The dining and tipping behaviour of Black South Africans: a segmentation approach

12The literature on dining and tipping behaviour has focused mostly on the United States of America (USA), while minimal research has been conducted in African countries. While there is a negative and grounded perception surrounding black diners being poor tippers in the USA, hardly any research has focused on the dining or tipping behaviour of this dining market from a developing country perspective. The intention of this exploratory research was to fi ll the current knowledge gap by segmenting black South Africans on the basis of their motives for dining out and to determine the differences in the dining and tipping behaviour of the different segments. To target potential black diners, a visitor survey was conducted at the Cape Town International Jazz Festival. A total of 256 usable questionnaires were returned and included in the analysis. Socialisation, gastronomy enjoyment, lifestyle and escape and status were identifi ed as the four motives for dining out. Based on these motives, different black dining segments were identifi ed and an OSI (Occasionalists, Socialisers and Indulgers) typology of diners proposed. The dining and tipping behaviour of these dining segments are furthermore infl uenced by several factors, with clear implications for both the server and restaurateurs. The results shed light on the dining and tipping behaviour of black South African diners and showed that this dining market cannot be regarded as bad tippers. 13

literature on dining and tipping behaviour has focused mostly on the United States of America (USA), while minimal research has been conducted in African countries. While there is a negative and grounded perception surrounding black diners being poor tippers in the USA, hardly any research has focused on the dining or tipping behaviour of this dining market from a developing country perspective. The intention of this exploratory research was to fi ll the current knowledge gap by segmenting black South Africans on the basis of their motives for dining out and to determine the differences in the dining and tipping behaviour of the different segments. To target potential black diners, a visitor survey was conducted at the Cape Town International Jazz Festival. A total of 256 usable questionnaires were returned and included in the analysis. Socialisation, gastronomy enjoyment, lifestyle and escape and status were identifi ed as the four motives for dining out.

Introduction
1T ipping is a vital source of income for a variety of occupations in the hospitality and tourism industry. Zeigler-Hill, Besserb, Vrabela and Noserc (2015) revealed that nearly 80% of employed individuals in the United States of America (USA) work in the service industry, of which nearly nine million of them work as servers in bars and restaurants. Azar and Tobol (2008) state that these servers earn in excess of $44 billion in tips per annum. According to Mkono (2011), one of the reasons why research in this field is important, in addition to the economic value attached to tipping, is that tipping behaviour differs from country to country, and one would therefore expect it to also differ from one culture to the next. In this regard, many restaurant servers believe that ethnic minorities are poor tippers. In the US restaurant industry in particular, Asians, blacks, and Hispanics are widely perceived to be poor tippers (McCall & Lynn 2009;Lynn & Thomas-Haysbert 2003;Lynn, 2014;Lynn, Pugh & Williams 2012). These perceptions, especially by US servers, are unfortunately largely consistent with empirical evidence (Noll & Arnold 2004;Lynn 2006b;2013;McCall & Lynn 2009).
Studies have found that black customers are less likely than whites to base their tips on bill size and that blacks leave smaller tips as a percentage of the bill compared to whites (Lynn 2004(Lynn , 2009Lynn & Thomas-Haysbert 2003;Lynn, Sturman, Ganley, Adams & McNeil 2008). Moreover, these race differences persist even after controlling for customer education and income, and for service quality (Lynn 2006a). Thus, the race differences in tipping cannot be dismissed simply as reflecting socioeconomic differences across racial and ethnic groups (Lynn & Brewster 2015). The perception that blacks tip less than whites poses numerous problems, ranging from discriminatory service -as it leads to many servers disliking waiting on black tables and delivering inferior service to those blacks seated in their sections (Rusche & Brewster 2008) -to restaurant executives' reluctance to open restaurants in minority communities (Wallace 2001;Amer 2002). Lynn (2011) suggests that solving these problems requires either a reduction in the race differences in tipping or the use of something other than tips to motivate waiters. Racial differences in tipping therefore need to be substantially reduced, which requires an understanding of their underlying causes (Lynn & Brewster 2015).
While the majority of research on tipping has been conducted in the USA, the opposite is true when one looks at developing countries, and specifically African countries. A review of the literature review indicated that the few studies on tipping in Africa included those by Mkono (2011), Saayman (2014) and Saayman and Saayman (2015). The studies by Saayman (2014) and Saayman and Saayman (2015) both focused on the tipping behaviour of white South Africans. No study has to date examined the tipping behaviour of black South Africans and whether the general US perception that black diners tip less, are confirmed or contradicted in the context of the country. The purpose of this research, however, was not to replicate and extend the results of previous studies. Because South Africa is a multi-ethnic society encompassing a wide variety of cultures with 11 official languages, attempting such a feat would be difficult.
Instead, this exploratory research followed a market segmentation approach as a first step in understanding black South African diners' dining and tipping behaviour. Segmentation ultimately attempts to match the expectations of different groups of customers with offers made from services suppliers, or in this instance, restaurants. Once the market has been segmented into different segments or target markets, a customised marketing mix aimed at each segment is designed to position the product for each market (Dolnicar, Grün & Leisch 2016). Although past studies have focused on the identification of factors that may differentiate dining segments (Oh & Jeong 1996), no study has to date (and to the authors' knowledge) segmented a specific ethnic group in order to evaluate any differences in dining and tipping behaviour. To perform this market segmentation, the most widely used segmentation method, namely a factor-cluster analysis, was adopted (Lee, Lee & Wicks 2004;Chang 2006). The purpose of this research was to determine the dining motives of black South Africans and, based on these motives, to segment and profile different black South African dining segments. Moreover, this research attempted to determine the dining behaviour of the different segments as well as the factors that influence the size of their tips and why they do or do not tip. Considering the negative perception surrounding black diners, a greater understanding of the dining and tipping behaviour of black South African customers might be useful for restaurateurs and servers alike (Brewster 2015).
Literature review 1The next sections provide an overview of the relevant literature relating to dining and tipping behaviour as well as the findings from cross-cultural tipping and market segmentation studies.

Predictors of dining behaviour
1Diners have an overabundance of restaurant choices (Harrington, Ottenbacher & Kendall 2011). To be successful in this competitive business environment, restaurant managers need to understand what motivates consumers to dine out and to choose one restaurant over another (Riehle, Grindy, Stensson & Smith 2014). Factors influencing consumers' decision making when selecting a restaurant have varied, depending on context and motivations for dining out (Choi & Zhao 2010).
When dining out, people constantly seek comfort as well as quality and an enjoyable environment (Choi & Zhao 2010). Harrington et al. (2011) found that restaurant customers often make dining decisions by simultaneously evaluating several criteria. Customers might consider food quality, price, promotions and recommendations, among other desirable benefits (Riehle et al. 2012). Arguably, the dining experience also relies on the value-for-money aspect of food and service components (Alonso & O'Neill 2010). Various motives have been identified for dining out at restaurants, which include efficiency, taste, socialisation, health, the external environment, economic factors and entertainment (Park, 2004;Choi & Zhao 2010). Additionally, people often dine at casual restaurants when exhausted, pressed for time or when already out (Epter 2009). However, besides serving excellent food, good service, the location of the restaurant, its decor and intangible aspects such as the ambience and atmosphere are key factors essential to diners choice of restaurants (Yüksel & Yüksel 2002;Sparks, Bowen & Klag 2003;Gäl, Akbay, Özcicek, Özel & Akbay 2007;Correia, Moital, Da Costa & Peres 2008;Kim, Ng & Kim 2009;Chang, Kivela & Mak 2011;Josiam, Kalldin & Duncan 2015).

Predictors of tipping behaviour
1A ccording to Thomas-Haysbert (2002), tipping is the voluntary act of leaving money in appreciation for a service received. The word "tip" is sometimes considered as an acronym for "to insure promptness" or "to insure prompt" service (Fong 2005). Tipping follows two main standardised guidelines. One guideline defines the tip as a fixed percentage of the bill. A second guideline suggests that the tip is compensation for service and a monitoring tool for quality (e.g. Lynn & Graves 1996;Bodvarsson & Gibson 1997;Lynn & McCall 2000). In agreement with these guidelines, the literature on tipping constantly indicates four main predictors of tip size, namely the bill, group size, service quality and service effort (or quantity). However, the tipping decision is not only influenced by these four factors. Various authors have found that several other factors influence the tipping decision (Azar 2010;Saunders & Lynn 2010;Becker, Bradley & Zantow 2012;Lynn 2015;Saayman & Saayman 2015). Rind and Bordia (1996) and Rind and Strohmetz (1999) divided tipping behaviour into the following three basic categories: the characteristics of the customer, the characteristics of the server/waiter(ess) and, lastly, the interaction between the server and the characteristics of the customer. Saayman (2014) added another category, namely external factors (e.g. weather). Each of these categories is now briefly discussed.
The first category has to do with the characteristics of the customer, which include sociodemographic variables such as gender, where all male groups tend to tip more than all female groups (Stillman & Hensley 1980), although Saayman and Saayman (2015) did find that females tip more frequently than males. However, when one controls for education and income, according to Thomas-Haysbert (2002), the results showed that the differences between white and black relating to tipping, seem insignificant. In general, income showed a positive relationship with tipping, as did education (Lynn & Thomas-Haysbert 2003;Saayman & Saayman 2015). One of the behavioural variables that has the greatest impact on the size of the tip is the size of the bill (Lynn & McCall 2000;Green, Meyerson & Schneider 2003;Saayman & Saayman 2015). Regarding the method of payment, a positive relationship was reported between credit card payment and size of the tip (Garrity & Degelman 1990;Lynn 2006aLynn , 2006b. Alcohol consumption also revealed a positive correlation between it and the size of the tip (Sanchez 2002;Conlin, Lynn & O'Donoghue 2003). Similarly, the mood of the diner has a positive relationship with tipping (Lynn & McCall 2000), while Greenberg (2014) found that if people are on holiday, they also tend to tip more. Although numerous studies have been conducted on tipping, very few have combined the reasons for dining out in their analysis to help clarify why people dine out and how this influences tipping. A study by Saayman and Saayman (2015) added this factor. They found that it should be part of understanding tipping behaviour at restaurants where they had identified three reasons for dining out, namely status, gastronomy (quality food) and its enjoyment, and socialisation.
On the question of why people do tip, several reasons or motives have been identified. These include the following: rewarding good service (Azar 2005;Lynn 2003Lynn , 2009Lynn , 2015Saayman & Saayman 2015), which also guarantees future service delivery (Lynn 2009;Saayman 2014); building honest character (Lynn 2006a); helping the servers (Azar 2004;Saunders & Lynn 2010;Saayman & Saayman 2015); pressure to conform to societal approval and status (Azar 2004(Azar , 2005Boyes, Mounts & Sowell 2004;Saayman & Saayman 2015); a sense of duty or obligation (Azar 2004;Lynn 2015); making the tipper feel positive (Lynn 2006a); and simply the fact that people support the practice of tipping (i.e. social convention) (Lynn 2006a;Lynn & McCall 2016). Unfortunately, less research has been focused on the reasons why people do not tip, compared with those who do.
The second category concerns the server and his or her interaction with the customer. Here, a variety of characteristics play a role, such as the attractiveness of the server, where more attractive servers receive a larger tip (Stillman & Hensley 1980;Hornik 1992;Lynn & McCall 2000;Lynn & Simmons 2000;Lynn & McCall 2016). Language has a positive relationship with tipping, especially when the server addresses the customer in his or her mother tongue (Van Vaerenbergh & Holmqvist 2013). A study by Seiter (2007) also revealed that flattery leads to a larger tip, which applies to gestures such as conveying messages and drawing faces on the bill (Rind & Bordia 1995, smiling at customers (Tidd & Lockard 1978) and briefly touching customers (Hornik 1992). The most important aspect of this category, the level of service, also revealed that when servers render a better or quality service, it leads to greater tips (Schwartz 1997;Lynn & McCall 2000;Conlin et al. 2003;Lynn 2001Lynn , 2003Chung & Heung 2007). Conversely, Lynn and McCall (2016) found that service quality and cost considerations seem to have only weak effects on tipping.
The third category considers external factors that also influence travel behaviour and consequently, the size of the tip. Rind and Strohmetz (2001) found that in sunny weather, diners give larger tips -hence it could be said that weather influences tipping behaviour. In addition, the size of the town or city also has a positive relationship with tips, which implies that people dining in larger cities tend to give more substantial tips (Garrity & Degelman 1990;Rind & Strohmetz 2001;Lynn & Thomas-Heysbert 2003) as well as the location and whether the restaurant is elegant. The aforementioned research collectively indicates that a variety of factors influence tipping behaviour, and while certain factors overlap among diners, there is currently no universal set of variables that explain tipping behaviour.

Cross-cultural tipping studies
1Cross-cultural comparisons have examined the differences in tipping practices between different countries. Lynn, Zinkhan and Harris (1993) found that tipping was less prevalent in countries with low tolerance for interpersonal status and power difference and in countries with values that emphasise social over economic relationships. According to Fong (2005), although the amount of tips and the persons who are tipped vary from culture to culture, there are norms that most would follow. However, Ogbonna and Harris (2002) emphasise the fact that although tipping is an internationally recognisable behaviour, the actual practice is heavily influenced by societal cultural considerations; some cultures are more committed to tipping than others. For instance, according to UNIGLOBE (2014), tipping is not preferred in most of the local restaurants in China, while 15 to 20% of tips are expected in most US and Canadian restaurants. In most European restaurants, the norm is to tip 10 to 15% if there is no service charge. Tipping is perceived as extremely insulting in Japan, while it is illegal to tip in countries like Argentina and Vietnam. In South Africa, the tip is between 5 and 10% when no service has been charged. It is evident that most developed countries are committed to tipping. The reason for this could be their different culture and higher per capita income (Fong 2005).
To date, most of the research on race differences in tipping has been conducted in the USA. Lynn (2006) found that knowledge of the restaurant tipping norm is greater among people who are white, in their forties to sixties, highly educated, wealthy, living in metropolitan areas, and living in the North East, than among their counterparts. These findings support the idea that differential knowledge of tipping standards underlies geodemographic differences in tipping behaviour. Other findings suggest the following: (1) blacks leave smaller average restaurant tips than whites; (2) black-white differences in tipping persist after controlling for socioeconomic status; (3) blacks tip less than whites even when provided with comparable levels of service; (4) blacks tip less than whites even when the server is black; and (5) blacks are much less likely than whites to know that it is customary/expected to tip between 15 and 20% of the bill size in US restaurants (Lynn & Graves, 1996;Lynn, Le & Sherwyn 1998).
Ethnic differences in tipping may be influenced by the following: (1) discrimination in the delivery of service to ethnic minorities since it is generally believed they tip less (Lynn & Graves 1996;Lynn 2000); (2) a greater number of female-headed households, large families/dining parties and low incomes among ethnic minorities (Dinkins 1994;, and/or (3) low familiarity with and internalisation of tipping norms among ethnic minorities (Lynn & Thomas-Haysbert 2003;Lynn 2004;Lynn 2006a). Lynn (2011), Brewster (2012a, 2012b and Lynn and Brewster (2015) confirmed that race differences in tipping can be sizably reduced (though not eliminated) by educating blacks about appropriate tipping norms through educational campaigns by restaurants. For example, restaurant managers can inform and remind their customers about the tipping norm percentage by putting appropriate information on menus, table tents and cheques or adding the words "suggested tips" to cheques and credit-card slips (Lynn 2004(Lynn , 2014(Lynn , 2015. In one of the few African studies, Mkono (2011) analysed tipping practices and policies in Zimbabwe's hotel industry from the perspective of restaurant waiters. The majority of respondents reported that discretionary tipping, as opposed to service charges, is the norm and that most hotels have a tip-pooling system. The findings show that guests who are expected to tip well, based largely on stereotypes as well as past tipping behaviour, receive better service treatment from waiters. In a South African context, Saayman (2014) and Saayman and Saayman (2015) found that the main reasons why white South Africans dine out include socialising, then experiencing the food (gastronomy) and, to a lesser degree, status. The reasons why white South Africans tip were financial, good service and social acceptability, while bad service and because they perceive it as inessential (unnecessary) were the main reasons for not tipping. To date, no study has focused on the dining or tipping behaviour of black South African diners or has attempted to segment the South African dining market.

Market segmentation of diners
1Only a few studies that have investigated different segments in the restaurant sector (Bojanic & Shea 1997). In one of the first studies, Lewis (1981) investigated the users and nonusers of family, theme and gourmet restaurants and found that segments differed in their opinions about the importance of several service attributes. Swinyard and Struman (1986) identified three customer segments, namely family diners, romantics and entertainers, based on an analysis of customers' value benefit factors, lifestyle factors, usage patterns and demographic descriptors. When investigating customer expectations of fast-food restaurants, Oh and Jeong (1996) found four different customer segments, namely neat service seekers, convenience seekers, classic diners and indifferent diners. Granzin and Olsen (1997), however, revealed the following three groups of consumers relating to fast-food restaurants: nonusers, light users and heavy users. Bahn and Granzin (1985) tested how nutritional concerns could affect restaurant patronage. They discovered four distinct segments, namely the health, gourmet, value and unconcerned segments. Their study revealed that these four segments were likely to patronise different restaurant types. For example, the health segment is unlikely to frequent fast-food restaurants compared to the value group. Shoemaker (1998) discovered the following five distinct segments among university students: perceptive shoppers, expedient shoppers, 24-hour social students, focused diners and demanding diners. Yüksel and Yüksel (2002) investigated tourists who had dined in independent Turkish restaurants and based their segmentation on the factors that influence restaurant selection (service quality, product quality/hygiene, adventurous menu, price and value, atmosphere, healthy food, location and appearance, availability of a non-smoking area and the visibility of the food preparation area). Five segments were identified, namely value seekers, service seekers, adventurous-food seekers, atmosphere seekers and healthy-food seekers. In a more recent study, Josiam et al. (2015) identified two dining segments based on their motives for dining out: kitchen-challenged diners (unwilling or unable to cook) and social suppers (those who eat out for social reasons, say, to accompany friends or family, or to celebrate a special occasion).
A review of these studies clearly demonstrates that distinct customer groups are evident in the total dining market. The existence of different segments in the dining market requires restaurant managers to attend to different factors that would appeal to the targeted segment(s) in their marketing communications (Yüksel & Yüksel 2002). The current research should not only add to the body of knowledge on the dining and tipping behaviour of black South Africans. The study should also provide a segmentation typology of these diners in the country and the factor that distinguishes them in terms of their dining and tipping behaviour.

Method of research
Study focus 1The Cape Town International Jazz Festival is currently ranked as the number four jazz festival in the world and is known as "Africa's grandest gathering". The festival is held annually at the Cape Town International Convention Centre in Cape Town. A highly anticipated part of the Cape Town International Jazz Festival is the complimentary live concert held at Cape Town's historical Greenmarket Square, which is a fitting precursor to the main event. The concert usually takes place on the Wednesday before the festival weekend. The six-hour performance allows audiences to catch a glimpse of what to expect at the weekend-long main festival and gives those who missed out on tickets -or could not afford to buy them -a chance to enjoy the music of some of the country's best jazz artists. Since the majority of visitors to the festival as well as at the free concert are black South Africans (Tourism Research in Economic Environs and Society 2015), both events were the ideal setting to research the tipping behaviour of both the low-to middle-class-income black South Africans (at the free concert) and the middle-to high-income black South Africans (at the festival).
Questionnaire 1Since this was a quantitative study, a structured questionnaire was used to collect the data. The questionnaire had three sections. Section A captured sociodemographic details: gender, age, province and country of residence, level of education, marital status, group size, length of stay, annual income, expenditure, tipping behaviour and frequency of dining out. This section also captured motivational factors for dining out, measuring 19 items on a five-point Likert scale of agreement, ranging from 1 = "strongly disagree" to 5 = "strongly agree". Section B measured 26 aspects of the respondents' tipping decision making, while section C measured 11 and 13 items respectively, as reasons for tipping and not tipping. In both of section C's parts, the items were also measured on a five-point Likert scale of agreement. The measurement methods used in sections B and C were assessed according to validity and reliability criteria. Sections B and C satisfied the criteria for content validity because the variables included in the section were based on the studies described in the literature review. The convergent construct validity of the variables in sections B and C was tested by means of factor analysis to determine the combination of factors in which the variables were most consistent (Zikmund, Babin, Carr & Griffin 2010). The reliability of the variables in sections B and C was measured using the alpha coefficient, which represents the average of all possible split-half reliability for a construct (Zikmund et al. 2010). The alpha coefficient indicates the quality of the measurement: a score ranging from 0.60 to 0.70 indicates fair reliability, 0.70 to 0.80 good reliability and 0.80 to 0.95 very good reliability (Zikmund et al. 2010).
Sampling method and survey 1In total, 300 questionnaires were distributed over a period of three days (150 at the free concert on 25 March and 150 at the main festival on 27 and 28 March 2015) using simple random sampling. A total of 285 questionnaires were returned sufficiently completed. Since their motives were used to cluster the respondents, only questionnaires in which this section was fully completed could be included in further analysis. Furthermore, since only black South Africans were analysed, all foreign respondents were excluded from the analysis. This resulted in 256 usable questionnaires. All questionnaires were completed at the main festival grounds, where fieldworkers moved around to minimise bias. Adult visitors were randomly selected as they entered or left these areas and when they were sitting down and relaxing at the various venues. The fieldworkers were trained to ensure that they understood the aim of the study and the questionnaire and could brief respondents about the purpose of the research. Fieldworkers were also instructed to target black attendees at the respective events.
Profi le of respondents 1More female (56%) than male (44%) respondents participated in the survey. The respondents were predominantly English speaking (73%), on average 40 years of age, and originated in the Western Cape (62%) or Gauteng (18%) provinces. The respondents spent an average of five nights in the area of Cape Town, travelled in a group of five persons, spent an average of R6081.67 per group, and were in a professional (17%) or management occupation (16%). Regarding their tipping behaviour, respondents indicated that they either always tip (42%) or sometimes tip (30%) when they dine out at least once a month (26%) or occasionally dine out (22%).

Statistical analysis
1The data were captured using Microsoft Excel ® and analysed using SPSS ® (Statistical Package for the Social Sciences) (2015). The analysis was conducted in three stages: four-factor analyses (respondents' motives for dining out, their tipping decision making as well as the reasons why they tip and do not tip), a cluster analysis based on respondents' dining motives, and an analysis of significant differences between different segments of black South African diners.
Principal component factor analyses, using an Oblimin rotation with Kaiser normalisation, were performed on, respectively, the 19 factors of motivation, 26 of tipping decision making, 11 reasons for tipping and 13 reasons for not tipping items to explain the variance-covariance structure of a set of variables through a few linear combinations of these variables. The Kaiser-Meyer-Olkin measure of sampling adequacy was used to determine whether the covariance matrix was suitable for factor analysis. Kaiser's criteria for the extraction of all factors with eigenvalues larger than 1 were used because they were considered to explain a significant amount of variation in the data. All items with a factor loading greater than 0.3 were deemed to contribute to a factor, while all items with loadings less than 0.3 were regarded as not correlating significantly with this factor (Steyn 2000). This resulted in 15 tipping decision-making factors, nine reasons for tipping and ten reasons for not tipping that could be included in further analysis. Any item that cross-loaded on two factors, with a factor loading greater than 0.3, was categorised in the factor where interpretability was best. A reliability coefficient (Cronbach's alpha) was computed for each factor to estimate its internal consistency. All factors with a reliability coefficient above 0.6 were considered acceptable in this study. The average inter-item correlations were also computed as another measure of reliability -according to Clark and Watson (1995), these should lie between 0.15 and 0.55.
The cluster analysis was performed on the scores of the motivational factors. This research did not take an a priori view of which data points should fall into which segment; instead, a hierarchical cluster analysis was used to explore the natural structure of the data, by means of Ward's method with Euclidean distances. Hierarchical clustering makes no assumptions about the number of groups or group structure. Instead, the members are grouped on the basis of their natural similarity (Johnson & Wichern 2007).
The analysis of significant differences used analysis of variances (ANOVAs) to investigate any significant differences between the identified market segments and their dining and tipping behaviour. Effect sizes (d) were utilised to identify any further significant differences between the segments. Cohen (1988), Ellis and Steyn (2003) and Steyn (2009) offer the following guidelines for interpreting the effect sizes: small effect: d = 0.2, medium effect: d = 0.5 and large effect: d = 0.8.

Results
1This section presents the results of the factor analyses, the cluster analysis and ANOVAs to investigate significant differences.

Results from the factor analyses
1Using an Oblimin rotation with Kaiser normalisation, the pattern matrix of the principal component factor analysis identified four motivational factors (Table 1), three tipping decision-making factors (Table 2) and two factors for why respondents tip and do not tip, respectively (Table 3). These were labelled according to similar characteristics and accounted for 56%, 56%, 51% and 58% respectively of the total variance. All had relatively high reliability coefficients (above 0.6), while the average inter-item correlation coefficients also implied internal consistency for all factors. Moreover, all items loaded on a factor with a loading greater than 0.3; the relatively high factor loadings indicated a reasonably high correlation between the factors and their component items. The Kaiser-Meyer-Olkin measures of sampling adequacy of 0.84, 0.90, 0.77 and 0.82, respectively, also indicated that patterns of correlation were relatively compact and yielded distinct and reliable factors (Field 2005). Bartlett's test of sphericity also reached statistical significance (p < 0.001), supporting the factorability of the correlation matrix (Pallant 2007). Factor scores were calculated as the average of all items contributing to a particular factor to interpret them on the original five-point Likert scales.
As indicated in Table 1, socialisation (3.68) was regarded as the most important motive for dining out, followed by gastronomy enjoyment (3.65), lifestyle and escape (3.51), while status (3.34) was regarded as the least important factor (although still important according to the Likert scale rating). the tipping decision-making factors, waiter competence and service (3.45) was the main factor that influenced tipping decision making (Table 2). This was followed by mood and ambience (3.09) while ethnicity and restaurant setting (2.98) was the least influential factor.  Table 3, social acceptability (3.62) was the primary reason for tipping, followed by server support (3.29), while bad service (3.70) followed by waiter incompetence (3.53) were the main reasons for not tipping. Identifi cation of segmented clusters 1An exploratory cluster analysis was performed on the motivational factors using Statistica (StatSoft, Inc. 2015). A hierarchical cluster analysis, using Ward's method of Euclidean distances, was used to determine the clusters' structures based on the motivation factors. A three-cluster solution was selected as the most discriminatory (see Figure 1). The results of the multivariate analyses were used to identify the three segments and to indicate any significant differences between them (p < 0.05).  Table 4, the ANOVAs indicate that all motivational factors contributed to differentiating between the three segments (p < 0.05) with large effect size differences. Segment 1, the largest, with 167 respondents, was labelled Occasionalists because they singled out certain motives indicating gastronomy enjoyment and socialisation as their main motives for dining out. Segment 2, the second largest, with 65 respondents, was labelled Socialisers because they were mainly motivated by socialisation and had the lowest mean values across the other three factors. Finally segment 3, the smallest, with only 24 respondents, had the highest mean values for all four motives and was labelled Indulgers, as they are particularly motivated by gastronomy enjoyment.

As, Tukey's post hoc multiple comparisons and effect sizes results
ANOVAs and Tukey's post hoc multiple comparisons and effect sizes were used to determine the differences between the respondents according to their sociodemographic, behavioural characteristics as well as dining preferences. According to Table 5, the three segments of black South African diners recorded statistically significant differences in terms of age (p = 0.055, although Tukey's posthoc tests indicated a significant practical difference), the percentage they regard as appropriate to tip (p = 0.021), spending on food and restaurants (p = 0.022) as well as based on the tipping decision-making factors (p < 0.05), reasons for tipping (p < 0.05) and reasons for not tipping (p < 0.05). Regarding age, the Socialisers were the youngest (average age of 37 years), while the Indulgers were the oldest (average age of 43 years). Furthermore, the Indulgers indicated a higher percentage (13.42%) as being appropriate to tip compared to the Socialisers who indicated a lower percentage (9.75%). All three segments differed significantly in terms of the tipping behaviour factors. Mood and ambience influenced the Indulgers more when deciding to tip (3.27) compared to the Socialisers (2.76), while waiter competence and service as well as ethnicity and restaurant setting were more important to the Indulgers and Occasionalists (3.90 and 3.54 and 3.19 and 3.11, respectively). Compared to the other two segments, Indulgers regarded both server support (3.62) and especially social acceptability (4.20) as important reasons to tip, while bad service (4.34) was a stronger reason for this segment not to tip compared to the other two.

Findings and implications
1The results of this research were the following findings and implications that could be useful for researchers, restaurateurs as well as servers: Firstly, the results confirm that different cultures have distinct motives for dining out, thereby confirming the research by Mkono (2011). The motives identified also correspond with the general motives for dining out as discussed in the literature review. A comparison with the findings of Saayman (2014) and Saayman and Saayman's (2015) studies among white South Africans showed that socialisation and gastronomy enjoyment are the main reasons for dining out, while status is less important. This research, however, revealed that over and above socialisation and gastronomy enjoyment, an additional motive exists, namely lifestyle and escape. This implies that although some deem dining out to be a luxury, our respondents indicated that it may form part of their lifestyle and is a way to escape their everyday environment. This result also confirms a shift in the leisure behaviour of black South Africans, which includes dining out (Kruger & Saayman 2014). The implication for restaurateurs is that restaurants should focus primarily on creating an atmosphere or ambience where people can socialise and get away from their daily routine, while experiencing good food and service.
Secondly, based on the identified motives, three distinct segments of diners were identified: Occasionalists, Socialists and Indulgers. The results therefore provide an OSI typology of black South African diners. These segments differ from the dining segments identified in previous research with the exception of the Socialists who correspond with the Social Supper segment as identified by Josiam et al. (2015). From a methodological point of view, this research thus confirms that diners' motivational factors are a useful segmentation base, thereby challenging the use of traditional sociodemographic segmentation bases. This approach proves to be effective, especially for restaurateurs and researchers wishing to understand a particular ethnic group's dining and tipping behaviour.
Thirdly, the results indicated that black South African diners do indeed tip, as 42% of the respondents indicated that they always tip and 30% sometimes tip. All three segments also tip in line of the 10% required in South Africa, indicating that black South African diners are aware of the tipping norm in the country. This finding therefore contradicts the general notion that black diners are bad tippers. This finding implies that these diners in South Africa cannot be regarded as a nontipping market. Servers should therefore be encouraged to deliver excellent service and favourable dining experiences to this dining market since the results showed that quality service is a prerequisite for tipping.
Fourthly, the results also confirm that black South African diners should not be regarded as homogeneous in terms of their dining and tipping behaviour. Results have shown that there are more behavioural variables that influence the tipping decision than sociodemographic variables. The three identified segments differed significantly from each other, based on their age, motives for dining out and especially in terms of their tipping behaviour. The influence of age, the only sociodemographic difference in the present study, has not been identified in previous research. It does, however, reveal an interesting finding regarding the progression of dining experiences. Socialisers were the youngest, followed by Occasionalists and Indulgers who were the oldest. Indulgers have the most diverse motives for dining out and the most evolved tipping behaviour in terms of how much they tip (tipping more than the customary 10%) as well as their reasons for tipping/not tipping. This cluster's greater age and experience could play a significant role in this behaviourpossibly having developed a more distinct palate for different food and flavours over time, the enjoyment of food, therefore, becomes the primary reason for dining, while the socialisation and escape reasons become secondary.
Fifthly, the research confirms the findings by Azar (2010) (2015), Saayman and Saayman (2015), and Saunders and Lynn (2010), namely that a combination of factors plays a role in black South Africans' tipping behaviour. Based on the results, all three tipping behaviour categories (the customer, the server/waiter and external factors) are relevant. The particular combination of factors found in this research, however, has not been identified in previous studies. These factors can therefore be regarded as distinct and especially influential in decisions to tip by black South African diners, thereby once again supporting the notion that tipping behaviour differs from culture to culture. The following three factors were identified, in order of importance: waiter competence and service, mood and ambience and ethnicity and restaurant setting. Lynn and McCall (2000) also identified mood as having a positive relationship with tipping, while restaurant setting and location were also identified by Garrity and Degelman (1990), Rind and Strohmetz (2001) and Lynn and Thomas-Heybert (2003). Ethnicity was also identified by Harris (1995), Lynn (2004) and Lynn et al. (2008). However, contradicting the Lynn et al.'s (2008) findings, there is no evidence that black South Africans tend to tip white waiters more, and this factor did not seem to play a role in the present study. In fact this research demonstrated the opposite. The implications of this finding are twofold: (1) restaurateurs should train waiters properly on the aspects of service delivery and hospitality since service delivery not only affects the size of the tip but also impacts on the main motives of diners, namely gastronomy enjoyment as well as the socialisation aspect and experience of diners, and (2) restaurants should pay careful attention to creating an inviting atmosphere and distinct setting as this greatly influences tipping decisions by black South Africans.
Sixthly, social acceptability is the primary reason for tipping, which is supported by Azar (2004Azar ( , 2005, Boyes et al. (2004) and Saayman and Saayman (2015), who found that there is pressure to conform socially for approval and status, while Azar (2004) and Lynn (2006Lynn ( , 2015 revealed it is a sense of duty or obligation, it makes one feel positive and because people support the practice of tipping. This was followed by server support, which included statements that it contributes to waiters' income and gave support to the rule of tipping. According to Lynn and Grassman (1990), Speer (1997) and Saayman and Saayman (2015), tipping is for financial reasons. Restaurateurs and waiters could use this finding to their advantage. Since patrons feel a moral obligation to tip, delivering excellent service and food would almost certainly result in larger bills and higher tips. Restaurateurs could also consider including a service charge on the bill, but this consideration would require more research.
Seventhly, confirming the findings of Saayman (2014) and Saayman and Saayman (2015), bad service is regarded as the main reason for not tipping, followed by waiter incompetence, which is a distinct factor identified in the present study. The implication of this finding is of vital importance for restaurants and the broader tourism industry in general, since it emphasises the need for training, which was also highlighted by the fact that diners are willing to pay larger tips if the service is good. Since most servers could be perceived as the "face" of a restaurant, they should be trained in a variety of functions, such as having a knowledge of wine, how to approach customers and how to be polite. Even though most servers are not permanently employed in the South African restaurant context -since most of them are students, scholars or young people who are still working towards a career, not necessarily in the hospitality industry -restaurants should still devote time to developing servers' skills and knowledge.
Lastly, all three clusters differed significantly in terms of the tipping behaviour factors as mood and ambience influenced Indulgers more, compared to Socialisers, while waiter competence and service as well as ethnicity and restaurant setting were more important to Indulgers and Occasionalists. Indulgers furthermore regarded both server support and especially social acceptability as the main reasons to tip, while bad service was a stronger reason for this cluster not to tip, compared to the other two. These results also indicate a progression in the particular factors, changing from more general to specific, that influence each diners' tipping behaviour. More research would thus be needed to validate this finding and the progression from Socialisers to Occasionalists, and vice versa, and ultimately Indulgers.
Conclusion 1The purpose of this research was to gain greater insight into the dining and tipping behaviour of black South Africans. This was achieved by segmenting this dining market on the basis of their motives for dining out to distinguishing different markets and their consequent tipping behaviour. The results revealed a number of interesting findings. Firstly, socialisation and gastronomy enjoyment were the main reasons for dining out, while status was less important, but more important compared to white South Africans' dining behaviour. This research, however, revealed an additional motive, namely lifestyle and escape. Secondly, by segmenting diners on the basis of these motives, the results clearly indicated that there are different types of black diners and that one can distinguish between them, namely Occasionalists, Socialisers and Indulgers. An OSI typology for black diners is thus proposed. Thirdly, black diners do tip regularly and also tip within the 10% tipping norm in the country, proving that general stereotypes surrounding this dining market are ungrounded in South Africa. The research furthermore confirmed that tipping behaviour is determined by the following three key components: the customer (mood), the server (waiter competence and ethnicity) and external factors (restaurant setting and ambience). The reasons why people do not tip were captured in two factors: bad service and waiter incompetence, as opposed to the reason why they do tip, that is, social acceptability and server support. Over and above the fact that the tipping decision of the different dining segments is influenced by several factors, the research has also clear implications for both the server and restaurateurs. One element that influences good or quality service and which has also been acknowledged by this research is the role that training could play in providing good services, which in turn, would impact on the size of the tip. The research should also make a contribution in terms of identifying the determinants of tipping in general, indicating that most of those determinants are behavioural. It furthermore sheds light on the dining and tipping behaviour of black diners in a developing country, thereby filling the gap in the current literature. The results confirm that while it is important to consider cultural differences, compared to US studies, these differences seem to be unjustified when it comes to dining out and tipping in South Africa.
Limitation and future research 1This study had some limitations in that it analysed potential black diners at a festival that mainly attracts high-income festivalgoers. The researchers were thus unable to generalise on the basis of this study alone. Future and ongoing research could fill this gap. It is therefore recommended that similar research be conducted, but on a larger scale in order to target black diners from across the country, and include all nine major ethnic groups so as to compare results and validate the proposed typology as well as dining and tipping behaviour. The results of this research should be regarded as a stepping stone towards a much larger study on the dining and tipping behaviour of black South Africans.