing is a much addressed topic in the service industry as regular/loyal customers contribute significantly to high profit margin and revenue, positive referral and continuing customer reciprocity (Morgan & Hunt, 1994; Ryu & Lee, 2013; Gummesson, 2014).
To understand the buyer-seller relationship better, insight into the motivation of the buyer to tip and how certain antecedents can manipulate this behavior are essential. In an overview of literature Azar (2007a) sums up service quality, patronage frequency (loyalty), bill size, service quantity, group size/ composition, differences between countries, interpersonal connection with the server and social norms as variables to tip. Later in 2011, Azar suggests that the two main underlying reasons for tipping in restaurants are that tipping is a social norm and that customers can use tipping to show their gratitude for the service they received.
The fast majority of articles on tipping however takes the social norm argument as a departure point, and neglects gratitude (Azar, 2007a; Conlin et al. 2003; Kershnar, 2014). In general, gratitude is a concept which has been overlooked by marketing scholars in the study of commercial relationships (Raggio, Walz, Godbole, & Garretson Folse, 2014; Palmatier et al., 2009).The social norm argument explains tipping behavior by means of an implicit contract between the server and the customer (Bodvarsson & Gibson, 1994) and it is seen as form of reciprocity which is aligned with a score-keeping mentality (Cropanzano & Mitchell, 2005).
7
Palmatier, Burke Jarvis, Bechkoff and Kardes (2009) however claim that emotional feelings in response to a favor or benevolence received are different from the norm of reciprocity. Recent research into relationship marketing has reinforced the importance of gratitude and stresses that taking reciprocity (as a score-keeping mentality) as a departure point in research ignores the underlying and important role of gratitude (Palmatier et al., 2009). Raggio et al. (2014) suggest that gratitude is an important antecedent for trust and commitment, which are proven precedents for loyalty and should therefore receive more attention.According to Palmatier et al. (2009) gratitude is a key mediator in understanding how relationship marketing investments increase e.g. (re)purchase intentions. In their qualitative research of 2014 Raggio et al. build on the study of Palmatier et al. (2009) and claim that for both parties to reap the full transformational effect of the expression of gratitude both parties have to come to a mutal understanding that the gratitude cycle has been accomplished. The steps of this cycle are ‘benefit – gratitude – expression – acknowledgement’ and is the initial component of their theoretical model. In this cycle the beneficiary first has to recognize a benefit which he then feels appreciation for (gratitude), when the beneficiary expresses his feelings of gratitude a non-financial exchange takes place which the intitial benefactor has to acknowledge to complete the cycle.
One of Raggio et al. (2014) observations therefore is that buyer and seller are both responsible to express gratitude, but an asymmetry exists in which sellers are expected to show more gratitude. The opposite seems to be the case in the concept of tipping, in which the buyer voluntarely increases the price of the service to express gratitude. This observation will be especially relevant in a context in which the social norm for tipping is less prevelant and guidelines less explicit. In a cross-country study Lynn et al. (1993) suggest that this is the case in countries with feminine values, in this context tipping emphasizes the social relationship. Most insight into tipping behavior however applies to a North American context in which masculine values prevail, here tipping stresses an economical relationship. Little academic research is available related to tipping in countries where tipping is less prevalent and it’s rules of the practice remain vague.
This study will be the first to explore the impact of acknowledgement by the service provider, as a response to a tip, on future intentions by the customer, being word of mouth and revisit intentions. It will test if the claim of Raggio et al. (2014) that the transformational power of
8
gratitude can only be achieved when both parties agree that the gratitude cycle has been completed is correct. Additionally, this study will show if the evaluation of the ratio between the level of acknowledgement and height of the tip by the customer has to be in accordance to have an effect. In general, it will be valuable to explore if the expressed acknowledgment by the service provider strengthens the relationship between customer and service provider and consequently increases future intentions by the customers, and eventually business results.The main research question has been defined as follows: what is the impact of acknowledgment by the hospitality service supplier, in reaction to the expression of gratitude shown by the customer in terms of tipping, on customer future intentions to revisit and of word of mouth?
In order to answer this question the literature review will answer the following sub-questions:
What is tipping and what motivate this behavior?
How do ‘social norm theory’ and feelings of gratitude’ affect tipping behavior?
What are cross-cultural differences in tipping behavior in relation to ‘social norm’ and ‘feelings of gratitude’?
What is relationship marketing?
What is the impact of gratitude in relationship marketing on buyer-seller relationships? What are factors which make buyers (and sellers) grateful?
How do feelings and acknowledgement of gratitude impact customer reciprocity (revisit intentions and word of mouth)?
The paper is organized in the following manner: firstly a review of literature provides insight in the phenomenon of tipping related to social norm compliance and gratitude and the concept of gratitude in relationship marketing. This overview will provide relevant insight in tipping behavior and the importance of gratitude to the buyer-seller relationship which will likely result in positive customer future intentions. Afterwards details about the quantitative research are provided. Findings from the studies, as well as support from the literature are incorporated into a theoretical model. Finally the consequences of the research are discussed and the academic and managerial implications are addressed.
This research will make an academic and managerial contribution. Firstly, it will provide insight into the impact of expressed gratitude which provides academic value regarding the role of gratitude in (B2C) relationship marketing. Additionally it will provide more insight into the consumer behavior related to tipping. Secondly, from a managerial perspective it will provide
9
guidance in anticipating on tipping behavior of customers by the service provider and consequently improving the relationship quality, hence customer’s future intentions.2. Literature review
2.1 Tipping
Tipping is a widespread custom governed by many and varied norms and motivations which often leave customers uncertain about how much to tip, or even whether they should tip at all (Lynn, Zinkhan & Harris, 1993). Tipping is a multi-billion-dollar phenomenon worldwide. To illustrate: the custom accounts for up to $40 billion in the United States (Holland, 2009), and approximately €425 million in the Netherlands (Versprille, 2014) a year.
The origin of the concept of tipping knows many stories dating as far back as the Middle Ages when lords would throw coins at beggars to ensure a safe journey (Lynn et. al. 1993; Azar, 2007; Whaley et al., 2014). The origin of the word ‘tip’ is also debated. Lynn et al. (1993, p. 479) sums up the following derivations: “(1) from the Latin word stips, meaning ‘gift’, (2) from the Dutch word ‘tippen’, meaning ‘to tap’ (as in tapping a coin on a table to attract a server’s attention), (3) the gypsy phrase ‘tipper me your money’, and (4) from the words ‘To Insure Promptitude’ that were placed on collection boxes in eighteenth-century English coffee shops.” In countries where tipping is custom an implicit contract dictates that good service is rewarded by tipping. In general a custom disappears over time when it is costly to follow, however the norm to tip has defied this observation. Azar (2007b) suggest that the custom survives due to the fact that it is the most efficient way to monitor and reward the efforts of service workers. Other research by Lynn (2001) however shows a weak correlation between the tip amount and the quality of service. Better service increased the tip only by approximately 1.5%; if tipping was aimed to reward or punish the quality of service Lynn claims that this increase would be too small for waiter to notice. Conlin et al. (2003) present an alternative reason, they suggest that the custom, although costly, does give a return as it takes away the negative effects of disutility of not adhering to the social norm.
10
Since the late 70’s experts have theorized economic models of tipping. In the early days the observation of Ben-Zion and Karni (1977) was that tips were consistent with self-interest seeking behavior only for the case of repetitive customers to ensure the same level or better service for their next visit. The motivations for tipping by one-time customers were left out of their model. Ever since, tipping literature knows two dominant schools of thought (Whaley et al., 2014), described as the employer-driven and customer-led. Segrave (1998) claims that from the first perspective tipping behavior is driven by consumers who are sympathetic toward low-wage servers and try to compensate for the irresponsible employers who cut on labor costs. The customer-led view presents the idea that consumers like to reward (or not) the quality of service received and show gratitude for the service received. In a review of tipping literature for economic and psychological reasons to tip these two schools are recognized. Returning motivations are: rewarding and responding to feelings of gratitude for good service while simulatiously assuring future service delivery (Azar, 2007; Bodvarsson, Luksetich, & McDermott, 2003; Lynn, 2001, Whaley et al., 2014), out of empathy for the service staf (Azar, 2007; Lynn, 2001, Whaley et al., 2014) and need to comply to social norms and in relation the need for social approval (Azar, 2007; Bodvarsson et al., 2003; Conlin et al., 2003; Lynn, 2001, Whaley et al., 2014).In a study by Azar in 2010 people were asked to indicate one or more out of seven potential motivation for tipping. With 85% “tipping in restaurants in the social norm in the US” scored highest, follow by “by tipping I can show the waiter my gratitude for his service” (68%).
2.1.1 Tipping and social norm
Kershnar (2014, p.247) suggests that “in contexts in which tipping is customary, there is a moral duty to tip or explicitly tell the server that you will not be tipping”. He draws this conclusion after researching both customers and server’s views on a scenario in which a customer is unable to tip his server. The strong and apparent views from both parties illustrate a duty to tip which is inherent in ordering food. This duty is dictated by an implicit contract which is governed by a convention in the United States that customers tip their servers 15-20%. Conlin et al. (2003) suggests that the main motivation of tipping is an internalized norm to support the social (external) norm. When people violate this norm they are left with negative feelings caused by less favorable social treatment from their peers. In their decision how much to tip, the customer
11
trades off material payoffs against the harmful effects. Additionally Azar (2007a) explains that compliance with the social norm of tipping by stating that people derive benefits from tipping, such as impressing others and improving their self-image. In another paper, also published in 2007, he states these reasons as the major motivations for tipping and labels these extrinsic and intrinsic motivations as ‘social pressure’ and ‘fairness’.Although the above research is all conducted in an American context which knows a strong social tipping norm, the social norm theory most probably also apply to some extend in the Dutch context, in which this research is conducted. Although the convention might be less explicit and social norm less prevalent, the custom of tipping is commonly used. 9 out 10 Dutch people indicate to regularly tip after a visit to a restaurant or bar (TNS Nipo/ RTL Nieuws, 2005). This vast majority suggests that a sense of duty to tip is likely to exist, also in the Netherlands. Consequently, the following is proposed:
H1 Feelings to comply to social norms related to tipping positively influence tipping behavior.
2.1.2 Tipping and gratitude
Although ‘gratitude’ is suggested to be one of the most prevalent motivations to tip, this antecedent has received little isolated attention in tipping literature (Azar, 2007), neither has it in marketing in general (Palmatier et al., 2009). A widely explored topic in tipping literature has been the motivation driven by psychological utility to reciprocate feelings of satisfaction caused by good service (Lynn & McCal , 2000). Little prior work exists about the relationship between gratitude and satisfaction. In their theoretical framework Raggio et al. (2014) suggest that gratitude precedes satisfaction, however their study offers mixed results, also suggesting that satisfaction first has to be achieved before experiencing gratitude. They claim that the constructs gratitude and satisfaction are independent and suggest that feelings of gratitude go beyond feelings of satisfaction; feelings of gratitude are a response to an experience of benevolence rather than offsetting the actual experience against the expected experience.
Regardless of the order, a strong relationship between the constructs satisfaction and gratitude exist. Azar (2010) suggests that people who tip in reaction to feelings of gratitude are more sensitive for the quality of service than those who tip to comply to social norm. In contrast,
12
people who indicate that social norm theory is their main reason for tipping do not think that it is necessary to tip much more for excellent service than for poor service. Research by TNS Nipo/ RTL Nieuws (2005) shows that the vast majority (84%) of tippers in the Netherlands gives a tip to show appreciation for the received service. If this appreciation is caused by feelings of satisfaction or gratitude cannot be derived, however in light of this study we propose the following hypothesis:H2 Feelings of gratitude of the consumer towards the service experience positively influence
tipping behavior.
2.1.3 Tipping – across countries
Although wide-spread, the concept of tipping is strongly culturally determined. The variation in tipping is considerable. In countries like the United States and Canada the principle is relatively straight forward and the tip percentage high; the guideline for the percentage to tip in restaurants is 15-20%. There are also countries, like Brazil, Turkey and Japan, where tipping is not part of its culture and can even be taken as an insult. Other countries adopt a custom of a moderate tip of 10% or less, is these countries (e.g. Germany, France, Argentina and China) the practice remains vague, one person gives 10 cents, the other 10 percent (Bujisic et al., 2014; Conlin et. al. 2003; Kershnar, 2014; Lynn et al., 1993). Conlin et. al. (2003) suggest two possible answers for cross-country differences. Firstly, the extent to which the tipping norm is related to possible benefical outcomes, these “efficiency benefits” differ across countries. Secondly, the degree the norm for tipping is enforced by “internalized feelings of guilt and shame”, the sensitivity for, and the effects of these feelings are experienced differently across cultures.
Lynn et al. (1993) is the only researcher who has done extensive cross-country research to test national differences in tipping practices and norms at a macrolevel across 30 countries. The predictor variables used in his research are the cultural dimensions of Hofstede, power distance, uncertainty avoidance, individualism and masculinity. He suggests that “tipping appears to be more prevalent in countries with a high tolerance of status and power differences” (p.483), “tipping was more prevalent the less tolerant countries were of uncertainty” (p.483), suggesting tipping gives customers a sense of control. Another conclusion was that “tipping was
13
more prevalent in communalistic countries (p.484). Finally, “tipping appears to be less prevalent in countries with feminine values” (p.484).The Dutch score for masculinity is 14, in contrast to the US score of 62 (Hofstede, 2014), this suggests that in general the Dutch national culture has a strong preference for feminine values. In countries with feminine values service, interdependence and people are important; tipping in this context emphasizes the social relationship. Most insight into tipping behavior however applies to a North American context in which masculine values prevail, here tipping stresses an economical relationship (Lynn et al.,1993).
Similar to Azar’s findings in his North American study in 2010, it is likely that social norm theory and feelings of gratitude will positively relate to the tip amount in countries like the Netherlands, like proposed in the hypotheses one and two. However, the final observation of the strong preference for feminine values in relation to tipping behavior suggests that gratitude will have a stronger effect on the tip amount than compliance to social norm in the Dutch context in which this research is conducted; therefore hypothesis three is as follows:
H3 ‘Feelings of gratitude’ will have a stronger effect on tipping than ‘compliance to social norm’ in a country with a preference for feminine values, like the Netherlands.
2.1.4 Effects of tipping
Tipping literature in general explores consumer’s motivations for tipping which explain the variance and irrationality in tipping behavior. Related to these motivations the psychological effects of tipping are addressed. However no literature on the effect of tipping on the exchange relation between consumer and provider and reciprocal consumer behavior has been found. Gratitude, strongly related with satisfaction, however are both assumed to positively affect tipping behavior. As a consequence the well supported satisfaction-trust-commitment theory of positive reciprocal consumer behavior (Garbarino & Johnson, 1999; Hennig-Thurau, Gwinner, & Gremler, 2002; Raggio et al., 2014) would apply. The tip is assumed to be a behavioral repsonse to the affective aspect of gratitude and will therefore be positively related with customer’s future intentions towards the service provider.
14
Overall, restaurant visits are characterized by variety seeking behavior, which despite having great potential for high commitment are also typified by low repeat behavior (Dick & Basu, 1994). This type of customers is described as ‘low relational’ by Garbarino and Johnson (1999) and for this type of customers overall satisfaction (and gratitude) is the primary mediating construct between attitudes and future intentions. Therefore we propose that if the tip amount goes up as behavioral response to feelings of gratitude and satisfaction, the tip amount consequently will have a predictive value for future intentions like revisit intentions and word of mouth. This results in the following hypothesis:H4 There is a positive relation between tipping behavior and customer’s future intentions.
2.2 Relationship marketing
Relationship marketing (RM) is defined as “all marketing activities directed toward establishing, developing, and maintaining successful relational exchanges” (Morgan & Hunt, 1994, p. 22). The objective of relationship marketing theory is understanding which key drivers have an impact on important company results and obtaining better insight in the causal realtions between these drivers and results (Hennig-Thurau et al., 2002). The increased emphasis on RM is a reaction to the assumption that strong customer relationship result in “guest satisfaction, positive word of mouth, business reeferrals, references and publicity” (Kim & Cha, 2002. p.321).
Gwinner, Gremler and Bitner (1998) claim that strong customer relationships are especially important in the service industry because of it’s interpersonal focus and the relative lack of objective measures for evaluating service quality. Research in both business-to-consumer and business-to-business markets has established a well supported commitment-trust theory in RM to achieve these strong customer relationships. Hence, RM investments are required to enhance both commitment and trust, which in turn positively mediates customer behavior, resulting in superior seller performance (Morgan & Hunt, 1994; Sirdeshmukh, Singh, & Sabol, 2002). A meta-analysis by Palmatier, Dant and Evans (2006), aimed to test the commitment-trust theory in RM, however revealed that the direct effect of RM investments had a greater effect on performance outcomes than could be explained by the mediating effect of commitment and trust alone, suggesting a missing mediating mechanism. Three years later Palmatier et al., (2009) find support to believe that this missing construct is ‘gratitude’.
15
RM investments are essential to reap the benefits of RM. According to Palmatier et al., (2009) these investments can vary from the delivery of extra service, adaptations in policies and providing small favors or considerations. These initiatives create short term feelings of gratitude (and in long term commitment and trust) with customers, which result in gratitude-based reciprocal behaviors, which in turn lead to enhanced seller performance. Although there are hospitality suppliers which have commited and trustworthy relationships with their customers, most relationship between customers and suppliers will be clasified as ‘low relational’ (Garbarino & Johnson, 1999) due to the variety seeking attitude customers have towards the service. This does not make relationship marketing investments less important, but it is assumed to make the mediating short term effect of gratitude, versus commitment and trust, more relevant.In the past three decades two conceptual approaches have tried to explain the determinants of succesfull relationship marketing outcomes, the relational benefits approach and the relationship quality approach.
2.2.1 Relationship benefits
Research and experience have proven that strong consumer-supplier relations are beneficial for the service provider, however Gwinner et al. (1998) claim that for a relationship to be long-term, both parties will have to benefit. In this view they conducted research into the motivations of consumers to be loyal to a company. These motivations were defined as relational benefits; “benefits consumers receive above and beyond the core service performance, in return for their enduring relationship with a service provider” (p. 102).
These relational benefits comprise of three dimensions, namely: social, confidence, and special treatment benefits. Social benefits are psycho-social benefits and include emotions of personal recognition, friendship and familiarity and result from personal connections. Confidence benefits reduce perceived risk and builds trust. Special treatment benefits are economic incentive-based preferential actions such as discounts, free upgrades and complimentary services (Ryu & Lee, 2013).
Gwinner et al. (1998) conclude that confidence benefits are most important in services, followed by social and special treatment benefits. The confident benefit ensures that you know what you can expect of your service provider and consequently reduces the perceived risk caused by the
16
intangible character of services. In relation to this study, confidence benefits of tipping could be the predictability and insurance and of good service in the future.2.2.2 Relationship quality
Relationship quality reflects the overall nature of the relationship between companies and consumers. As relationship marketing has gained ground, so did the concept of relationship quality. Over time the concept has known various component or dimensions including cooperative norms (Baker, Simpson and Kelley, 1998) and customer orientation (Dorsch, Swanson and Kelley, 1998; Palmer and Bejou, 1994). However, most researchers conceptualize relationship quality as a “high-order construct composed of commitment, satisfaction and trust” (De Wulf, Odekerken-Schroder, & Iacobucci, 2001; Hennig-Thurau et al. 2002; Ulaga & Eggert, 2006 p.314).
Relationship quality refers to relationship strength; Ryu and Lee (2013) claim it provides insight into customer perceptions of the extent to which their expectations and desires are satisfied by their relationship with services providers. A high relationship quality suggests that customers trust and rely on the performance of service providers. In line, Kim and Cha (2002) refer to relationship quality as the perceptions and evaluations of individual employee’s communication and behavior, such as respect, courtesy, warmth, empathy and helpfulness.
In conclusion relationship quality focuses on the nature of the relationship and relational benefits focus on the receipt of user-oriented benefits, both concepts view the fulfillment of customer needs as central for relationship success (Hennig-Thurau et al., 2002). This study uses the approach of relationship quality as it explores the role of gratitude in customer-supplier interactions; relationship quality has proven to be a strong predictor of performance outcomes like brand loyalty, (re)purchase intentions, share of wallet and positive word of mouth (Garbarino & Johnson, 1999; Palmatier et al. 2009).
2.2.3 Relationship marketing and gratitude
Gratitude has always been an integral component in the study of social relationships but the component has been overlooked by marketing scholars in the study of commercial relationships (Raggio, Walz, Godbole, & Garretson Folse, 2014; Palmatier et al., 2009). This is remarkable as
17
gratitude shows potential to “transform transactional exchanges into relational ones” (Bolton, Lemon, & Verhoef, 2004, p.272). The study of Palmatier et al., (2009) into RM reinforces the importance of gratitude and stresses that taking reciprocity as a departure point in research ignores the underlying and important role of gratitude.To feel gratitude customers have to voluntarily appreciate and acknowledge certain given benefits. The RM investment (the benefit) has to be perceived as genuine by the customer and demonstrate the service provider is acting in the best interest of the customer. If this is the case the benefit will be perceived as a benevolent act (Atuahene-Gima & Li, 2002; Fazal e Hasan, Lings, & Mortimer, 2014). An actual benefit is not necessarily required to experience gratitude. Only the intention or perceived intent to benefit the customer can already result in feelings of gratitude and therefore good intentions can already be sufficient.
The ingrained psychological pressure to reciprocate after receiving a RM investment has been recognized by researchers in different fields (Palmatier et al., 2009). Raggio et al. (2014) explain reciprocity as closely aligned with a conditioned score-keeping mentality in which people seem to be hardwired to repay others for what they receive, in a positive or negative sense. A response related to feelings of gratitude however is not characterized by mindless behavior, nor does the expresser expects something in return. In case of gratitude the beneficiary genuinely appreciates and acknowledges certain given benefits. To experience feelings of gratitude “the recipient must recognize that the benevolence is intentional and moreover, attribute good intentions to the giver” (Palmatier et al., 2009. p.5). The recipient can then opt to keep these feelings of gratitude to himself or express his feelings of gratitude (gratitude-based reciprocal behavior) to the giver. In this sense gratitude is described as the emotional core of reciprocity. Komter (2004) suggests that through this type of mutual giving, people become tied to what has been described as a “web of feelings of gratitude” (p.203).
Although Palmatier et al. (2009) underpinned the importance of gratitude, they failed to demonstrate how gratitude strenghtens customer-supplier relationships. To unraffle the “web of feelings of gratitude” and shine light on the role of gratitude in commercial relationships Raggio et al. (2014) conceptualized the web by introducing a gratitude cycle. In their model recognition
18
and appreciation are key conditions for gratitude to exist; “without recognition, a beneficiary may not perceive himself as one, and without appreciation, feelings of gratitude will not exist” (p.4). The steps of this cycle are ‘benefit – gratitude – expression – acknowledgement’ (figure 2.1) and is the initial component of their theoretical model. In this cycle the beneficiary first has to recognize a benefit which he then feels appreciation for (gratitude), when the beneficiary expresses his feelings of gratitude a non-financial exchange takes place which the intitial benefactor has to acknowledge to complete the cycle.The effect of the completed gratitude-cycle is measured by the perceived quality of the relationship, a metaconstruct composed of satisfaction, trust and commitment, which reflects the overall nature of the relationship between companies and consumers (Hennig-Thurau et al., 2002)
They claim that for both parties to reap the full transformational effect of the expression of gratitude both parties have to come to a mutal understanding that the gratitude cycle has been accomplished. To do so, the person who expresses it’s gratitude has to agree the receiver has acknowledged his expressed gratitude.
Figure 2.1 Gratitude cycle, adapted from Raggio et al., 2014
Additionally, Raggio et al. (2014) observes that buyers and sellers are both responsible to express gratitude, but an asymmetry exists in which sellers are expected to show more gratitude. The opposite seems to be the case in the concept of tipping, in which the customer voluntarely increases the price of the service and so invests extra effort and money in his
2. Gratitude 3. Expression 4. Acknowledgment 1. Benefit
19
expression of gratitude. In the context of tipping, the moment the beneficiary expresses his feelings of gratitude by means of tipping (step 3) he becomes the benefactor and the receiver of the tip the beneficiary. In earlier research Buck (2004) illustrates this step in his ‘gratitude of exchange’, in which the beneficiary receives something valued, and the benefactor gives up something valued, this exchange involves the the beneficiary (the service provider) to humbly give thanks for the benefit received (acknowledgement). In this search for balance, the equity theory of Adams (1965) is also relevant to mention. It states that there has to be a fair distribution of resources within interpersonal relationships, this observation makes it plausible to assume that the acknowledgement by the service provider of the customer’s extra effort is expected by the customer, and therefore the completion of the above cycle is of importance.This study will explore the moderating effect of acknowledgement by the service provider, as a response to a tip, on future intentions by the customer, being word of mouth and revisit intentions. It will test if the claim of Raggio et al. (2014) that the transformational power of gratitude can only be achieved when both parties agree that the gratitude cycle has been completed is correct.
For this study we assume that a tip is (in some extent) a behavioral response to an emotion in which appreciation for a benefit is felt. Both the expressions (tipping) of gratitude by the consumer and acknowledgement of gratitude by the service provider can vary from non-existing to explicit. The figure below applies the gratitude cycle to a tipping context.
Figure 2.2 Gratitude cycle applied to tipping context
2. Buyer appreciation of benefit: GRATITUDE (emotion) 3. EXPRESSION of gratitude by tipping (behavior of buyer) 4. ACKNOWLEDGMENT of tip by seller 1. Buyer recognition of BENEFIT (cognitive) extra service, adaptations in policies,small favors/ considerations (RM investments)
20
The completion of these steps above should optimize the relationship between buyer and seller according to Raggio et al. (2014). Therefore the following is proposed:
H5 The acknowledgement of the tip by the service provider moderates the impact of tipping behavior on customer future intentions; when the acknowledgment is more explicit the tip will have a stronger effect on revisit and word of mouth intentions.
In line with the gratitude cycle of Raggio et al. (2014) and the equity theory of (Adams, 1965) we hypothesized that the acknowledgement of the tip has to be explicit to achieve a fair relational exchange of gratitude. However, we also assume that it has to be in line with the perceived height of the tip. This assumption is based on finding of Adams (1965) and Mayser and von Wangenheim (2012) that individuals perceive distress when they feel that they are under or over benefited. Although the threshold for making customer feel overvalued is higher than making customers undervalued, both feelings result in feelings of discomfort, stress and guilt. The reaction of the service supplier therefore can either be evaluated as fair or unfair by the customer, according to their perception of the height of the tip. In case of a fair evaluation of the equity ratio remains in balance, however an unfair evaluation leads to a feeling of inequity (Allen & White, 2002) and research shows that in response the customer will act less favorable to restore the equity ratio.
In this study the feeling of inequity is caused by either the undervalued or overvalued reaction in response to the tip. In case the tip is perceived as low, an overenthusiastic thank-you reaction (acknowledgement) might be considered inappropriate and embarrassing for the customer. In contrast, when the customer believes to give a relatively high tip, the tip acknowledgement of the tip might be expected to be stronger than when a moderate tip is given. We therefore hypothesize that:
H6 The acknowledgement of the tip by the service provider has to be perceived as ‘in line’ with
21
2.3 Conceptual model
Figure 2.3 summarizes the theoretical framework and the hypotheses that have been derived.
Figure 2.3 Conceptual model
H1 Feelings to comply to social norms related to tipping positively influence tipping behavior.
H2 Feelings of gratitude of the consumer towards the service experience positively influence
tipping behavior.
H3 ‘Feelings of gratitude’ will have a stronger effect on tipping than ‘compliance to social norm’
in a country with a preference for feminine values, like the Netherlands.
H4 There is a positive relation between tipping behavior and customer’s future intentions.
H5 The acknowledgement of the tip by the service provider moderates the impact of tipping behavior on customer future intentions; when the acknowledgment is more explicit the tip will have a stronger effect on revisit and word of mouth intentions.
H6 The acknowledgement of the tip by the service provider has to be perceived by the customer as ‘in line’ with the height of the tip, in order for customers to indicate future intentions as more likely. H 5 H 6 + H 2, 3 + H 4 Compliance to Social norm (by customer) Customer future intentions: - revisit intentions - word of mouth Level of explicit acknowledgement (by seller) Tipping (expression of gratitude by customer) Feelings of gratitude (by customer) + H1
22
3. Research method
3.1 Data collection procedure
In order to test the proposed hypotheses a descriptive research design was adopted. The study used a cross-sectional survey approach to collect data related to consumer tipping behavior and the effect acknowledgement of this tip has. Respondents were asked to recall the last time they tipped in a restaurant context; this approach provided a personal, realistic and approachable scenario for the respondent to relate the succeeding questions to. Firstly, the respondents were asked to comment on the gratitude they felt towards the staff and their intentions to revisit and participate in positive word of mouth. Next, the respondent was asked to indicate the amount of the tip left behind and if he perceived the tip to be high or low and if the tip was acknowledged by the service provider. Finally, questions were posed to test the level of compliance to social norm related to tipping. The order of questions aimed to prevent socially desired answers from the respondents related to the relation between the height of the tip, the perceived gratitude for the received service and future intentions. To avoid any language barriers the questionnaire was translated into Dutch.
The questionnaires were designed in Qualtrics, and distributed via the internet. SPSS was used for descriptive statistics and analyses. Prior to issuing the questionnaires, a representative group (in age and educational background) of eighteen respondents were asked to fill out the survey and give feedback, small adjustments in the questionnaire were made accordingly.
3.2 Sample
Data was collected via convenience and self-selection sampling in the Netherlands. Care was taken to achieve a sample which was representative for the Dutch population involved in tipping; in gender, age (>18yrs) and geographically. Self-administrated, internet-mediated questionnaires were distributed via the author’s network and invitations to participate in and distribute the survey were posted on social media. Additionally, hardcopy invitations (appendix I) were handed out to restaurant guests upon their departure of the restaurant. A total of 327 surveys were collected within 14 days of distribution. Of the 327 surveys, 2 respondents indicated never to tip and 21 were incomplete, resulting in 304 valuable surveys. This type of sample ensured a relatively diversified group of respondents, hence enhancing the external validity of the study.
23
3.3 Measure development
To maximize the level of reliability and internal validity measures were borrowed from previous studies, of which some were adapted to suit a restaurant setting. Participants responded to a series of questions and multi-item Likert measures on a five-point scale, ranging from ‘strongly disagree’ (1) to ‘strongly agree’(5).
The first set of questions was developed to set the scene. By asking the name of the restaurant, the type, the amount of people dining, if the visit was for leisure or business, the respondent was forced to recall the situation in which he last tipped. Additionally, the respondent was asked when the visit took place and if he visited the restaurant before.
With respect to the customer reciprocity (revisit intentions and word of mouth) scales were taken from the Marketing Scales Handbook (Bruner, 2009; 2013). The four item construct to test for revisit intention (repatronage intention) originally was tested by Hess, Ganesan and Klein (2003). As for the word of mouth construct a four item construct was originally composed and tested by Bruggen, Foubert and Gremler (2011). In this section of the questionnaire, scales to test for the level of gratitude were also included. These scales were adapted from Palmatier et al. (2009). All three constructs present a satisfactory construct validity and reliability.
Next, the respondent was asked questions in relation to his tip. This study is conducted in a context where guidelines for the tip percentage are vague. The Netherlands belong to the group of countries in which tips can vary from 10% to ten cents. Therefore the perception of whether a tip is high or low is very personal. To take this in consideration we suggest that not only the absolute tip amount and the tip percentage will be of value but also the perception of height of the tip by the consumer. The respondent was therefore asked to indicate the total amount of the bill, the amount they tipped and if they perceived this tip to be low, a little low, not high/low, a little high or high.
No existing scales to measure acknowledgement were found. The reciprocity construct originally used in the research of Arnett, German and Hunt (2003) was able to measure a similar effect and has been used to measure the level of acknowledgement by the service provider in reaction to the given tip. The scale is composed of six, five-point Likert-type statements which validly and reliably measure the degree to which a person believes that his support of a particular organization is truly appreciated, three out of the six statements have been selected
24
and adapted to a tipping setting. Finally, the respondent was asked to indicate if the acknowledgement of the tip by the service provider was in line with the height of their given tip.To test for social compliance in relation to tipping, scales were borrowed from research into tipping motivation scales by Whaley et al. (2014). Finally respondents were asked for several demographical data, like gender, age, educational background and province of residence.
All previously mentioned scales have been borrowed from English academic papers and translated to Dutch by conducting back-translation. Firstly, the English items were translated to Dutch by the author, afterwards the translation was translated back to English by two academics (English and Dutch) both fluent in English. These two translations were compared to the original scales. The scales which differed from the origal scales were reassessed.
For detailed construct items, see appendix II. For the full questionnaire (in Dutch) see appendix III.
4. Results
This chapter provides more insight in the descriptive data of the sample and correlations of the variables of the hypothesized model.
4.1 Descriptive data of sample
The questionnaire has been completed by 327 respondents. Of the 327 surveys, 2 respondents indicate never to tip and 21 were incomplete, resulting in 304 valuable surveys. The distribution between men and women is very close to equal; 50.3% males; 49.7% female. The average age is 41.7 years old and the vast majority of the respondents resides in the East of the Netherlands (60%) and has a higher educational background (74%). The overrepresentation of the latter two is most probably due to the self-selective nature of the data-collection and the networks used to find respondents. The author addressed her own social and professional network to find
respondents, as she lives and works in Deventer at a University of Applied Science in the East of the Netherlands. However, no literature has been found to suggest that people from the East of the Netherlands or with a higher educational background tip significantly differently.
25
professional background in the hotel industry and the support on social media by hospitality related accounts e.g. Nationale Horecagids, may enforced the self-selection of respondents who are positively biased regarding the hospitality industry. For detailed distribution of the sample data refer to table 4.1 below.Table 4.1 Descriptive data of sample (N=304)
Freq. in %
Gender Male 154 50.3
Female 150 49.7
Age (years) M = 41.7, SD =14.8, Min 19, Max 81
<30 68 22.4
30 - 40 99 32.6
41 - 55 63 20.7
>55 74 24.3
Province of residence North (Fri, Dre, Gro, Fle) 16 5.3 East (Gel, Ove) 182 60.0 West (N-Hol, Z-Hol, Utr) 84 27.8 South (Zee, N-Bra, Lim) 22 7.2 Education Lower 80 26.3
Higher 224 73.7
Out of the 304 respondents the vast majority of respondents (74.3%) stated to tip nearly always or even every time they visit a bar or restaurant (table 4.2). The respondents (N=2) who indicated to never tip, and who have been excluded from the sample, are unlikely to be representative for the population due to the self-selective approach of this study. People’s interest to participate in the study was raised by asking them about the last time they dined out and tipped, it is not likely that people who do not tip would feel addressed to participate in the survey. Although 74.3% states to (almost) always tip, out of the 304 respondents only 47.2% agree that tipping in restaurants is the social norm in the Netherlands, 23.8% disagrees with this statement and 29% is indifferent (M = 3.22, SD =1.095).
26
Table 4.2 Frequency tipping in hospitality context
Freq. in %
Frequency tipping Sometimes 24 7.9
Often 54 17.8
Almost every time 170 55.9 Every time 56 18.4
4.2 Variables and measurements
4.2.1 Reliability of scales
Firstly the counter-indicative item in the ‘revisit intention’ construct was recoded. Next, the reliability for all construct items was tested. Although all constructs showed sufficient reliability in first instance (all Cronbach’s Alpha scores >.70) the constructs ‘word of mouth’ and ’revisit intention’ would benefit from deleting an item. Within ‘word of mouth’ the originally counter-indicative item was removed and within ‘revisit intention’ the item ‘I am likely to say positive things about ____ in general to other people’ was removed. As a result the Cronbach’s Alpha scores changed from .873 into .928 for ‘word of mouth’, and from .847 to .880 for ‘revisit intention’. The strong Cronbach’s Alpha scores for all constructs allowed for the items in the constructs to be averaged, for further details see table 4.3 below.
Table 4.3 Cronbach’s Alpha scores for construct scales
N=304 Cronbrach’s
Alpha
N of items M SD
Feelings of gratitude .893 3 3.90 .835 Compliance to social norm .833 5 2.87 .855 Revisit intention .880 2 3.98 .966 Word of mouth .928 3 3.93 .891 Acknowledgement tip (N=252) .936 3 3.78 1.0
4.2.2 Skewness and kurtosis
The data was checked for skewness and kurtosis and, as expected for most scores, showed signs of non-normally distribution. The questionnaire was only posed to people who tipped, suggesting some sense of positive sentiment, therefore the buildup of high scores was
27
expected. As a rule of thumb some researchers accept skewness and kurtosis values from -2 to +2, more conservative views use a range from -1 to +1. As presented in table 4.4 most constructs show evidence of negative skewness, with the exception of ‘tip percentage. However, after inspecting the shapes of the distribution the values were assessed as acceptable. Further support for this decision is the large sample (N=304), this could also cause the rejection of normal distribution, therefore no further action to transform the variables was taken (Tabachnick & Fidell, 2001).Table 4.4 Skewness and kurtosis scores
N=304 Skewness Std. Error Kurtosis Std. Error
Feelings of gratitude -.994 .140 1.149 .279 Compliance to social norm -.138 .140 -.242 .279 Tip percentage 1.327 .140 4.073 .279 Tip – perception of height -.682 .139 2.335 .278 Revisit intention -1.074 .140 .708 .279 Word of mouth -1.030 .140 1.181 .279 Acknowledgement tip (N=252) -1.676 .153 2.709 .306 Acknowledgment – Tip ratio (N=252) -1.062 .153 2.256 .306
4.2.3 Correlation Matrix
Table 4.5 presents the correlations between the variables. Included in this correlations matrix are the independent variables (feelings of gratitude, compliance to social norm, tip percentage/ perception of height), the moderating variables (acknowledgement of tip/ ratio of acknowledgement), the dependent variables (revisit intentions and word of mouth) and the control variables (gender, age and education level). The analysis presents several significant and relevant (r = higher or close to .3) correlations between certain independent, moderating and dependent variables, these are described below.
The positive correlation between the tip percentage and the customers perception of the height of their tip suggests that when the percentage goes up, so does the perception, however the relationship is moderate (r = .354 p <.01). Only the perception of the height of the tip (not the actual tip percentage) presents a close to moderate relationship with word of mouth intentions and the intention to revisit (respectively r = .264 p <.01; r = .224 p <.01).This suggests that when the perception of the height of the tip increases, respondent’s intention to participate in word of
28
mouth and revisit the restaurant in the future also goes up. Also, the higher the respondent perceives it’s tip to be, the more positive he assesses the acknowledgement of his tip (r = .275 p <.01).As for the correlation with feelings of gratitude the perception of the height of the tip (r = .285 p <.05) is positive. When feelings of gratitude go up, so does the percentage of the tip and the perception of it.
The construct of gratitude shows more significant and relevant correlations. Feelings of gratitude towards the restaurant staff correlates significantly with revisit intention, word of mouth, the level of acknowledgement of the tip and the ratio of acknowledgement. A positive relationship exist between gratitude and revisit intentions (r = .596 p <.001) and the intention to participate in word of mouth (r = .738 p <.001). These strong correlations suggest that the more gratitude respondents feel toward the restaurant staff the more likely they are to revisit and even more likely to speak positively about the restaurant. This direct relation is much stronger than the indirect relation via tipping behavior. The strong correlations between gratitude and the acknowledgement of the tip (r = .494 p <.001) could suggest two things, either the respondents who feel more gratitude have a positive bias towards the acknowledgement of the tip and therefore are more receptive of the acknowledgement, or the acknowledgement of the tip enforces feelings of gratitude. The same applies for the correlation between gratitude and if the level of the acknowledgement is in correspondence with the height of the tip (r = .418 p <.001).
Additional significant correlations exist for the moderating variables acknowledgement of tip and the accordance of the acknowledgement with height of the tip. Both variables correlate positively with the intention to revisit (respectively r = .383 p <.001; r = .486 p <.001) and participate in word of mouth (respectively r = .296 p <.001; r = .422 p <.001). This suggests that the more respondents feel that their tip is acknowledged by the restaurant staff and the more the acknowledgement exceeds expectation in correspondence with the tip, greater the chance that respondents will revisit and speak positively about the restaurant. Also, both moderating variables correlate with each other (r = .757 p <.001), indicating that when the respondent assesses the acknowledgement of the tip as more positive the evaluation of the reaction/tip ratio also goes up.
29
Both dependent variables revisit intentions and word of mouth correlate positively (r = .751 p <.001), if respondents intent to speak more positively about the restaurant they are also more likely to revisit the restaurant again.30
Table 4.5 Correlation matrix
Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 1. Gender 1.51 0.5 - 2. Age 41.7 14.81 .373** - 3. Education level 0.73 0.444 -.163** -.354** - 4. Tip – absolute 6.95 8.34 .161** .12* -.175** - 5. Tip percentage 8.11 4.15 -.047 -.029 -.075 .081 -6. Tip - perception height 3.01 0.682 .11 .06 .038 .146* .354** - 7. Feelings of gratitude 3.9 0.835 .09 .07 -.113* .15** .166** .285** (.89) 8. Compliance to social norm 2.87 0.86 -.119* -.059 -.058 .055 .08 .06 .06 (.83) 9. Revisit intention 3.98 0.966 .02 .004 -.062 .075 .11 .224** .596** .05 (.88) 10. Word of Mouth 3.93 0.891 .05 .018 -.081 .134* .122** .264** .738** .067 .751** (.928) 11. Acknowledge-ment of tip 3.78 1.002 .11 .096 -.12 .135* .108 .275** .494** .155* .383** .486** (.936) 12. Ratio acknow-ledgement tip 2.87 0.724 .09 .159* -.175** .13* .041 .113 .418** .052 .296** .422** .757** -
**. Correlation is significant at the 0.01 level (2-tailed) *. Correlation is significant at the 0.05 level (2-tailed)
31
4.2.4 Mean tip percentages – between groups
The tip percentages were calculated for each respondent by dividing the tip size by the bill amount and then multiplying this quotient by 100. On average the respondents gave an 8.1% tip (SD = 4.15). The median lays at 7.4%, the difference is due to the extreme outliers (minimum tip of 0.7%, maximum of 31.8%). Although the tip percentages differ within certain groups, only some differ significantly. Independent sample t-tests showed that respondent below age 30 tip significantly higher (M = 8.9 SD = 5.27, p < .05) and those between 30 and 40 years old tip significantly lower (M = 7.4 SD = 3.26, p < .05). The size of the bill also makes a difference, with bills below or equal to €25 respondents tip significantly more (M = 11.3 SD = 5.83 p < .01) and when the bill exceeds €50 the tip percentage decreases significantly (€50 < x ≤ €100, M = 6.9 SD = 3.21 p < .05) (> €100, M = 7 SD = 3.04 p < .05). The amount of times respondents have visited the restaurant also makes a difference, respondents who have never visited the restaurant before (M = 6.9 SD = 3.48 p < .01) tip lower, those who visit the restaurant regularly tip higher (M = 8.9 SD = 4.7 p < .05).
The average tip percentage differed between groups, the mean tip percentages are presented in table 4.6 below.
Table 4.6 Mean tip percentage
in % Overal (N=304) 8.1 Gender Male 7.9 Female 8.3 Age (years) <30 8.9* 30 - 40 7.4* 41 - 55 8.3 >55 8.1
Province of residence North (Fri, Dre, Gro, Fle) 10.1 East (Gel, Ove) 7.9 West (N-Hol, Z-Hol, Utr) 8.5 South (Zee, N-Bra, Lim) 6.9 Education Lower education 8.6 Higher education 7.9
32
Bill size < €25 11.3** €25 - €50 8.7 €50 < x ≤ €100 6.9** > €100 7.0** Previous visit restaurant Has never been before 6.9** Has been before 8.2 Has been regularly 8.9* **. Tip percentage in group is significant different at the 0.01 level (2-tailed)
*. Tip percentage in group is significant different at the 0.05 level (2-tailed)
4.2.5 Tip percentage vs. perceived height of tip
Table 4.4 shows a moderate positive correlation between the tip percentage and the perception respondents have regarding their tip (r = .353 p <.01). Further analysis of the data illustrates this observation further. Respondents who indicate to have given a low tip indeed gave the lowest tip on average and those who regarded their tip to be high also gave the highest tip (also on average). These difference are significant (Kruskal-Wallis test χ²= 37.76; df = 4; p <.001). Refer to table 4.6 for the detailed mean rank scores. The average tip percentage for low perceived tips is 3.96% (SD = 2.18, min. 0.71%, max. 9%), for ‘a little low’ 5.68% (SD = 2.73, min. 1.56%, max. 12.5%). The majority of the respondents perceived their tip not to be low of high, the average tip percentage for this group is 8.1% (SD = 3.63, min. 2%, max. 25%). For those who indicated to have given a little high tip the average tip percentage was 9.98% (SD = 5.45, min. 2.%, max. 31.82%), only 3 respondents classified their tip a high, they gave an average tip percentage of 15.4% (SD = 5.15, min. 9.52%, max. 19.4%).
Table 4.6 Kruskal- Wallis test: percentage vs. perception
Do you perceive this to be a low or a high tip? N Mean Rank Tip percentage
Low 12 52.83
A little low 25 96.92
Not low/ not high 215 155.34
A little high 49 185.80
High 3 266.67
33
4.3 Hypothesis testing
The next section will test the hypotheses derived from the proposed conceptual model (figure 4.1) below.
Figure 4.1 Conceptual model
Please note,
t
ipping behavior has been measured in three ways. Firstly, respondents were asked to indicate the absolute amount of the given tip. Secondly, the respondent was asked to indicate if he perceived this tip to be low, a little low, not high/low, a little high or high. Thirdly, for each respondent the tip size was divided by the bill amount and then multiplying this quotient by 100 by the researcher to calculate the tip percentage.Hypothesis 1, 2 and 3
1. Feelings to comply to social norms related to tipping positively influence tipping behavior. 2. Feelings of gratitude of the consumer towards the service experience positively influence
tipping behavior.
3. Feelings of gratitude’ will have a stronger effect on tipping behavior than ‘compliance to social norm’ in a country with a preference for feminine values, like the Netherlands.
A multiple regression analysis was used to test if the independent variable ‘feelings of gratitude’ had a stronger predictive effect on tipping behavior than the independent variable ‘compliance
H 5 H 6 + H 2, 3 + H 1 + H 4 Compliance to Social norm (by customer) Customer future intentions: - revisit intentions - word of mouth Level of explicit acknowledgement (by seller) Tipping behavior (expression of gratitude by customer) Feelings of gratitude (by customer)
34
to social norm’, simultaneously illustrating the isolated effects of both independent variables. For the absolute tip amount (table 4.7), the multiple regression indicated that the two predictors explained 2.5% of the variance (R²=.025, F(2,300) = 3.858 p < .05). The two predictors explained 3.4% of the variance in the tip percentage (R²=.034, F(2,300) = 5.358 p < .05 – table 4.8) and the analysis showed the two predictors explained 8,3% of the variance in the perception of the tip (R²=.083, F(2,300) =13.616 p < .05 – table 4.9).Table 4.7 Multiple regression model: tip – absolute amount
R R² R² change B SE β t Model 1 .158 .025* .019 Social Compliance .451 .558 .046 .807 Feelings of Gratitude 1.491 .572 .149** 2.605 * statistical significance: p < .05 ** statistical significance: p < .01
Table 4.8 Multiple regression: tip percentage
R R² R² change B SE β t Model 1 .186 .034** .028 Social Compliance .338 .275 .070 1.227 Feelings of Gratitude .835 .282 .168** 2.955 ** statistical significance: p < .01
Table 4.9 Multiple regression: tip – perception of height
R R² R² change B SE β t Model 1 .288 .083*** .077 Social Compliance .030 .043 .039 .039 Feelings of Gratitude .226 .044 .284*** 2.955 *** statistical significance: p < .001
All three models (tip amount/ percentage and perception) showed that ‘feelings of gratitude’ significantly predicted tipping behavior: tip amount (β= .149, p < .01), tip percentage (β= .168, p < .01) and perceived height of the tip (β= .284, p < .001).
