Research Article: 2021 Vol: 25 Issue: 3
Peter Kwasi Kodjie, University of Professional Studies, Accra
Ibn Kailan Abdul-Hamid, University of Professional Studies, Accra
Winston Asiedu Inkumsah, University of Professional Studies, Accra
Robert Yaw Amponsah, University of Professional Studies, Accra
This investigation provide an empirical evaluation of the influence of celebrity characteristics on brand resonance outcomes in the luxury automobile market of Ghana. A survey of users or owners of luxury automobiles was done. One hundred and sixteen questionnaire were used for further analysis. A structural equation modelling was used. The completed questionnaires were cleaned and used for further analysis. The study found that out of the three constructs that define celebrity characteristics in the luxury automobile industry only celebrity attractiveness had a positively significant effect on brand resonance outcomes.
Celebrity, Brand, Automobile, Luxury Brand Endorsement.
Many brands with luxurious appeal adopt celebrities in their advertising campaigns. For instance, studies have estimated that celebrities showed up in 15% of US adverts, 24% and 45% for (India and Taiwan) adverts respectively (Crutchfield, 2010). Celebrity endorsement is conceptualised as an agreement between an individual who enjoys public recognition (a celebrity) and an entity (a brand) to use the celebrity to promote the entity (Bergkvist & Zhou, 2016). Celebrity endorsements are known to be effective in that the pre-constructed images of celebrities generally provide many beneficial advantages in promoting luxury brands (Liu, 2012). Specifically, celebrity endorsements increase ad viewers’ attention, polish brand images, facilitate the introduction of new brands, reposition existing brands (Erdogan 1999), and generate favourable attitudes toward ads and brands (Hung et al., 2011), as many prior studies have empirically demonstrated. This craze for the use of celebrities has much following both in the developed economies and also gaining grounds in emerging economies (Sharma, 2016; Turner, 2004). Mainly, in Japan and South Korea, 70 per cent of commercials use a celebrity. The African setting is not left out of this growth with Morocco, Nigerian, South African, Ivory Coast and Ghanaian firms using celebrities to generate favourable attitudes towards ads and brands (Zakari et al., 2018; Cohen, 2019). The sub-Saharan African luxury goods market is burgeoning and it is poised to become the world’s second-fastest-growing region for the consumption of luxury goods. Specifically, its luxury markets are expected to grow by 30 percent over the next 5 years (Cohen, 2019). This expected growth is due to a growing urbanization, the continent’s changing demographic profile with a sharp rise in the youthful population which is embracing technological advances, an expanding middle class, in addition to an appetite for luxury goods accompanied by a growing desire for status and the ability to acquire status through the purchase of luxury goods (Cohen, 2019).
Brand endorsers possesses characteristics that can generate favorable attitudes towards luxury brand (Zakari et al., 2018). Extant literature have shown the impact of celebrity endorsers on brand recall and endorsed brand attitudes (Till et al., 2008; Kahle & Homer, 1985). Also, the economic advantage of celebrities have been assessed to measure its efficacy (Kang et al., 2019).
Trustworthy celebrities used to endorse brands, do influence consumer preference for those brands (Amos et al., 2008). Using celebrities to endorse brands also help increase other intangible determinants of the brand’s value which is essential to its edge and growth in the sector (Kotler and Keller, 2016). But, regardless of current attention on returns on advertising investment (Taylor, 2010), there is limited studies on how celebrity characteristics can explain such dimensions as behavioural loyalty, attitudinal attachment (personal attachment which involves love for the brand), sense of community (feeling of kinship or affiliation with others who use the brand), and active engagement (going the extra mile as a customer and becoming a brand evangelist or ambassador). More so, limited number of studies exist in establishing a link between celebrities and brand resonance dimensions all have found space within developed economies such as Europe and United States of America (Christodoulides & de Chernatony, 2010). The little that have been done in emerging markets have only found space in Asian jurisdictions (Sharma, 2016).
A study on sub Saharan luxury market which raked in $6 billion in revenue in 2017 (Cohen, 2019), will add value to the body of knowledge and improve strategies for practitioners. In addition, appreciating the sub-Saharan consumer can offer luxury brand marketers with valuable information for building a strong brand in the African market and building appropriate approaches for achieving an urge over rivals. This study investigates whether the celebrity-user gender congruity can differentiate the effect of celebrity characteristics on brand resonance dimensions. In addition, the extant literature has assess the effect of for instance credibility and gender and has produced mixed results (Gilal et al., 2020; Knoll & Mattes, 2017). Therefore, studies have indicated the male endorsers are more impactful for male consumers while female brand ambassadors are more impactful for female consumers (Gilal et al., 2020; Luoh & Lo, 2012). Such differences in findings have brought to the fore, whether industry professionals choose a female ambassador over a male brand ambassador when they aim at reaching male customers or the other way round.
The remainder of the paper is structured as follows: initially a background of the luxury automobile industry is provided followed by a review of the relevant theoretical literature and the supporting theory on celebrity endorsement and brand resonance dimension. Further hypothesis is developed and is followed by the research methodology, discussions, implications, conclusions and limitations.
Luxury Automobile Industry
The luxury automobile industry is one of the most valuable luxury markets. Changing dynamics of consumers in emerging markets like China and India has been key to the burgeoning growth of the industry (Statista, 2019). However, the American market remains an essential target country for the industry. Mercedes Benz leads the luxury car market share with BMW, Lexus, Audi, Tesla, Volvo, Land Rover and Jaguar brand round off the list of leading luxury vehicles in the world (Wagner, 2019). The industry audience of focus per a recent analysis was those aged 35-59, baby boomers generation, with high disposable income (SAR, 2019).The African market has high growth prospects. With an estimate population of 1.1 billion, it has only 1% of the global new car sales. As well, close to 90% of such sales are found in South Africa.
That notwithstanding the Sub-Saharan region with a potential for an average of 2.5 million cars is expected to offer high returns to players who are willing to enter the jurisdiction. The Ghanaian luxury automobile industry was initially characterized by major household makes like Jaguar, BMW and Mercedes Benz.
Associative Network Theory
The theory holds that knowledge that the consumer has about a brand is characterized by a network of nodes and connecting links (Keller, 2013). The nodes are the stored information or concepts in memory while the links are the association between the nodes (Keller, 1993). When a piece of stored information or idea is brought to the fore, it becomes an avenue of initiation of all other information in the memory of consumer reflected in a process of spreading activation (Spry et al, 2011; Dwivedi et al., 2015). In addition, the connecting link tend to vary in strength, which determines the amount of brand knowledge retrieved from memory (Dwivedi et al., 2015). As a result, it is this brand knowledge in memory that is reflective of brand equity (Keller, 2013). In endorsement of brands by celebrities, the celebrity and the brand are both made up stored information and connecting links in the customer’s mind (Till, 1998). Consumer’s behaviour towards the brand and celebrity are integral to the connecting link (Till & Busler, 2000). The endorsement process binds the brand and the celebrity, though prior to this they were not bound (Spry et al, 2011). The underlying theme is to the effect that through spreading activation, meaning and association from the celebrity get transferred to the brand. Celebrity endorsers are seen as individuals who are universally recognised by the generality of society as a result of their achievement in their area of work (Zakari et al., 2018). These individuals who may be found in varied area such as acting, entertaining, modelling, singing, sports among others, are also perceived as an indirect avenue of making brands salient and meaningful and as such an alternate approach to building brand resonance (Keller, 2013).
Celebrity Endorsement Characteristics
In a bid to make celebrity endorsement very effective, there must be a link between the celebrity, brand and customers (Albert et al., 2017). Marketers should therefore not just select any celebrity as an endorser of their brand but must consider the fit of the celebrity in promoting the brand to the targeted customers (Albert et al., 2017; Pradhan et al., 2014). There must be a bond between the celebrity and the brand being endorsed. Celebrity characteristics is thus very critical in enhancing the effectiveness of the celebrity endorsement process (Kapitan & Silvera, 2016). Studies have assessed the celebrity effects based on severally characteristics like attractiveness, likability, expertise and trustworthiness (Eisend & Langner, 2010; Rossiter & Smidts, 2012; Osei-Frimpong et al., 2019). These characteristics also constitute what Amos et al. (2008) & Ohanian (1990) refers to as the consumer-perceived credibility. Celebrity attractiveness is multi-dimensional in nature (Zakari et al., 2018). Attractiveness goes beyond the physical features, to encompass personality and athletic abilities (Erdogan, 1999). Physically attractive celebrities are generally found to be more preferred among consumers than non-physically attractive ones (Chan et al., 2013). Trustworthiness is explained as the extent to the level of confidence consumers put in celebrity’s intent in conveying the claims they consider essential (Ohanian, 1990). In addition, Kang et al. (2019) referred to expertise as how much knowledge, experience and professionalism a celebrity has with respect to the marketing offering under review. Customers see celebrities with enough knowledge as possessing the knack to enable them believe the celebrity’s point of view (O’Mahony & Meenaghan, 1998). Osei-Frimpong et al. (2019) therefore argue to the effect that all these characteristics of celebrities to a large extent influence emotional attachment to a brand albeit leading to brand resonance.
Brand Resonance is characterised in terms of intensity, or the depth of the psychological bond that customers have with the brand, as well as the level of activity engendered by this loyalty (Keller, 2013). Studies abound to the effect that customers showing a high degree of brand knowledge will offer a great desire to patronise a brand (Fleck et al., 2012; Sharma, 2016). Extant studies on brand equity adduce the fact that two perspectives abound (Christodoulides & de Chernatony, 2010). CBBE (consumer-based brand equity) from a customer perception viewpoint using antecedents such as brand awareness, association, perceived quality, and a customer behaviour point of view using loyalty supported by willingness to pay premium. Brand equity is an outcome of building a strong brand requires in addition to changing behaviour in terms of repeat purchase (loyalty), also establishing attitudinal attachment, sense of community, and active engagement (Keller, 2013).
Celebrity Characteristics and Brand Resonance Outcomes
A variety of characteristics of celebrities determine consumer behaviour towards a particular value proposition. Recent studies contend that consumer choice of a brand is influenced by secondary associations of which celebrities and their characteristics form a great part (Keller 2013; Osei-Frimpong, 2019). Celebrity characteristics such as attractiveness, expertise and trustworthiness have been investigated empirical with varied results in the literature (Tzoumaka et al. 2016; Zakari et al., 2019). These characteristics are influenced by the source credibility model which contends that the effectiveness of a celebrity’s endorsement is contingent on the perceived levels of expertise, trustworthiness and attractiveness (McGuire, 1969). To it, endorsement from “a credible source (e.g. celebrity) can influence beliefs, opinions, attitudes and behaviour through a process called internalisation, which occurs when receivers accept a source influence in terms of their attitude and value structures”(Wang et al., 2017 pg.2). For instance, in their study, Eisend & Langer (2010) claimed that attractiveness is a critical characteristic which consumers utilize in their decision making of patronizing brands and that an attractive endorser can influence positively consumer’s attitudes towards brands. This in agreement with Osei-Frimpong et al. (2019) and Erdogan (1999) who argue that attractiveness which forms part of consumer-perceived credibility has a positive effect on brand attitudes and loyalty. This notwithstanding, Eisend & Langer (2010) found out that the impact of celebrities who are unattractive can worsen consumers behavioral outcomes towards a brand.
Further, the expertise of the endorser which reflect their knowledge of the offering under review have also been found to directly influence the consumer to change his/her behavior (Wang et al., 2017). This is in sync with earlier studies which claim that a celebrity with adequate expertise has the ability to make consumers believe their viewpoint (O’Mahony & Meenaghan, 1998). Liu (2012) concurs with the above but as well indicated that the frequency of mentions and emotional polarity proportion regarding celebrity expertise influences sustainable firm value as well as overall consumer attitudes. However, Thomas-Hunt et al (2003) argued that perceived expertise, different occupations, or unique personal abilities create different levels of social status as well as other behaviour pattern. In addition, trustworthiness which is seen as the celebrity’s reliability, honesty and integrity and is hinged on perceptions of the consumer, have been found to make endorsement effective in changing behaviour (Shimps, 1997; Bozic, 2017). The above is in sync with other scholars who claim that recommendations from honest and reliable endorsers lead to attitude change as consumer believe in such endorsements (Saldanha et al., 2018; Djafarova & Rushworth, 2017).
However, extant literature have indicated that trustworthy celebrities were much effective when it came to endorsement of non-durable goods than to durable goods (Eren-Erdogmus et al., 2016). In addition to assessing the effect of celebrity characteristics on brand resonance dimensions for luxury automobile brands, the current research as well aimed at exploring whether theoretically significant moderator like gender congruity can set apart the influence of characteristics like celebrity attractiveness, expertise and trustworthiness on brand resonance dimension. The moderation of celebrity-user gender congruity into the association between celebrity attractiveness, expertise and trustworthiness and brand resonance dimensions is significant, in that several studies have shown that gender congruity can make a significant contribution in changing attitudes and behaviour of customers (Gilal et al., 2020; Gilal et al., 2020b; Su & Reynolds, 2017).
For instance, studies of Knoll & Matthes (2017) on effectiveness of celebrity endorsement indicated that matching endorsers produce more positive behaviour and better behavioural expectation relative to non-matching endorsers. Researchers have suggested that male celebrities were considered to influential for male customers whereas female endorsers have become the face of brands that appeal to females who have the means to acquire (Liu & Brock, 2011). In addition, claims are made to the effect that male celebrities have essential characteristics to male consumers which make them distinct from female brand ambassadors (Gilal et al., 2020; Branchik & Chowdhury, 2017). Further, female brand ambassador hold important characteristics which appeal to female customer and make them distinct than male celebrities Gilal et al., (2020); Klaus & Bailey, (2008).
Based on the above discussion, the theoretical framework Figure1 and the hypothesis are indicated below;
H1a. Celebrity attractiveness positively influences behavioural loyalty
H1b. Celebrity attractiveness positively influences emotional/attitudinal attachment
H1c. Celebrity attractiveness positively influences sense of community
H1d. Celebrity attractiveness positively influences active engagement
H2a. Celebrity expertise positively impact behavioural loyalty
H2b. Celebrity expertise positively influence emotional/attitudinal attachment
H2c. Celebrity expertise positively influence the sense of community
H2d. Celebrity expertise positively influence active engagement
H3a. Celebrity trustworthiness positively impact behavioural loyalty
H3b. Celebrity trustworthiness positively influence emotional/attitudinal attachment
H3c. Celebrity trustworthiness positively influence a sense of community
H3d. Celebrity trustworthiness positively influence active engagement
H4. Gender congruity moderates the impact of celebrity attractive, celebrity expertise and celebrity trustworthiness in luxury automobile acquisitions
This study’s approach is quantitative rather than qualitative because of the adoption of the positivist stance on the onset. The choice of quantitative is because the nature of the data is in a form of numbers which allows for quantitative analysis. As well because the study’s intent is to test proposed hypothesis and investigate the influence of celebrity endorsement characteristics on brand resonance outcomes from the luxury automobile market, a quantitative approach is the likely option. In addition, because this study’s path is linear in outlook, the obvious option is the quantitative, hence its choice. Survey is a popular research strategy in management research that is widely accepted and used also in luxury goods purchase and use. This widely used strategy is associated with the deductive approach and is used within the positivist paradigm (Neuman, 2011). The survey strategy uses structured observation, structured interviews and questionnaires for data collection. Clients of luxurious automobile dealers in Ghana found within DVLA database constitute the population for this research. Accordingly, a sample size of 300 made up of male and female luxury car users aged between 30-60years and found within DVLA database was used for this study. The non-probability sampling technique involves the selection of samples. Clients within the DVLA database who are well informed as well as a prudent targets for celebrity influence and can better understand and answer the questions.
The design of the questionnaires was primarily based on multiple-item measurement scales adopted from previous research on celebrity characteristics and consumer behaviour outcomes. The first section of the questionnaire elicited demographic information on age, gender, occupation, income levels and educational qualifications together with the brands of luxurious automobiles bought and driven. The second section obtained information on celebrity characteristics. The third section as well obtained data on resonance outcome with respect to for example behavioural loyalty, attitudinal/emotional attachment, sense of community and active engagement. The questionnaire was a Likert scale type, and anchored on 1 “strongly disagree” and 5 “strongly agree”. Prior to administration of the survey, a panel comprised of six graduate research students reviewed the measurement items (Malhotra and Birks 2013). Questionnaires were developed in English. Subsequently, the questionnaires were answered by the respondents who were contacted at various meeting points – workplaces, university campuses, churches and homes. After six week, 300 responses were obtained out of which 116 were found usable for the analysis after a thorough cleaning of data.
The various items used in measuring the constructs were developed based on literature. The customers were asked to assess the influence of celebrity characteristics on brand resonance outcomes of luxurious automobiles they currently use or intend to buy. They responded on a Likert scale of 1 “strongly disagree” and 5 “strongly agree”. The items used in measuring the celebrity characteristics (attractiveness, expertise and trustworthiness) were adapted from Priyankara et al. (2017) and Wang et al. (2017). The items used in measuring brand resonance outcome (behavioural loyalty, attitudinal/emotional attachment, sense of community and active engagement) were also adapted from Keller (2013). The study operationalised gender congruity as “0”= explaining better gender congruity and “1”= explaining gender incongruity, stimulated by Gilal et al. (2020) and Pradhan et al. (2017). The study controlled for the effect of the participant’s age, income and education level which was in agreement with prior studies (Gilal et al., 2020; Pradhan et al., 2017).
The IBM SPSS was used for “descriptive statistics”, including “frequency tables”, central tendency and regression. The constructs and their items were loaded in a factorial confirmatory model (CFA). The robust maximum-likelihood method (MLR) used in Mplus 8 software package was applied for estimating the CFA model. The model was then assessed and verified for fitness, validity and reliability. Measuring model fit comprises two main conditions: absolute fit and comparative fit (Bagozzi & Yi, 2012). Mostly absolute fit criteria include chi-square (χ2), root-mean-square-error of approximation (RMSEA), root-mean-square-residual (RMSR), goodness-of-fit-index (GFI) and adjusted goodness-of-fit index (AGFI). The absolute fit is based on the variances between the observed and the covariance matrix (Hair et al., 2014). The comparative fit, on the other hand, reflects whether, in accounting for the observed data, the model used is more enhanced than other models. Hair et al. (2014) confirms that comparative fit index (CFI), normed fit index (NFI) and relative non-centrality index (RNI) are some of the criteria used in the comparative fit. The reliability and validity of the questionnaire scales are important for obtaining the most accurate results. Reliability was determined by Cronbach’s alpha coefficient and composite reliability. Assessing Reliability using Cronbach’s alpha means considering a threshold of 0.6 as a minimum, as compared to the composite reliability, which has an acceptable minimum of 0.5 (Hair et al., 2014). The validity, on the other hand, covers convergent validity and discriminant validity. Fornell-Lacker criterion and cross-loadings are used to assess discriminant validity. The square root of the average variance extracted (AVE) is greater than the correlation shared between the construct and the other constructs. AVE loading should have a minimum of 0.5 to be satisfactory (Hair et al., 2014) in assessing Convergent validity.
Descriptive statistics relating to the characteristics of respondents are discussed below. Out of the 116 responses used for the analysis, 67.9% were male, with 31.8% representing females. In terms of age, the greater percentage of 95.1% was found between the age range of 40 and above and 30-39, with 40 and above recording the highest age percentage of 4.9%. Respondents varied in terms of their education levels with greater number of respondents indicating they had their master’s degree (53.4%), doctorate (24.9%), diploma (11.1), and undergraduate degree (10.5%). For the type of luxury brands, 55.4% of the respondents indicated that they drove Mercedes Benz, followed by 25.6% indicating they drove a BMW. Range Rover respondents also accounted for 2.6%, while Lexus users who responded accounted for 16.4%.
Also, for the number years respondents have driven a luxury car, 45.2% of the respondents indicated that they have driven a luxury car for more than 10 years, followed by 35.5% showing they have driven luxury car between 5-10years. Those who have driven luxury car for less than five years accounted for 19.3%. Further, for the occupation, 57% of the respondents indicated they were in entrepreneur ventures, while 43% said they were CEOs. Because this study’s focus was on construct associations and not descriptive insights, the study did not take into consideration weighting the sampling elements.
The constructs and their items were loaded in a factorial confirmatory model (CFA). The robust maximum-likelihood method (MLR) used in Mplus 8 software package was applied for estimating the CFA model. In measuring the goodness-of-fit (GOF) of the CFA model, indexes such as chi-square test statistic, comparative fit index (CFI), Tucker-Lewis index (TLI) and root mean square error of approximation (RMSEA) were used. GOF for the model of the data is arrived at once values were above the threshold of 0.90 for CFI and TLI, and below 0.08 for RMSEA (Hair, Babin, & Anderson, 2014). The results indicated that the CFA model had a good fit for the data. The GOF values were: MLR ꭓ2(231)= 403.719, CFI = 0.917, TLI = 0.901, and RMSEA = 0.093; 90% confidence interval (C.I.) of 0.067–0.080. In establishing the reliability of each of the scales, Cronbach’s alpha (α) composite reliability (CR) and factor determinacy (FD) were used for such assessment (Appendix B). Both Cronbach’s alpha (α) and composite reliability (CR) values for the latent variables ranged from 0.76 to 0.91. The above values exceed the threshold of 0.7 proposed by (Bagozzi & Yi, 1988). The factor determinacy (FD) values for all the latent variables were above 0.88, exceeding the proposed threshold of 0.80 (Muthen and Muthen, 2012). In addition, loading estimates were greater than 0.660 and statistically significant. As well, t-values for the items ranged from -1.125 to 19.083 (p < 0.000). Convergent validity was established as there was no evidence of cross-loading (Hair et al., 2014).
For discriminant validity, the study accessed it based on Fornell and Larcker’s (1981) criterion by testing if the square root of the Average variance extracted is greater than its correlation with each of the remaining constructs. Initially, the AVE for each construct was calculated. The average variance extracted of the constructs ranged from 0.606 to 0.791 (Appendix B). The values calculated were in agreement or exceeded the acceptable threshold of 0.5 (Fornell & Larcker, 1981). It is also evident from (Table 1) that the square root of the AVEs for each construct is greater than the cross-correlation with other constructs.
|Table 1 Construct Standardised Loadings, FD, CR, Reliabilities, and Intercorrelations|
Path Analysis for Celebrity Characteristics and Brand Resonance Outcomes
The influence of celebrity characteristics on brand resonance outcomes was assessed using path analysis with structural equation modeling in Mplus 8. MLR was the estimator method used. Also, the GOF values for the model were MLR ꭓ2(238)= 347.390, CFI = 0.931, TLI = 0.920, and RMSEA = 0.077; 90% confidence interval (C.I.) of 0.048–0.063. As shown in Table 2, the regression results used the parameters based upon the independent variables to test the earlier framework in Figure 1 above. For hypothesis 1a-d, 2a-d, 3a-d; the study hypothesized the degree to which the three constructs that define celebrity characteristics in the luxury automobile industry is positively related with brand resonance outcomes. However, per the results shown in Table 2 below only celebrity attractiveness is positively related to the brand resonance outcomes within the luxury automobile industry. All the other hypotheses were rejected. Therefore, it is instinctive from the result to take cognizance of the fact that in general terms, celebrity attractiveness is positively related to the brand resonance outcomes within the luxury automobile industry in APPENDIX A.
|Table 2 Hypotheses and Path Analysis|
|Relationships among constructs|
For the analysis of interrelationship among the brand resonance outcome, ACEN and SECOM; SECOM and ATTA; and ATTA and BELO was calculated. The outcome of the calculation suggests a significantly positive effect of SECOM on ATTA and ATTA on BELO, but no relationship was found for ACEN and SECOM. The outcome of the interrelationship is shown in Figure 1 and Table 2.
The paper investigated the influence of celebrity characteristics on brand resonance outcomes from the luxury automobile market. The results indicate that, out of the three constructs that define celebrity characteristics in the luxury automobile industry only celebrity attractiveness had a positively significant effect on brand resonance outcomes. This in agreement with Osei-Frimpong et al. (2019); Erdogan (1999) whose empirical study found that attractiveness has a positive effect on brand attitudes and loyalty. As well, it is in agreement with Zakari et al., (2018); Wang et al., (2017) who found that celebrity attractiveness play a critical role in influencing beliefs, opinions, attitudes and/ or behaviour through a process called internalization.
The empirical test reveals a significantly positive effect of sense of community on attitudinal attachment and attitudinal attachment on behavioral loyalty, but no relationship was found for active engagement and sense of community. From the foregoing, luxury automobile brands should decide on their long-term goals of their automobiles and as well take note that attractiveness as a characteristics of a celebrity align with their automobile because confirmation has been adduced from this study. Further studies can study the moderating role of the above characteristics on luxury automobile reputation in today’s global warming prone environment.
|Appendix A Measuremt of Aotpmobiles Promotions|
|Celebrity Attractiveness Priyankara et al. (2017) and Wang et al. (2017)||α=0.82|
|ATTR_1||Luxury automobile adverts have physically attractive (beautiful, elegant, classy) celebrity endorsers||0.777||0.601||3.716||-0.98||0.505||CR=0.860|
|ATTR_2||Luxury automobile promotions have attractiveness as an important characteristic for a celebrity endorser||0.799||0.64||3.741||-0.63||0.031||FD=0.951|
|ATTR_3||In luxury automobile sector, physically attractive endorser influences my attitude to engage.||0.811||0.659||3.733||-0.82||0.347||AVE=0.606|
|ATTR_4||In luxury automobile promotions, I remember a brand that is being endorsed by attractive celebrity more||0.725||0.525||3.638||-0.63||-0.348|
|Celebrity Expertise Priyankara et al. (2017) and Wang et al. (2017)|
|EXPER_1||In luxury automobile promotions, I buy a product if the celebrity endorsing it is an expert.||0.762||0.592||3.167||-0.27||-1.153||α=0.83|
|EXPER_2||In luxury automobile promotions celebrity endorsers have practical knowledge in the industry||0.769||0.591||2.836||-0.28||-1.034||CR=0.849|
|EXPER_3||In luxury automobile promotions, I think a brand that is being endorsed by an expert celebrity is more trustable||0.756||0.567||3.249||-0.93||0.639||FD=0.951|
|EXPER_4||In luxury automobile promotions , I prefer celebrity endorsers with sufficient experience in his or her field||0.773||0.591||2.698||-0.66||0.048||AVE=0.685|
|Celebrity Trustworthiness Priyankara et al. (2017) and Wang et al. (2017)||α=0.86|
|TRUS_1||In luxury automobile promotions, celebrity endorsers are reliable||0.824||0.68||3.5||-0.53||-0.443||CR=0.917|
|TRUS_2||In luxury automobile promotions, celebrity endorsers are dependable||0.841||0.71||3.526||-0.45||-0.442||FD=0.966|
|TRUS_3||In luxury automobile promotions, celebrity endorsers are honest||0.906||0.821||3.414||-0.28||-0.534||AVE=0.734|
|TRUS_4||In luxury automobile promotions, celebrity endorsers are trustworthy||0.835||0.696||3.431||-0.27||-0.618|
|Behavioral loyalty (Keller, 2013)||α=0.85|
|BELO_1||I feel loyal to my luxury automobile endorsed by my favourite celebrity.||0.816||0.672||3.586||-0.3||-0.579||CR=0.818|
|BELO_2||It is likely that brands endorsed by my favorite celebrity will be my first choice||0.721||0.539||4.009||-0.93||1.05||FD=0.948|
|BELO_3||I will not buy another brand of luxury car if a brand endorsed by my favorite celebrity is available||0.783||0.585||3.724||-0.48||-0.371||AVE=0.70|
|Attitudinal/emotional attachment (Keller, 2013)||α=0.78|
|ATTA_1||I really love my brand of luxury car||0.66||0.301||4.043||-1.24||1.84||CR=0.769|
|ATTA_2||My luxury car brand is the only automobile I prefer||0.682||0.614||3.742||-0.86||0.729||FD=0.940|
|ATTA_3||I feel luxury car is the only automobile brand that I need||0.829||0.671||3.647||-0.77||0.164||AVE=0.729|
|Sense of Community (Keller, 2013)||α=0.87|
|SECOM_1||I feel like I almost belong to a club with other customers of my luxury automobile brand||0.916||0.841||3.957||-1.12||1.444||CR=0.919|
|SECOM_2||I really identify with people who use this brand.||0.923||0.851||4.009||-0.92||0.878||FD=0.968|
|SECOM_3||This is a brand used by people like me.||0.826||0.679||4.06||-1.08||1.303||AVE=0.791|
|Active Engagement (Keller, 2013)||α=0.84|
|ACEN_1||I am always interested in learning more about my luxury automobile brand||0.806||0.649||3.888||-0.84||0.383||CR=0.866|
|ACEN_2||I am always interested in going the extra mile to access a service about my luxury automobile brand||0.845||0.716||3.853||-0.66||-0.205||FD=0.953|
|ACEN_3||I am always interested in recommending to others to buy my luxury automobile brand||0.829||0.684||3.621||-0.48||-0.838||AVE=0.684|