The role of celebrity involvement in influencing domestic tourists’ intention to visit Tanzania’s attractions: attitude as a mediating variable

Abstract

Purpose

Although previous studies have examined the influence of celebrity involvement in behavioural intentions, the role of celebrity dimensions such as attraction, self-expression and centrality in influencing tourists’ intention in the context of developing countries such as Tanzania remains largely unaddressed. This study, therefore, examined the relationship between celebrity involvement and domestic tourists' intentions to visit tourist attractions, attitude being the mediating variable.

Design/methodology/approach

A questionnaire was self-administered on a convenient sample of 279 domestic tourists in the Tanzania’s four largest regions, namely, Dar es Salaam, Mbeya, Arusha and Mwanza. Employing a quantitative research approach, structural equation modelling was performed to test the cause-and-effect relationships between celebrity involvement and tourists’ intentions before testing the mediating role of attitude in such a relationship. Confirmatory factor analysis was also performed to test the measurement models.

Findings

Attraction emerged to be the main determinant of the celebrity dimension that significantly influenced domestic tourists’ travel intentions, whereas attitude partially mediates such a relationship. Moreover, Bongo Fleva musicians, particularly Diamond Platnumz, one of the leading celebrities in this genre, were found to influence most of the respondents’ travel intentions – he posted a picture on his Instagram account of him touring the Serengeti National Park.

Research limitations/implications

The study focused on domestic tourists residing in four of the Mainland Tanzania’s largest regions, hence excluding those residing on the islands of Unguja and Pemba. Due to cultural differences, including the islands not only could unleash new perspectives on celebrity involvement dimensions but also could have introduced new determinants of travel intentions.

Practical implications

This study offers guidance to tourism businesses on designing their marketing campaigns that they should harness celebrity’s attractive qualities effectively. The focus should be directed not only towards linking destinations with celebrities but also on stimulating positive perception of those destinations, aligning with the attitudes of their followers.

Social implications

The study has set out a new perspective for researchers, practitioners and tourism businesses to refine their promotional strategies and for academicians to gain a deeper understanding of visitor behavioural intention dynamics.

Originality/value

This study has proposed and verified that attraction is a dominant determinant compared to self-expression and centrality in explaining tourists’ travel intentions and attitudes, which play a significant role in explaining such a relationship. Although the study employed a modified theory of planned behaviour in a celebrity involvement study, the findings have broadened the understanding and its applicability in the context of a developing country.

Citation

Kara, N.S. (2024), "The role of celebrity involvement in influencing domestic tourists’ intention to visit Tanzania’s attractions: attitude as a mediating variable", International Hospitality Review, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/IHR-05-2023-0031

1. Introduction

Tourism is a business in which customers make several choices before travelling. Therefore, in a world of diverse attractions, tourism firms must aim to distinguish one product from another through advertisement. To achieve this, celebrity endorsement is commonly used in promoting different types of products or services (Yuan, Wang, Liu, & Ma, 2023). Celebrity endorsement in promoting products has been one of the crucial research areas in marketing (Wang & Liu, 2023). It is an agreement between publicly known individual and an entity (e.g. brand for a purpose of promoting the brand (Liu & Liu, 2020). In simple terms, a celebrity refers to an individual well known in the public domain who is using his/her fame to promote various brands (Ugwu & Okeke, 2023). The current study adopts Ugwu and Okeke’s (2023) definition and describes Tanzanian athlete, beauty queen, Bongo movie star, TV personality and Bongo Fleva musician who use Instagram as celebrities. Arguably, the celebrity’s use of Instagram can help boost various tourist attractions available in Tanzania.

Recently, issues of social media influencers have been receiving increasing attention (Masuda, Han, & Lee, 2022). The increased use of social media platforms has inevitably resulted in an increase in social media influencers and celebrities. Subsequently, these celebrities have become so famous in the eyes of their followers and fans that they are contracted to endorse various brands. In this regard, the rapid use of the Internet and social media has accelerated the growth of such celebrities (Hu, bin Sidek, Abd Rahman, Yusof, & Chen, 2023). Popular social media platforms such as YouTube, Instagram, Facebook and X (formerly known as Twitter) have changed the way communication is conducted between suppliers and customers. Celebrities are now hired to use their social networks to endorse various brands. Implicitly, they have become a key marketing tool for influencing consumers’ purchase decisions.

Inevitably, marketers have been forced to employ celebrities to attract their fan base. After all, fans can easily relate to the brand endorsed by their favourite celebrities than those endorsed by non-celebrities (Phua, Jin, & Hahm, 2018). In this era, celebrities have become powerful marketing agents (Wang & Liu, 2023); they are public figures and image promoters (Andrews & Jackson, 2001). As a result, consumers prefer them over non-celebrities. Since celebrities are more widely known on social media platforms than non-celebrities (Su, Li, Zhang, & Wang, 2023), customers tend to associate their attractiveness and trustworthiness with the brands they endorsed (Silvera & Austad, 2004). In fact, celebrities have power to make advertisements believable and in brand recognition (Min, Chang, Jai, & Ziegler, 2019). Moreover, celebrities capture customers’ attention more easily than non-celebrities (Agrawal & Kamakura, 1995). For this reason, celebrities have become key icons for changing consumers’ attitude towards brands (Mittal, 2021).

The contribution of celebrities to branding and income boosting for firms should not be underestimated in tourism industry. Celebrity endorsements in destinations constitute a viable tool for boosting destination earnings and fame (Kotler, Haider, & Rein, 1993). In the Global North, celebrity endorsement has been used extensively (Al Mamun, Naznen, Yang, Ali, & Hashim, 2023; Jia, 2023). Celebrities are contracted to showcase destinations’ beauty to travellers (Chan, Lee, & Wong, 2018), which boosts earnings. In Australia, the tourism promotion branded by Chris Hemsworth, an actor, raised more than USD195 million in 2019 (Salemme, 2019). Such positive contribution has elevated the acceptance of celebrity endorsement to the tourism industry (Atkin & Block, 1983; Petty, Cacioppo, & Schumann, 1983).

In the Global South, particularly in Africa, the use of celebrity to promote tourism is emerging. Among other countries, Tanzania has also started engaging celebrities in promoting its tourism destinations. Sports personalities (e.g. Mbwana Samata), musicians (e.g. Diamond Platinumz), beauty queens (e.g. Witness Teddy) and political ambassadors (e.g. Julius Yulin Liu) have been engaged in promoting the Tanzanian destinations. For instance, Witness Teddy and eminent TV and radio presenters were used by the Tanzania National Park Authority (TANAPA) when launching a campaign during Christmas and New Year festival called “Tourism and TANAPA”. These celebrities were invited to visit various tourist destinations in the country (Mwachweo Communication News, 2019). Furthermore, many Tanzania celebrities, including movie stars, musicians and beauty queens, were also invited to attend Kili Challenge event. The event aimed at encouraging local people to climb Mount Kilimanjaro, and it raised 172 million Tanzania shillings (Jamhuri wa Muungano wa Tanzania, 2021).

Recently, a presidential documentary dubbed “The Royal Tour” features the president of the United Republic of Tanzania, Honourable President Samia Suluhu Hassan, visiting different tourist attractions. She promoted the existing tourist attractions to the outside world (The Citizen, 2021). The Royal Tour documentary premiered on 18th April 2022 in New York. Despite the government’s efforts to employ celebrities to promote tourist attractions, information on the role of celebrities in influencing travel intention in Tanzania is unclear.

The existing studies have focused mainly on products and services, whereas those that are focused on tourism destinations have remained largely under-researched (van der Veen, 2008). Few empirical works on tourism destinations have established the link between celebrity involvement and travel behaviour (Lee, Scott, & Kim, 2008; Van der Veen & Song, 2014; Yen & Croy, 2016; Yen & Teng, 2015). Some studies have reported a positive relationship between celebrity involvement and behavioural intentions (Lee et al., 2008; Salamoura, Giannopoulos, & Moumouri, 2020; Yen & Teng, 2015). In contrast, others have reported that celebrity involvement did not always lead to behavioural intentions (Ohanian, 1991; Stafford, Stafford, & Day, 2002). These conflicting findings regarding the roles of celebrity in influencing travel intention inspired calls for further assessments.

The foundation of this study is based upon the theory of planned behaviour (TPB), which originates from social psychology. Despite its origin, the theory has been extensively employed in tourism studies to understand visitors' travel behaviour (Kaplan, Manca, Nielsen, & Prato, 2015; Yuzhanin & Fisher, 2016). However, the use of the theory in illuminating its role in explaining the influence of celebrity involvement on travel intentions among domestic travellers in the context of Tanzania needs further inquiry. Since Tanzania is endowed with multiple tourist attractions ranging from wildlife, mountains, beaches and cultural diversity, a study to assess how celebrities can be used to attract tourists to different destinations is timely. In addition, the mediation tests in this study offer details on how celebrity involvement affects tourists’ travel intentions.

2. Literature review

2.1 Theory of planned behaviour (TPB)

Ajzen’s (1991) theory posits that individual behaviour is influenced by people’s behavioural intention to pursue a particular behaviour, making the latter an immediate predictor of one’s action. Ajzen and Fishbein (1980) defined intention as the likelihood of an individual wishing to do, pursue something or self-predict or expect what an individual would do. Behavioural intention, a key variable in the model, is explained jointly by an individual’s attitude, perceived behavioural control and subjective norms. Thus, the development of beliefs regarding a behaviour’s possible outcomes and the assessment of these outcomes can translate into an overall positive or negative evaluation of the behaviour (attitude) capable of influencing behavioural intention (Fishbein & Ajzen, 2011). Attitude is an important variable under TPB defined as a positive or negative psychological emotion that emerges when an individual performs certain behaviour (Eagly & Chaiken, 1993). Impliedly, when individuals have a positive attitude, they tend to behave positively, and vice versa. The subjective norm is also an important variable in the model since it includes individuals’ normative beliefs regarding how others perceive, judge or endorse his or her pursuit of a certain behaviour. This variable is well explained by the influence of social pressure (friends or family members) regarding an individual’s performance of certain behaviour. Such subjective norms refer to the perceptions of other people intending to perform certain behaviour and might not reflect the reality of what others think (Yuzhanin & Fisher, 2016).

Perceived behavioural control focuses on how easy or difficult it is to perform a certain behaviour. The TPB assumes that perceived behavioural control influences an individual’s actual behaviour indirectly through their intention. This theory assumes that the individuals believing to have limited control over performing it because of factors beyond their ability, would be limited to perform it, even if the attitudes and subjective norms are positive (Madden, Ellen, & Ajzen, 1992). Factors that limit an individual from performing certain behaviours include lack of funds, time and limited opportunities.

Although in the TPB behavioural intention is influenced by attitude, subjective norms and perceived behavioural control, this study only borrowed the attitude and behavioural intention constructs from it and used them together with the celebrity dimension items (attraction, centrality and self-expression) to build its model. Therefore, in addition to assessing the role of celebrity involvement in influencing individual domestic travel intention, the study factored in the role of friends or family (subjective norms), and the perception of the difficulty inherent in undertaking a certain behaviour (perceived behavioural control). Moreover, the actual behaviour was excluded from the model since the study was narrowed down to predicting travel intentions. This study is not the first to apply some of TPB constructs in behavioural intention studies. Cuomo, Foroudi, Tortora, Hussain, and Melewar (2019) employed some TPB constructs when examining the effect of celebrity endorsement (credibility, familiarity, luxury brand values and brand sustainability awareness) on attitudes towards purchase intentions.

Despite the TPB being extensively used to determine individual behaviour, it is not devoid of criticisms. It serves as a causal model for predicting individual behaviour, but one’s attitude partially influences behavioural intention. Moreover, the TPB mainly focuses on rational reasoning while ignoring the roles that an individual’s sub-consciousness and feelings play (Sniehotta, Presseau, & Araújo-Soares, 2014). It also ignores other crucial factors such as unconscious motives and spontaneous moments. Furthermore, the TPB also ignores the role of individual motivation and past behaviour in explaining the actual behaviour (Lam & Hsu, 2004). On top of that, factors such as lack of financial resources and limited opportunities may affect one’s attitude, hence ending up affecting individual behaviour. Even though there is conflicting finding regarding attitude as a predictor of behaviour, past tourism studies have acknowledged that attitude is a key factor in predicting travel behaviour (Pike, 2006; Um & Crompton, 1991). Empirical tests further attested that attitude to destination influences travellers' preferences and intentions (Court & Lupton, 1997; Milman & Pizam, 1995).

Regardless of criticisms raised on TPB, the theory has widely been applied in tourism studies to assess behavioural intentions. Some of the studies include the work of Huang, Chang and Backman (2019), who applied TPB to understand tourists' revisit intentions in the context of creative tourism. Yet another study by Tara, Albattat, and Jacquline (2020) also employed it to identify the factors that affect customers’ intention to stay in branded luxurious hotels in Malaysia. Apart from these studies, other researchers have also utilized the TPB to investigate residents’ perceptions of tourism and their intention to support tourism developments (Wong, Wu, & Kim, 2022). Others have addressed the role of cultural worldview in predicting heritage tourists' behavioural intention to visit Hanok Village in South Korea (Lee et al., 2021). Furthermore, Moorthy et al. (2021) employed the theory to examine social media usage among Malaysian millennials in planning for their domestic holidays. Thus, such a widespread application of the TPB in behavioural intention studies further validates its relevance in assessing behavioural intention in tourism studies.

2.2 Celebrity involvement

The celebrity involvement concept originated from leisure involvement (Lee et al., 2008). Celebrity involvement refers to the process of including an unobservable state of motivation, arousal or interest in leisure activities (Havitz & Dimanche, 1997). Most of the celebrity involvement studies have measured the construct using three main dimensions (attraction, centrality and self-expression) as described by McIntyre and Pigram (1992). The attraction combines pleasure, enjoyment and satisfaction, which are evident in the perceived importance of pleasure/hedonic activities. Centrality handles the significance of the activity within one’s overall life (i.e. expression of individual lifestyle), whereas self-expression covers one’s self-identity and social identity about the activity he/she undertakes (i.e. how leisure activity provides an opportunity for an individual to showcase the desired image). In this study, celebrity involvement was measured using the three dimensions of attraction, centrality and self-expression borrowed from Lee et al. (2008).

2.3 Celebrity involvement and travel intention

In tourism, celebrities have played a crucial role in influencing the tourists’ destination choice process and travel behaviour (Zhu, Hoc Nang Fong, Gao, Buhalis, & Shang, 2022). The trust that people have on celebrities has a positive effect on each phase of their travel decision (Pop, Săplăcan, Dabija, & Alt, 2022). Celebrity involvement in terms of attractiveness, self-reflection and centrality reportedly has a significant influence on destination image, place attachment and tourist loyalty (Zhou et al., 2023). Furthermore, celebrity involvement can lead to perceived destination authenticities, hence influencing travel intentions (Glover, 2009; Lee, Busser, & Yang, 2015; Zhu et al., 2022). The effectiveness of celebrity involvement in promoting brands is reflected by the level of fans' personal attachment (Zhu et al., 2022). It is further reported that celebrity attachment leads to a positive sense of presence, which positively influences tourist enjoyment, and, in the end, it predicts tourists' travel intention. Based on this information, it could be hypothesized:

H1a.

Celebrity attraction (ATR) significantly influences tourists' travel intentions (TI).

H1b.

Self-expression (SE) significantly influences tourists' travel intentions (TI).

H1c.

Centrality (C) significantly influences tourists' travel intentions (TI).

2.4 Celebrity involvement and attitude towards the destination

In leisure studies, celebrities have helped to promote destination images positively to their fans (Teng & Chen, 2020). The celebrities’ favourable qualities (e.g. attractiveness, expertise and trustworthiness) influence tourists’ attitudes and motivate their intentions to visit the destinations they endorsed (Teng & Chen, 2020). Since celebrities are ambassadors of the destinations, they can influence the audience’s attitudes and behavioural intentions based on the congruence between celebrities and destinations (Wu & Lai, 2023). Besides, fans trust the credibility of their celebrities (Teng, Su, Liao, & Wei, 2020). They are easily persuadable, which allows them to develop the intention of visiting the endorsed destination. However, if a fan’s attitude to a celebrity is negative, that fan can develop a negative attitude towards the endorsed destination, and vice versa. Wang, Kao, and Ngamsiriudom (2017) examined the causal relationship between celebrity credibility, brand attitude and purchase intention using the air transportation sector in Taiwan. Impliedly, once customers perceive an endorsed product to be trustworthy, desirable and respectable, their attitude towards the celebrity who endorsed the brand becomes positive. In the end, it may lead to purchase intention. Since a celebrity endorser serves as an antecedent of brand attitude (Spears & Singh, 2004), it could be hypothesized that:

H2a.

Celebrity attraction (ATR) significantly influences tourists’ attitude towards a destination.

H2b.

Self-expression (SE) significantly influences tourists’ attitude-towards a destination.

H2c.

Centrality (C) significantly influences tourists’ attitude towards a destination.

2.5 Mediating effect of attitude

Attitude is “a psychological tendency expressed by evaluating a particular entity with some degree of favour or disfavour” (Eagly & Chaiken, 1993, p. 1). Since attitude is one of the most important constructs in explaining one’s behaviour, various consumer-based studies found attitude to mediate the effects of attitude to a brand (Biehal, Stephens, & Curio, 1992). The foundation for understanding the role of attitude in predicting one’s behaviour can be traced back to the theory of reasoned action (TRA) and the TPB. The two theories suggest that behavioural intention is mediated by one’s attitudes. This implies that attitude is a key mediator of behavioural intentions. This factor prompts an individual to behave in a certain way, as reported in tourism behaviour (Lee, 2007; Sparks, 2007). Also, attitude is a crucial predictor of tourist participation, satisfaction (Ragheb & Tate, 1993) and external behaviours (Ajzen, 1991; Lee, 2007) in studies that involve celebrities.

The match between a celebrity and the endorsed brand is important in influencing an individual’s behavioural intentions. Without connections between celebrity endorsers and the brand, a negative attitude can result. For example, customers could treat a celebrity endorsing the brand as someone just paid to promote the brand. On the other hand, celebrities with connections to the brands they endorse are perceived as experts by their target customers (Ohanian, 1991). As such, we hypothesize that:

H3.

Attitude towards a destination mediates the relationship between the celebrities’ involvement dimensions and the tourists’ travel intentions.

3. Methods

3.1 Sample size and sampling strategies

Data for this study were collected between October and December 2021. The study targeted Instagram users mostly in the age between 25 and 34 years, who intended to visit Tanzania’s tourist attractions after seeing pictures posted by local celebrities on their Instagram accounts. Data for early 2022 show that Tanzania had 3.15 million Instagram users (Simon, 2022), aged between 25 and 34 (NapoleonCat, 2020). This age group is consistent with global data on Instagram users of 2022 who were similarly aged between 25 and 34 years (Woosuite, 2023). Since the actual data of Instagram users in specific regions in Tanzania are not known, it was difficult to adopt a random sampling design, hence the use of convenience sampling. Moreover, the tourists had to meet the following additional criteria to qualify for participation. (i) They must be residents of Tanzania, have Instagram accounts and follow local celebrities such as Bongo Fleva musicians (such as Diamond, Vanessa Mdee and Sheta), Bongo movie stars (such as Ray Kigosi and Irene Uwoya), TV personalities (like Millard Ayo), sports personalities (like Mbwana Samata) and beauty queens (such as Flaviana Matata, Nancy Sumari). (ii) They intend to visit Tanzania tourist attractions after seeing posts by local celebrities of their visits to various tourist places on their Instagram accounts. (iii) They stay in Dar es Salaam, Mwanza, Mbeya and Arusha during the data collection. Bongo Fleva musicians refer to all local musicians who sing songs with beats fusing Tanzanian traditional tunes with modern music compositions (i.e. “Bongo Fleva”). The movie stars were drawn from a pool of all local actors and actresses in Tanzania. The involvement of the biggest four cities in the country also helped to ensure that the sample represented the study population and reduce the non-random convenience sampling effect. Therefore, 300 survey instruments were initially distributed conveniently and physically to domestic tourists during the day (i.e. lunch break) and collected after working hours. The data cleaning process, which preceded the data analysis, excluded 21 questionnaires for having missing information, leaving a sample 279 usable questionnaires, with a response rate of 93%. This sample size qualified for meaningful analysis with the structural equation modelling (SEM) for being higher than the minimum sample of 200 recommended for it (Byrne, 2001).

3.2 Pilot study and data collection

A pilot study was done to ensure the instrument measured what it was intended to do. The survey was piloted conveniently on five academic staff teaching tourism at The Open University of Tanzania. Later, the survey instrument was piloted on 50 respondents in Dar es Salaam to evaluate its appropriateness. Several items were corrected after the pilot study, including adjusting some of the wordings. The piloting informed the final self-administered questionnaire applied during data collection in the field.

3.3 Measurements

A leisure involvement scale borrowed from Lee et al. (2008) was used to measure celebrity involvement. The scale has three constructs (attraction, centrality to lifestyle and self-expression). Each construct had 4 measurement items (12 items). Respondents were asked to indicate their agreement with the statements provided on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Higher scores implied a greater degree of the celebrity’s’ involvement based on Lee et al. (2008).

Attitude towards a destination was measured by asking respondents, “How would you describe your overall attitude towards Tanzania?” A 7-point semantic differential scale ranges from good-bad, dislike-like, pleasant-unpleasant, positive-negative and favourable-unfavourable, according to MacKenzie and Lutz (1989).

The travel intention construct was measured using 3 items adopted from Lam and Hsu (2006) and Hsu, Kang, and Lam (2006). Respondents were asked to indicate their travel intention by rating the given statements presented on a 7-point Likert scale ranging from 1 (very unlikely) to 7 (very likely).

Those who indicated celebrity influence further named the endorsed attraction(s) they intended to visit. Additionally, the respondents determined the celebrity category (athletes, Bongo Fleva musicians, Bongo movie stars, beauty queens and TV personalities) that influenced their travel intentions to various attractions in the country. Table 1 presents the results of the measurement items under each construct.

4. Results

4.1 Data normality

Skewness and kurtosis of the data were assessed to determine data normality. Sposito, Hand, and Skarpness (1983) suggested that skewness and kurtosis values ranging between −3 and +3 indicate normal distribution of the data. The value of skewness and kurtosis in this study fell within this recommended range. Table 2 attests to data normality.

4.2 Respondents’ profiles

Descriptive statistics such as frequencies and percentages were used to analyse respondents' demographic information. The results (Table 3) indicate that there were more male respondents (53.8%) than female respondents (46.2%). Most of the respondents were single (77.1%) and were aged 25–26 (52%), followed by those aged 27–28 (19.7%). Students (45.9%) constituted the largest segment of the sample and had a university education (66.7%). A good number of respondents earned less than 300,000 Tanzanian shillings per month. In terms of residency, most of the respondents used in the study were from Dar es Salaam. The data imply that most domestic tourists were predominantly young, educated, single and had an average monthly income.

4.3 Reliability and validity results

Cronbach’s alpha was used to test for the internal consistency of the research instrument. It is reported that any item with a Cronbach’s alpha of 0.65–0.80 is considered reliable (Vaske, Beaman, & Sponarski, 2017). Internal consistency of the instrument was proved as the Cronbach’s alpha values for the constructs ranged between 0.832 and 0.927, which were above the acceptable minimum cut-off point of 0.65–0.80 as suggested by Vaske et al. (2017). On the other hand, convergent validity was examined by using the average variance extracted (AVE). The results (Table 4) indicate that the values of AVE for the entire constructs were above the cut-off point of 0.5, as indicated by Fornell and Larcker (1981). Furthermore, the values of the standardized regression weights fell between 0.72 and 0.93, which is above the accepted cut-off point of 0.5. This proves that convergent validity was attained. Furthermore, discriminant validity was also examined using an approach developed by Fornell and Larcker (1981). The shared variance (SV), the squared correlation between each pair of constructs, was compared with AVE. The results in Table 4 indicated that the value of AVE for each construct fell between 0.611 and 0.769, which was greater than the acceptable level of 0.5 (Fornell & Larcker, 1981). The results also indicate that all the values of SV were lower than AVE values, hence supporting discriminant validity. The results further indicate that the composite reliability (CR) ranged from 0.835 to 0.930, which is greater than the minimum acceptable value of 0.6 (Hair, Black, Babin, Anderson, & Tatham, 2006).

4.4 Celebrity influence

In this study, descriptive statistics was used to identify a category that could influence domestic travel intentions – 42% of the respondents cited being influenced by celebrities, followed by those mentioning friends (33.3%) and families (25%). This implies that most domestic tourists' travel intentions are influenced by famous people rather than people who are in their regular social cycles. Table 5 presents group and celebrity categories that influence visitors' travel intentions.

4.5 Category of celebrities influenced visitors’ travel intentions

In this study, respondents were also asked to mention celebrity categories that would influence their travel intention. In all, 52% mentioned that Bongo flavour musicians such as Diamond would influence their travel intention. This was followed by TV personalities (15.1%) and athletes (14%), and very few (8.6%) are influenced by beauty queens such as Wema Sepetu, Flaviana Matata and Nancy Sumari. This implies that it would be profitable to hire Bongo flavour artists to be countries’ tourist ambassadors because they have high persuading power to influence visitors’ travel intentions significantly than other celebrities including TV personalities, athletes and beauty queens. Table 5 summarizes the results on the celebrity category that influenced travel intentions.

4.6 Name of celebrity and the type of attraction that would influence visitors' travel intentions

This study identified the name of the celebrity that influenced tourists' travel intentions the most. The results indicated that most of the respondents (10%) mentioned that they would be willing to travel when Bongo Fleva singer Diamond Platnumz posted a picture of the Serengeti National Park. Other influential Bongo Fleva artists included Nandy (8.6%) when she posted a picture of Ngorongoro National Park; Harmonize (7.9%) when he posted a picture of Mkomazi National Park, Zuchu (6.5%) when she posted a picture from the Zanzibar archipelago; and Ali Kiba (5.7%) when he posted a picture of Mikumi National Park. This implies that, although celebrities play a crucial role in influencing tourists' travel intention to various national parks, each celebrity influences visitors’ differently. Powerful celebrities such as Bongo flavour artists can be hired to endorse unknown tourist attractions to the outside world. Their persuasion power might attract a significant number of visitors from different parts of the world. Appendix presents a summary of the names of the celebrities and the attractions they posted that could influence visitors’ travel intentions.

4.7 Structural equation modelling (SEM)

A two-stage approach developed by Anderson and Gerbing (1988) informed the analysis of the hypothesized relationships. IBM Amos and SPSS Statistics software Version 23 also aided the analysis. The goodness-of-fit indices are divided into three parts: absolute, relative and parsimonious. Hair, Anderson, Tatham, and Black (2007) observe that at least one of the fit indices from each category should be reported to evaluate the overall model fit. Therefore, the study used goodness-of-fit indices such as GFI (>0.90), RMSEA (<0.10), CFI (>0.95), NFI (>0.90), PNFI (>0.50) and PCFI (>0.50) to evaluate the modal fit. The accepted cut-off point for each index is indicated in the brackets based on Meyers, Gamst, and Guarino’s (2006) recommendations.

4.7.1 Measurement model

Confirmatory factor analysis (CFA) using Anderson and Gerbing’s (1988) approach was carried out to determine the uni-dimensionality of each construct before establishing causal relations. The results indicated that all the standardized parameters were statistically significant (p < 0.001), as recommended by Hair et al. (2007). Factor loadings for the measurement items were above the recommended value of 0.40 as suggested by Hair et al. (2007). The initial results indicated that the chi-square was significant (χ² = 435.123, df = 160, p < 0.001), χ²/df = 2.7, GFI = 0.868, CFI = 0.932, PNFI = .755, PCFI = 0.785, RMSEA = 0.07. The results indicated that the data were not a good fit for the model. The decision to check for the model misfit entailed examining the modification index. Results affirm the presence of a high value of the modification indices between e13 and e12, hence the decision to delete e12 (AT4). Further inspection of the modification index indicated that some of the errors (e1 and e2, e3 and e4) were highly correlated. Thus, a decision was made to co-vary them to attain an adequate model fit. All the co-varied error terms fell within construct error covariance. As such, the covariance was among the error terms of items that represented the same constructs, which validates how the measurement remained unaffected (Hair, Black, Babin, & Anderson, 2014). The final model indicated an adequate model fit as the chi-square was significant (χ² = 270.688, df = 140, p < 0.000), χ²/df = 1.9, GFI = 0.912, CFI = 0.964, PNFI = .760, PCFI = 0.789, RMSEA = 0.05. All the goodness-of-fit indices were also within the acceptable minimum cut-off points as recommended by Meyers et al. (2006), except the value of RMSEA, which was greater than the acceptable value of 0.01. Since the choice of the strictness of the indices’ threshold depends on the sample size and model complexity, the lower the sample size, the less the complexity of the indices’ threshold, and vice versa (Hair et al., 2014); as a result, RMSEA, which ranges from 0.05 to 0.08, is a close fit (Schumacker & Lomax, 2010). Figure 1 presents a summary of the measurement model.

4.7.2 Structural model

The structural model (SEM model 1) was performed using a maximum likelihood estimation approach. The results from SEM model 1 indicate the data fit well for the proposed model because the chi-square was significant at 270.688, df = 140, χ²/df = 1.933, GFI = 0.912, NFI = 0.929, CFI = 0.964, PNFI = 0.760, PCFI = 0.789, RMSEA = 0.058. The overall result in the model indicates that four of the hypothesized paths were statistically significant, supported at p < 0.05 and p <0.01. More specifically, hypothesis 1a, which proposed that ATR is significantly influencing TI, was supported (Ꞵ = 0.232, p < 0.05). On the other hand, hypothesis 1b, which hypothesized that SE is significantly influencing TI, was not supported (Ꞵ = 0.109, p = 0.214), and so was hypothesis 1c, which hypothesized that C would significantly influence (Ꞵ = −0.08, p = 0.415). Additionally, hypothesis 2a, which hypothesized that ATR is significantly influencing tourists’ AT towards the destination, was supported (Ꞵ = −0.261, p < 0.001), but not hypothesis 2b, which hypothesized that SE would significantly influence tourists’ AT towards the destination (Ꞵ = −0.02, p = 0.807). However, hypothesis 2c, which hypothesized that C would significantly influence tourists AT towards the destination, was supported (Ꞵ = 0.39, p < 0.001). Table 6 summarized parameter estimates for SEM model 1, while Figure 2 offers a summary of factor loadings for SEM model 1.

4.7.3 Mediation test

A mediation test was performed following the approach developed by Baron and Kenny (1986), which is the most widely applied technique for testing mediation. This is not the first study to employ this technique – other celebrity endorsement studies by Zhou and Whitla (2013) and Pradhan, Duraipandian, and Sethi (2016) also employed it. In this study, mediation testing was conducted using hypothesis 3. In that proposed hypothesis, it was stated that visitors AT to destination mediates the relationship between (H1a-H2c) and TI. First, the direct-effect model examined the effect of ATR, SE and C on TI in the absence of the mediator (i.e. AT): ATR and SE demonstrated a positive relationship with TI, whereas C demonstrated a negative relationship with TI. The overall finding indicated that the model adequately fit the data (χ² = 270.7, df = 141, p = 0.000, χ²/df = 1.920, GFI = 0.912, NFI = 0.929, CFI = 0.964, PNFI = 0.766, PCFI = 0.795, RMSEA = 0.05). The parameter estimate for the hypothesized path between AT and ATR was statistically significant (β = 0.19, p < 0.001), whereas the remaining hypothesized paths were insignificant. This finding implies that TI is directly and significantly explained by ATR than other celebrity dimensions. (See SEM model 2 in Table 7).

Second, a fully mediated structural model was performed (i.e. ATR, SE, C to AT and TI to AT): C was positively associated with AT (β = 0.39, p < 0.001), and TI positively linked with AT (β = 0.13, p < 0.001). The results further indicated that ATR and SE negatively correlated with AT. When parameter estimates were examined, the hypothesized paths between AT and ATR were statistically significant (β = −0.26, p < 0.001), and the path between AT to C was also statistically significant (β = 0.39, p < 0.001). The remaining hypothesized path was insignificant. The overall finding implies that the relationship between ATR and TI is indirectly explained by AT. This implies that the ATR dimension plays a crucial role in influencing visitors’ attitudes towards a destination. Thus, having a celebrity with an attractive dimension in tourism can influence how potential visitors perceive the endorsed destination. In short, when tourists have positive attitude towards a destination due to celebrity attractiveness, then most of them may be likely to consider visiting the endorsed destination. The overall result for the SEM model 3 indicates that the data were a good fit for the model (χ² = 271.4, df = 142, p = 0.000, χ²/df = 1.911, GFI = 0.964, NFI = 0.929, CFI = 0.964, PNFI = 0.771, PCFI = 0.801, RMSEA = 0.05).

Third, a partially mediated model (SEM model 4) was examined in which a direct path from the exogenous variables (i.e. ATR, SE and C) to the endogenous variable (TI) was included in the fully mediated model. Table 7 indicates that a hypothesized route from ATR to TI was the only statistically significant path; the remaining paths were insignificant. Thus, the model was a good fit for the data (χ² = 272.4, df = 143, χ²/df = 1.911, GFI = 0.911, NFI = 0.928, CFI = 0.969, PNFI = 0.776, PCFI = 0.806, RMSEA = 0.057). The bootstrap method was then performed to determine the significance of the indirect effect on the structural model. The 1,000 bootstrap samples with 95% using bias-corrected confidence intervals were performed. The overall result indicates that AT partially mediates the relationship between ATR and TI. This implies that visitors’ attitudes towards a destination may not be singly explained by celebrity attractiveness. Other factors such as culture, affordability, safety, convenience, unique experience and individual preferences can also influence visitors’ attitudes. The results are presented in Table 7:

5. Discussion

The role of celebrities in influencing travellers’ behaviour has been documented (Beeton, 2005; Pop et al., 2022). Nevertheless, the topic remains largely under-researched in tourism, particularly in the developing country. In Tanzania, the current study found celebrities influence the local tourists’ intentions to travel to tourism destinations. The findings are in line with Lee et al. (2008), Yen and Teng (2015), Salamoura et al. (2020), Zhu et al. (2022) and Lou (2022). Impliedly, celebrities have the power to influence the decisions of their social media followers. The social media platforms are an easy way for celebrities to communicate with their enthusiast followers and the followers of their followers (Jin, Muqaddam, & Ryu, 2019). As a result, their post-sharing on social media platforms is a recipe for influencing the visitors' decisions to travel to a tourist destination (Ferreira, Santos, & Santo, 2022).

Moreover, the study found that Bongo Fleva musicians such as Diamond Platinumz influence many visitors’ travel intentions, followed by Millard Ayo (TV personality), Mbwana Samata (athletes) and beauty queens Wema Sepetu (cf Table 5). Implicitly, TTB (2021) should continue using celebrities in endorsing Tanzania’s tourist attractions and products. Diamond, as the most preferred icon linked with Serengeti, can further popularize the national park among local tourists and happens to be the most visited game park for both local and foreign visitors (Faria, 2021). Also, Diamond is the first local celebrity in Tanzania to have more than 12 million followers on Instagram (Starngage, 2022). Not only Tanzania, but also other African countries which are endowed with multiple tourist attractions are argued to use celebrities as a marketing strategies to promote their destinations (UNWTO, 2023).

Although Lee et al. (2008) and Yen and Teng (2015) examined the influence of celebrity involvement on travel behaviour, studies addressing the influence of specific celebrity dimensions such as attraction, self-expression and centrality on travel intention in Tanzania are rare. In this regard, the current study has revealed a positive and significant relationship between celebrity attraction and travel intention. This finding corroborates the findings of Zhou et al. (2023) and Wang and Scheinbaum (2018) regarding how consumers’ perception of a celebrity’s attractiveness and trustworthiness plays a crucial role in inducing the people’s purchasing intention for an endorsed brand. It seems physically attractive communicators can change people’s beliefs easily (Baker & Churchill, 1977) and influence purchase intentions more than unattractive communicators. This simply means that people who are more attracted to their favourite celebrities are more inclined to consider travelling to destinations endorsed by those celebrities. Besides, people trust the credibility of social media influencers such as Instagram influencers (Djafarova & Rushworth, 2017), which has a significant effect on their perception and purchase intentions of their followers (Xiao, Wang, & Chan-Olmsted, 2018; Schouten, Janssen, & Verspaget, 2019). In short, celebrity power, to a significant extent, plays a crucial role in influencing travel intentions (Zhu et al., 2022). Thus, they should not be ignored because they have the power to influence human behaviour (Selkie, 2022).

This study further found a positive relationship between celebrity self-expression and travel intention. This finding is in line with Zhou et al. (2023), who also found celebrity involvement (self-expression) to improve the destination image significantly, and place attachment, in turn, led to tourist loyalty. Impliedly that celebrity self-identification and social identity play a crucial part in influencing visitors' travel intention. Therefore, it is crucial to maintain a celebrity’s desired image because it is linked to their hedonic/leisure activities. This finding also corroborates with Chan et al. (2018), who found that celebrity endorsement displays a favourable effective image that affected visit intentions. Consumers generally associate the endorsed brand with the celebrity’s image (Evans, 1988; Mehta, 1994). As such, any negative information regarding the celebrity endorsing a certain brand may affect not only consumers' perceptions of that celebrity but also the endorsed brand (Till & Shimp, 1995), which could also apply in any of the developing countries besides Tanzania. After all, many of them depend on tourist attractions to beef up their national coffers and economies.

Furthermore, this study has found a negative relationship between celebrity centrality and travel intention. Implicitly, the more a celebrity posts tourist activities on their Instagram account, the fewer domestic travellers’ will be motivated to visit the endorsed attractions in future. This finding contradicts Zhou et al. (2023) and Soo-Jin (2008) who reported a positive relationship between celebrity involvement and travel intentions. Surprisingly, most of the respondents in this study were young college students aged 25–26. One would expect a positive relationship because millennials tend to pay attention to celebrity activities to replicate what they admire including purchasing the endorsed product (Lippe, 2001). Such a relationship turned out to be negative probably because they did not relate their lifestyles to those of the celebrities. People do consume brands with a personality that matches their own (Fortini-Campbell, 1992). This finding implies that the selection of a celebrity to endorse tourist destinations needs to be properly considered to include those who are powerful and with whom followers can identify since a powerful celebrity can make an individual respond to his/her requests (Byrne, Whitehead, & Breen, 2003).

This study also found that attitude partially mediates the relationship between attraction and travel intention. Attitude can serve as a mechanism for explaining how celebrity involvement relates to the travel intention. In other words, the impact of celebrity attraction on travel intention can be indirectly explained by the attitudes that visitors have towards the destination. When people are more attracted to a celebrity endorsing a particular destination, their overall attitude towards that destination could be positive (Zhang, Xu, & Gursoy, 2020). This finding corroborates with that of Atkin and Block (1983) and Van der Veen and Song (2014) who similarly found that celebrity endorsers play a pivotal role in changing people’s attitudes and influencing behavioural intention. This means that even if a celebrity may not be a real expert on the endorsed product if perceived positively (favourable attitude), that celebrity could be more persuasive and generate favourable reactions towards the brand and purchase intentions (Ohanian, 1990). Celebrities, as ambassadors of tourist destinations, can influence the audience’s attitudes and behavioural intentions based on the congruence between celebrities and destinations (Wu & Lai, 2023). The connection between the endorser and the endorsed destination can increase the possibility of future travel intentions to the endorsed destinations. Moreover, individuals with high levels of celebrity involvement tend to develop positive attitudes towards the products or brands their favourite celebrities endorse (Ilicic & Webster, 2011).

6. Conclusion

The study reaffirms that celebrities can significantly influence domestic travellers’ intentions to visit various tourist attractions in Tanzania. The results particularly show that Bongo Fleva musicians outperform other celebrities in the context of Tanzania when it comes to endorsing tourist destinations such as national parks. Moreover, among the celebrities dimensions (attractiveness, centrality and self-expression) used, the study also found that attractiveness substantially influences travel intentions. In addition to attractiveness, celebrity’s attitude is partially mediating the relationship between celebrity involvement and travel intentions. The study advises destination managers and other tourism stakeholders to use local celebrities to promote destinations. However, since celebrity attractiveness is not the sole determinant factor, celebrity popularity could be another avenue for assessment. The study contributes new knowledge on the use of celebrity in re-branding tourism destinations in Africa, and modified theory of planned behaviour. Furthermore, by identifying the celebrity dimension that influences tourists’ travel intention in Tanzania, the study has set a new perspective for researchers, practitioners and tourism businesses to refine their promotional strategies and for academicians to gain a deeper understanding of visitors’ behavioural intention dynamics.

Figures

Figure 1

CFA measurement model

Figure 2

The original structural model (SEM model 1)

Table 1

Measurement items and the constructs

Construct	Measurement items	Item label
Attraction (ATR)	The celebrity activity is very important to me	TR1
	The celebrity activity offers me relaxation when pressures build up	TR2
	Participating in celebrity activity is one of the most satisfying things I do	TR3
	I really enjoy celebrity activity	TR4
Self-expression (SE)	Participating in celebrity activity says a lot about whom I am	SE1
	You can tell a lot a person by seeing them participating in celebrity activity	SE2
	When I am participating in celebrity activity I can really be myself	SE3
	When I am participating in celebrity activity others see me the way they want to see me	SE4
Centrality (C)	I find a lot of my life is organized around the celebrity	C1
	The celebrity activity has a central role in my life	C2
	I enjoy discussing celebrity activity with my friends	C3
	I find a lot of my life is organized around the celebrity activity	C4
Attitude (AT)	Your attitude towards tourist destinations and attractions in Tanzania is good-bad	AT1
	Your attitude towards tourist destinations and attractions in Tanzania is like-dislike	AT2
	Your attitude towards tourist destinations and attractions in Tanzania is pleasant-unpleasant	AT3
	Your attitude towards tourist destinations and attractions in Tanzania is positive-negative	AT4
	Your attitude towards tourist destinations and attractions in Tanzania is favourable-unfavourable	AT5
Travel intention (TI)	I intent to visit tourist attractions in the next 12 moths	TI1
	I want to visit tourist attractions in the country	TI2
	It is likely that I will visit tourist attractions in the future	TI3

Source(s): Courtesy of Lee et al. (2008), MacKenzie and Lutz (1989), Lam and Hsu (2006) and Hsu et al. (2006)

Table 2

Skewness and kurtosis results

Items	Mean	Std. deviation	Skewness	Kurtosis
ATR1	3.39	1.29	−0.44	−0.95
ATR2	3.46	1.23	−0.67	−0.59
ATR3	3.39	1.27	−0.47	−0.82
ATR4	3.54	1.33	−0.56	−0.91
SE1	3.58	1.10	−0.55	−0.56
SE2	3.67	1.08	−0.63	−0.41
SE3	3.68	1.08	−0.59	−0.46
SE4	3.73	1.13	−0.69	−0.42
C1	3.10	1.44	−0.22	−1.36
C2	3.14	1.44	−0.25	−1.33
C3	3.32	1.40	−0.33	−1.26
C4	3.27	1.46	−0.34	−1.30
AT1	5.43	2.23	−1.13	−0.39
AT2	5.34	2.10	−0.99	−0.48
AT3	5.35	2.14	−0.99	−0.52
AT4	5.20	2.27	−0.88	−0.83
AT5	5.22	2.31	−0.91	−0.84
TI1	5.45	1.56	−0.80	−0.22
TI2	5.74	1.51	−1.04	0.07
TI3	5.87	1.46	−1.19	0.50

Note(s): ATR: attraction, SE: self-expression, C: centrality, AT: attitude, TI: travel intention

Source(s): Developed by researcher (2022)

Table 3

Respondents’ demographic characteristics

	F	%
Gender
Male	150	53.8
Female	129	46.2
Marital status
Married	64	22.9
Single	215	77.1
Age
25–26	145	52
27–28	55	19.7
29–30	36	12.9
31–32	21	7.5
33–34	22	7.9
Occupation
Student	128	45.9
Working	108	38.7
Unemployed	43	15.4
Education
Primary	10	3.6
Ordinary-level secondary	27	9.7
Advanced-level secondary	56	20.1
University	186	66.7
Income
Less than 300,000 T. shilling	178	63.8
300,001–500,000 T. shilling	52	18.6
500,001–700,000 T. shilling	21	7.5
More than 700,001 T. shilling	28	10
Region
Dar es Salaam	106	38
Arusha	59	21.1
Mbeya	55	19.7
Mwanza	59	21.1

Source(s): Developed by researcher (2022)

Table 4

Reliability and validity results

Construct and measurement items	Factor loadings	CR	AVE	SV	Cronbach alpha
Attraction (ATR)		0.897	0.685	0.432	0.907
The celebrity activity is very important to me	0.87
The celebrity activity offers me relaxation when pressures build up	0.86
Participating in celebrity activity is one of the most satisfying things I do	0.79
I really enjoy celebrity activity	0.78
Self-expression (SE)		0.861	0.611	0.320	0.877
Participating in celebrity activity says a lot about whom I am	0.88
You can tell a lot a person by seeing them participating in celebrity activity	0.87
When I am participating in celebrity activity I can really be myself	0.76
When I am participating in celebrity activity others see me the way they want to see me	0.68
Centrality (C)		0.892	0.675	0.432	0.890
I find a lot of my life is organized around the celebrity	0.80
The celebrity activity has a central role in my life	0.91
I enjoy discussing celebrity activity with my friends	0.78
I find a lot of my life is organized around the celebrity activity	0.78
Attitude (AT)		0.930	0.769	0.044	0.927
Your attitude towards tourist destinations and attractions in Tanzania is good-Bad	0.85
Your attitude towards tourist destinations and attractions in Tanzania like-dislike	0.93
Your attitude towards tourist destinations and attractions in Tanzania is pleasant-unpleasant	0.92
Your attitude towards tourist destinations and attractions in Tanzania is favourable-unfavourable	0.81
Travel intention (TI)		0.835	0.629	0.052	0.832
I intent to visit tourist attractions in the next 12 moths	0.81
I want to visit tourist attractions in the country	0.84
It is likely that I will visit tourist attractions in the future	0.72

Note(s): CR: composite reliability, AVE: average variance explained, SV: shared variance

Source(s): Developed by researcher (2022)

Table 5

Group and celebrity categories that influence visitors’ travel Intention

Group category	Frequency	%	Valid%	Cumulative%
Friends’	93	33.3	33.3	33.3
Celebrity	117	41.9	41.9	75.3
Family members	69	24.7	24.7	100.0
Total	279	100.0	100.0
Celebrity category
Bongo flavour musicians (e.g. Diamond, Vanessa Mdee, Jux)	145	52.0	52.0	52.0
Bongo movie stars (e.g. Ray Kigosi, Irene uwoya)	23	8.2	8.2	60.2
Athletes (e.g. Mbwana Samatta)	39	14.0	14.0	74.2
TV personality (e.g Millard Ayo)	42	15.1	15.1	89.2
Beauty queens (Wema Sepetu, Flaviana Matata, Nancy Sumari etc.)	30	10.8	10.8	100.0
Total	279	100.0	100.0

Source(s): Developed by researcher (2022)

Table 6

Standardized parameter estimates for the SEM model 1

Hypothesis				Path coefficient (β)	P-values	Conclusion
H1a	TI	←	ATR	0.23	0.002	Supported
H1b	TI	←	SE	0.11	0.214	Not supported
H1c	TI	←	C	−0.09	0.415	Not supported
H2a	AT	←	ATR	−0.26	0.006	Supported
H2b	AT	←	SE	−0.02	0.807	Not supported
H2c	AT	←	C	0.39	***	Supported
H3	TI	←	AT	0.14	0.038	Supported

Note(s): TI: travel intention, ATR: attraction, SE: self-expression, C: centrality, AT: attitude, P: p-value

Source(s): Developed by researcher (2022)

Table 7

Direct and indirect effects

GOFI	SEM model 1 original model	SEM model 2	SEM model 3	SEM model 4
X2	270.7	270.747	271.404	272.43
Df	140	141	142	143
P	0.000	0.000	0.000	0.000
X2/df	1.97	1.920	1.911	1.911
GFI	0.912	0.912	0.912	0.911
NFI	0.929	0.929	0.929	0.928
CFI	0.964	0.964	0.964	0.969
PNFI	0.760	0.766	0.771	0.776
PCFI	0.789	0.795	0.801	0.806
RMSEA	0.05	0.058	0.057	0.057

Note(s): X²: chi-square, df: degree of freedom, X²/df: Ratio of Minimum Discrepancy to degree of Freedom, p: p-value, GFI: goodness-of-fit index, NFI: normed fit index, CFI: comparative fit Index, PNFI: parsimonious normed fit index, PCFI: parsimony comparative fit index, RMSEA: root mean square error of approximation

Source(s): Developed by researcher (2022)

Table A1

Name of the celebrity and tourist attraction

Name of celebrity	Frequency	%	Valid%	Cumulative%
Sam Mahela Mount Kilimanjaro	2	0.7	0.7	0.7
Jux Manyara National Park	8	2.9	2.9	3.6
Diva Serengeti National Park	1	0.4	0.4	3.9
Farhia Middle Mount Kilimanjaro	2	0.7	0.7	4.7
Wema Sepetu Ngorongoro National Park	17	6.1	6.1	10.8
Millard Ayo Lake Manyara National Park	27	9.7	9.7	20.4
Nandy Ngorongoro National Park	24	8.6	8.6	29.0
Vanesa Arusha National Park	3	1.1	1.1	30.1
Maua Sama Tarangire National Park	4	1.4	1.4	31.5
Samata Mount Kilimanjaro	25	9.0	9.0	40.5
Alikiba - Mikumi National Park	16	5.7	5.7	46.2
Rose Mhando zanzibar	2	0.7	0.7	47.0
Diamond - Serengeti National Park	28	10.0	10.0	57.0
Rayvany -Arusha National Park	4	1.4	1.4	58.4
Ray Kigosi -Tengeru National Park	3	1.1	1.1	59.5
Jackline Mengi pictures zanzibar	2	0.7	0.7	60.2
Nancy Sumari Serengeti National Park	8	2.9	2.9	63.1
Hashim Thabit Mount Kilimanjaro	8	2.9	2.9	65.9
Zuchu Zanzibar	18	6.5	6.5	72.4
Harmonize - Mkomazi National Park	22	7.9	7.9	80.3
Christina Shusho- Zanzibar	2	0.7	0.7	81.0
Aslay - Mikumi	3	1.1	1.1	82.1
Juma Lokole - Lake Manyara	1	0.4	0.4	82.4
Jokate- Magoroto Lushoto	2	0.7	0.7	83.2
Idris Sultan -Mount Kilimanjaro	2	0.7	0.7	83.9
Gabo - Matema beach	1	0.4	0.4	84.2
Barnaba- Ruaha National Park	2	0.7	0.7	84.9
Mimi Mars- Serengeti	2	0.7	0.7	85.7
Mboso- Mikumi National Park	4	1.4	1.4	87.1
Belle 9- Ruaha National Park	1	0.4	0.4	87.5
Rose Ndauka- Ngorongoro National Park	2	0.7	0.7	88.2
Jay melody - Mount Kilimanjaro	1	0.4	0.4	88.5
JB- Mkomazi National Park	1	0.4	0.4	88.9
Meja Kunta - Katavi National Park	1	0.4	0.4	89.2
Godluck Gosbert- Mikumi National Park	1	0.4	0.4	89.6
Mboni Masimba- Udzungwa National Park	2	0.7	0.7	90.3
Hemed - Mikumi National Park	4	1.4	1.4	91.8
Irene Uwoya -Gombe National Park	5	1.8	1.8	93.5
Prof Jay - Udzungwa National Park	1	0.4	0.4	93.9
Salim Kikeke - Mount Kilimanjaro	2	0.7	0.7	94.6
Marioo- Mount Kilimanjaro	1	0.4	0.4	95.0
Steve Nyerere - Kitulo National Park	2	0.7	0.7	95.7
Matumla- Mikumi National Park	5	1.8	1.8	97.5
Chuchu Hans Serengeti National Park	3	1.1	1.1	98.6
Johari Katavi	3	1.1	1.1	99.6
Hamza Kasongo-Mount Kilimanjaro	1	0.4	0.4	100.0
Total	279	100.0	100.0

Source(s): Developed by researcher (2022)