The Impact of Perceived Effectiveness of Celebrity Endorsement on Perceived Brand Personality

The application of celebrity endorsement has been an intensive practise due to head-on brand competitions. It forces companies to depend on celebrities as one of the responsive strategies. However, it is crucial to examine the absolute effectiveness of celebrity endorsement towards the endorsed-brands as to rationalize marketing expenses incurred. This is a critical matter for the industries or sectors where celebrity endorsement takes place quite competitively. Alongside the said preview, this study focused to evaluate the perceived effectiveness of celebrity endorsement on perceived brand personality. Accordingly, it has examined the relationship between celebrity endorsements related dimensions with perceived brand personality to explain how it does make resultsdriven celebrity endorsement. Findings revealed that there is a positive relationship between perceived effectiveness of celebrity endorsement and perceived brand personality. Accordingly, it was noticed that the attractiveness and trustworthiness of celebrity endorsement prominently influence perceived brand personality as reporting significant levels of coefficient values and probability in both cases. Further, expertness of the celebrity is also playing a major role. In line with the findings, this paper contributes for the managerial practices referring how to use effective celebrity endorsement strategies to make effective brand building strategies. This could be used as a decisional guide to companies in managing celebrities to enhance brand performance without tarnishing the respective brand image and brand personality. Finally, it has presented the future research directions as a contribution to knowledge by reviewing the research gaps found in Sri Lankan context. Citation: Dissanayake DMR, Weerasiri RAS (2017) The Impact of Perceived Effectiveness of Celebrity Endorsement on Perceived Brand Personality. J Account Mark 6: 244. doi: 10.4172/2168-9601.1000244


Introduction
Background of the study: Review on practise and knowledge perspectives on celebrity endorsements Celebrity endorsement has been figured out as one of the widely accepted brand building application in different product contexts and market contexts as per the empirical evidences found in many studies. Sri Lanka is also fund as one of the contexts of uses celebrity endorsements at a significant level across the industries. Adding to the said, it refers that celebrity endorsement practise in Sri Lanka has been a growing trend as per the practise-related evidences found in different products and services sectors [1]. According to the studies found in Sri Lanka, the overall impact of celebrity endorsement towards endorsed brands had been investigated with different brandrelated behaviours. Meanwhile, it has referred that Sri Lankan scenario of celebrity endorsements is prominently dominated by sport sector personalities, particularly cricketers. The significance of the industry practises of celebrity endorsements has been reviewed in different studies by concluding service sector brands prominently use celebrities to develop brand related performances [2]. Celebrity endorsement is a form of brand or advertising campaign that involves a well-known person using their fame to promote a product or service. In brief, celebrity has been claimed as name which is made by related news [3] and celebrity endorsement relates to reference group [4] and some studies said celebrities as credible spokespersons or opinion experts who commit for the endorsed brands [5]. Further, it has referred that celebrity as one who traces recognition from the society in line with the personality of the celebrity [6]. Accordingly, celebrity has been explained in terms different notions as mentioned in the review aforesaid. As per the practice context of Sri Lanka, mostly used forms of celebrity endorsements are advertisements and attending PR events. Meanwhile, the application hindsight of celebrity endorsement practices of Sri Lanka has been reviewed in the studies of Dissanayake [7], Dissanayake and Ismail, [2] and Sri Lanka Celebrity Index Report [1]. This study highlights the notion multiple endorsement as a trendy practise claiming further studies to examine how over endorsements or multiplicity influences brand evaluation leading to brand-related behaviours. Meanwhile, impact generated through celebrity characteristics and factors related to celebrity practices on the endorsed message is referred as effectiveness of celebrity endorsements [8]. It reflects celebrity effectiveness as a function of both celebrity characteristics-related factors claimed as source-based factors, and the celebrity managements-related factors. Accordingly, the notion of effectiveness of celebrity endorsement has been said in many studies to refer the influence made by source-based and managementbased factors [9][10][11][12]. The notion of brand relationship behavior has been referred in many studies to examine the impact of perceived effectiveness of celebrity endorsement towards endorsed brand with reference to different market contexts and product contexts [5,11]. Sri Lanka is noted as a significant context to examine how celebrity endorsement influences to perceived brand evaluation depending on industry significance and empirical gaps proposed by recently held studies Dissanayake [7], Dissanayake and Ismail [2], Sri Lanka Celebrity Index Report [1][2][3][4][5][6][7][8][9][10][11][12].

Research problem
As per the studies found in the domain of celebrity endorsement, it has a clear argument on examining the existing models with different product forms and market contexts to validate the empirical findings of recently held studies. Meanwhile, Sri Lanka has been referred as a valid context to further examining the impact of the effectiveness celebrity endorsement [1,2,7]. Those studies have validated the notion of examining effectiveness of celebrity endorsement towards the endorsed brand as a needful in addressing to research gaps in Sri Lankan context. Meanwhile, Becerra and Badrinarayanan [13], Doss and Carstens [14], Tzoumaka et al. [11] and Dissanayake [7] have referred brand-relationship behaviors as a notable concept to be examined with the effectiveness of celebrity endorsement. Further, Aaker [15] referred that brand personality as a concept of brand related property which is perceived to consumers. Meanwhile, the effect of celebrity endorsement into a particular brand has been explained by McCraken [16] introducing "Meaning Transfer Model". It says that the meaning of a celebrity endorsement derived by a celebrity transfers to endorsed brand via three stages of process. Firstly, meaning assorted to endorsement will move to the endorsed brand and then the endorsement converts the particular product into a specific brand personality whilst finally consumer engages with a consummation behavior with the endorsed brand. This has been further verified by Ohanian [17] saying familiarity, attractiveness, likability and similarity of the celebrity are the properties that influence the effectiveness of the endorsement for a brand. Additionally, Shimp [18] mentioned TEARS model as the determinants of the perceived effectiveness of celebrity endorsement. TEARS model refers Trust worthiness, Expertise, Attractiveness, respect and Similarity as the variables that drive the effective endorsement. This is in par with the source-based factors proposed by Seno and Lukas [8] as it describes factors related to the celebrity character. Therefore, it could argue to examine how those celebrities related factors which determine the perceived effectiveness of celebrity endorsement influence brand personality as a context of brand related behaviors [15][16]. According to the argument found in Meaning Transfer Model [16] celebrity has to be a congruent mode to create perceived brand personality referring to endorsed brand. This is again a verification for the research gaps postulated by Sri Lanka Celebrity Index Report [1], Dissanayake [7] and Dissanayake and Ismail [2] as it collectively denotes the need of examining brandrelated influenced by perceived effectiveness of celebrity endorsements. Alongside, this paper defines its research problem as "how perceived effectiveness of celebrity endorsement influences perceived brand personality" as addressing to the research gaps postulated in Sri Lankan context. As per the research problem rationalized, this paper has been based on the overall research questions of "What is the influence of Perceived Effectiveness of Celebrity Endorsement (PECE) on perceived Brand Personality (PBP) as a context of Brand-Related Behavior?" [13][14][15][16][17][18].

Objectives of the study
In line with the research problem and the niche of the research question, this paper intends to address the following research objectives.
• To examine the overall impact of Perceived Effectiveness of Celebrity Endorsement towards Perceived Brand Personality.
• To identify the comparative importance of the factors determine the Perceived Effectiveness of Celebrity Endorsement on Perceived Brand Personality.
• To postulate managerial implications to figure out effective celebrity endorsement-related practices to build perceived brand personality.

Significance of the study
This paper addresses the knowledge gaps postulated in Sri Lanka as to examine the post effect of celebrity endorsement-related practiced towards the endorsed brand. Accordingly, it has used the practicerelated evidences to figure out the key product categories namely Services and Fast Moving Consumer Goods (FMCG) to investigate the brand-related impact in the face of Perceived Brand Personality as one of the avenues to investigate brand related behaviors. Alongside, the findings of this study could be used as a decisional guide for the brand managers to figure out how different factors related to PECE could influence PBP. Accordingly, industry practitioners could demarcate the celebrity characters to be considered in building PBP as it requires for brand positioning strategies. Further, findings and conclusion of this paper could be a comprehensive guide for the future studies to review empirical evidences and research priorities.

Limitation of the study
This study is farmed to investigate the TEARS Model [18,19] as the source-Base factors in investigating perceived effectiveness of celebrity endorsement even though different theories and models are found in literature sources. Further, respondents were preserved for the services and FMCG sector brands by providing some qualifying questions in the questionnaire to respond for the questions. This study had to refer limited secondary sources and industry related insights to get expose for the practise significance of Sri Lanka as it finds less number of sources.

Celebrity endorsement
There are number of theories proposed by authors trying to explain how celebrity endorsement in advertising works. McCracken [16] has proposed a three stage meaning transfer model. It proposes that an endorser's credibility and expertise first transfer meaning to his or her public image; the endorser then transfers the meaning to the brand being endorsed. Finally, the meaning is transferred to consumers. Celebrity endorsement enhances the perceived quality of the brand and is correlated with intention to use [20,21]. Yoo and Donthu [21] defined consumer-based brand equity as a combination of brand loyalty, brand awareness, perceived quality of a brand, and meaningful memories linked to a brand. The credibility is a significant matter in brand related behaviors. It has found that positive correlation between the endorser's credibility and consumer based brand equity is mediated by the brand credibility [22]. According to McCutcheon et al. [23] relationships with celebrities can be classified into two types. Entertainment oriented individuals enjoy learning about celebrities and discussing the information with others casually, while intense attachment based individuals think continuously and associate Para socially with the celebrities. Hung et al. [24] found that entertainment oriented consumers superficially evaluate the meaning transfer involved in a celebrity endorsement; while intense attachment based consumers evaluate it carefully with detailed considerations. Another previous study delineated gender differences in forming attitudes toward celebrity endorsements. It emphasizes that Female consumers responded more favorably to celebrity endorsements than males, and they preferred female celebrity endorsers than males [25]. Celebrities can be an influential reference group acting as a powerful asset for marketers. Celebrities can give testimonials about the benefits of using a product, endorse a product, or act as a spokesperson for a brand for an extended period [26]. Marketers often choose celebrity endorsers who are attractive, credible, or who have expertise while at the same time matching up with the desired brand image [27]. Physically attractive celebrity endorsers can improve a brand's image and encourage consumers to purchase that brand [27]. Knowledgeable and skillful celebrity endorsers who attribute poses expertise in a particular area can make consumers more willing to purchase the advertised brand [20]. The trustworthiness of the celebrity is a supportive element as said in source credibility, but researchers found that it is not a factor in increasing consumers' intention to try a brand [20,27]. Conclusively, it refers that effectiveness of celebrity appeals depends on attractiveness, expertise, trustworthiness of the celebrities, and congruity between image of brand and the celebrity [20][21][22][23][24][25][26][27].

Determinant of the effectiveness celebrity endorsement
There are different models and theories found in the existing knowledge to review the factors that determine the degree of effectiveness of the endorsement made by celebrities. This paper reviewed some prominent notions that describe source-based factors and management based factors pertaining to the effectiveness of celebrity endorsement [8]. Meanwhile, studies found in Sri Lankan contexts, namely Dissanayake [7] and Dissanayake and Ismail [2] Wanninayake and Dissanayake [28] and Weerasiri and Wanninayake [12] have also used the term "effectiveness of celebrity endorsement" to refer the outcome of the endorsement resulted through managementbased and source-based factors. This is in line with the empirical justifications made by Amos et al. [9], Atkin and Block [10], Tzoumaka et al. [11]. As per the propositions mentioned by Seno and Lukas [8], perceived effectiveness could be proposed as mechanism of both management-based and source-based factors. Accordingly, the notion of source-based factors in celebrity endorsement could be explained in details as per the content suggested by Shimp [18][19] referring to TEARS model that determine effectiveness of the endorsement made by a celebrity. This model suggests "Trustworthiness" explained by honesty, integrity and believability, "Expertise" indicates for knowledge, skills and experience and "Attractiveness" described by physical look, personality and lifestyle as the first three determinant of the endorsement effectiveness [17]. Additionally, "Respect" is noted to indicate how celebrity is perceived to be esteemed due his/ her achievements. Finally, "Similarity" refers to matching the celebrity with gender, age, life style and other related matters. The contribution made by TEARS Model has been referred in different studies such as Folse et al. [29], Liu et al. [30], Okyere and Asamoah [31], Rachita et al. [32] and, Suprio and Kumar [33] as a valid verification to conceptualize the notion of effectiveness of celebrity endorsement [28][29][30][31][32][33].

Brand and brand personality
Brand is a name that influences buyers by explicating the benefits namely saliency, differentiability, intensity and trust [34]. The mental associations connected to the power of a name influences emotional relationships. Accordingly, brand is an attitude of nonindifference knitted into consumer's hearts [34]. Meanwhile, brand personality traits, defined as a "set of human characteristics associated with a brand" and it includes sincerity, excitement, sophistication, ruggedness, and competence [15]. Further, those are recognized as "an efficient way to distinguish a brand from its competitors at the symbolic level" [35]. Those sources concluded the thought on brand personality as something that effectively increase its brand equity via consistent set of traits. This is the added-value that a brand gains, aside from its functional benefits. Sri Lank is a noted context for sport sector celebrities being used in brand building strategies including advertising. The notion of how sports management connects to brand personality has been examined in different studies [36,37]. As initiated by Aakerr et al. [38] and Aaker [15], it was able to construct a Brand Personality Scale [15]. However, the said concept of brand personality has been still found as relatively a less studied context. Further, it has referred that brand personality has been an important concept to be examined with celebrities as it finds no enough studies carried out [39]. Arsena et al. [40] referred the contribution of Aaker [15] to examine how celebrities' traits transfers to brand personality. Relationship between brand personality and the sport sector celebrities in the forms of teams is noted in the study of Carlson et al. [41]. Meanwhile, Dissanayake [7] emphasizes that brand evaluations and brand relationship behaviors resulted via perceived effectiveness of celebrity endorsements need to be further examined in Sri Lankan context as it notes a significant trend in the practice perspectives. Brand personality has also been recognized as a concept related to the brand-related behaviors in many studies as complimenting to the argument of Dissanayake [7]. Therefore, it provides a valid notion to examine the concept of brand personality along with celebrity endorsement to investigate how celebrity-related factors, known as source-based factors [8] influence perceived brand personality [34][35][36][37][38][39][40][41].

Methodology
This study follows the inductive approach as it tests the existing models and empirical thoughts in the context of Sri Lanka in par with the research gaps and knowledge-related validity as brief earlier. Accordingly, it used convenient sampling methods to do the data collection from 160 respondents who represent socio-demographic diversities (152 responses were considered in the final analysis). Structured questionnaire was used as the data collection instrument since it selected survey method as the research approach. Data analysis was supported by descriptive and inferential statistical sources. Findings have been presented at the end of the paper along with the managerial implications. Further, data were refined with the testing of reliability before analyzing with inferential statistical tools. According to Hair et al. [42] when repeated application of a survey instrument finds in consistent scores, it could consider as reliable saying reliability is concerned with the consistency of the research findings. This study also adopts Cronbach's alpha to measure the reliability of the items. Cronbach alpha is the most commonly used measure for evaluating the reliability of survey instruments. It was referred that the Cronbach's alpha of around 0.70 is normally acceptable in exploratory research [43] suggesting Cronbach alpha of between 0.7 and 0.98 is "highly reliable" and that anything below 0.35 should be rejected. We used SPSS-20 software package to assist data analysis and charts were used to present the data. Descriptive statistics were done for the calculation of mean, standard deviation and variances for the independent and dependent variables. The correlation test was applied to determine the relationship between the brand personality and the determinant of the effectiveness of celebrity endorsement as said in TEARS model [18,19]. Regression was used to identify the strength of the relationship between two variables and ANOVA test was conducted identifying the significance of the relationship in testing hypotheses [42,43].
In line with the foresaid literature based review, the independent variables are framed under the concept of Perceived Effectiveness of Celebrity Endorsement and the dependent variable is consistent with the concept of Brand-Related Behavior. Accordingly, it has operationalized the variables and indictors followed by deductive approach. Variables of effectiveness of celebrity endorsement were operationalized with TEARS Model [18,19]. The items for each variable mentioned in the literature review were tested in the questionnaire. Perceived brand personality was operationalized based on the five dimensions namely Sincerity, Ruggedness, Sophistication, Competence and Excitement [15] (Figure 1).

Hypotheses of the study
According to the objectives of the study and the niche of the research question stated, it has tested the following hypotheses in line with the variable relationships found in the conceptual framework.
H1: There is a significant relationship between Perceived Trustworthiness of the celebrity endorsement and the Perceived Brand Personality.
H2: There is a significant relationship between Perceived Expertise of the celebrity endorsement and Perceived Brand Personality.
H3: There is a significant relationship between Perceived Attractiveness of the celebrity endorsement and Perceived Brand Personality.
H4: There is a relationship between Perceived Respect related to celebrity endorsement and Perceived Brand Personality.
H5: There is a relationship between Perceived Similarity of the celebrity endorsement and Perceived Brand Personality.

Reliability of the variables
A reliability test was carried out for the variables and the results are shown in the below table.
It indicates that Cronbach's alpha values are more than 0.7 in each variable which would either be independent or dependent. Thus, there is a consistency in the responses found in the date set [42][43][44] (Table 1).

Profile of the sample
Gender: Sample includes a ratio of 52% of male respondents while 48% of the total sample was females. Above figure showcases a representation of the population of the study. Perhaps this can be explained as males are little more exposed than females about celebrity endorsements taking place currently.

Education Level
In terms of the education levels of the selected sample, 48% of the sample population was graduates and 29% was diploma holders. Only 4% was having an education only up to O/Ls and 19% have qualified in A/Ls.

Income distribution
In the income distribution perspective, majority was included in the 20,000-40,000 (Sri Lankan Rupees) category reporting 42%, and 25% of the respondents were in 41000-60000 category while 23% were in the above 80,000 category. Only 10% of the respondents represent less than 20000 category.

Descriptive statistics: central tendency of main variables
It has tabulated the mean values of the independent and dependent variables to approach an idea on how data has been found around the central tendency. As it indicates in the summarized table, almost all the variables are found as above averaged level in the five point Liker scale whilst a notable mean value is found in the dependent variable (perceived brand personality dimensions). Trustworthiness has been found as most significant variable in terms of mean value compared to the rest of four independent variables ( Table 2).

H1: There is a significant relationship between perceived trustworthiness of the celebrity endorsement and the brand personality
The SPSS output tables have been given below prior to the explanations on correlation values and regression values.
The above table shows that there is a significant positive relationship between trustworthiness and brand personality by expressing nearly 31% of relationship between the variables. Further, it has given below the ANOVA table to conclude the variable relationship in line with H1. According to table, the R is the correlation between two variables which is same as correlation analysis value of 0.308. And most important part in the table is that R square which indicates that the Trustworthiness only have 30% of variation in brand personality and other 70% of variation in brand personality are explained by other factors related to brand personality. In ANOVA, the main concern is about significance level. Since the significance level is less than 0.05, it is considered as statistically significance relationship between two variables. According to coefficient table, constant is 2.348, when trustworthiness is zero, still Brand Personality has a value of 2.348 in statistic terms. Then trustworthiness increases by one unit, the brand personality will increase by 0.412. According to B coefficient table, this is positive moderate relationship since it is between +0.3 and +0.7. Finally, P value is 0.001 which is less than 0.05, it shows that the statistical significance relationship between two variables and therefore hypothesis (H 1 ) is accepted (Tables 3-6).

H2: There is a significant relationship between perceived expertise of the celebrity endorsement and brand personality
It has presented the statistical output tables for the correlation values and ANOVA as visual presentation for the explanations on hypotheses testing. The above table shows that there is a positive relationship between expertise and brand personality of 35% of explanation at a significant level. However, H 2 has been tested as per the statistical verification of the ANOVA given below. According to table, the R is the correlation between two variables which is same as correlation analysis that indicates 0.356 and most important part in the table is that R square which confirms that the Expertise only have 35% of variation in brand personality and other 65% of variation in brand personality is explained by other factors. Besides, referring to ANOVA, the main concern is to comment on the significance level. Accordingly, it reports the significance level as less than 0.05 justifying as statistically significance relationship between two variables is established. According to coefficient table, constant value indicates as is 2.316. Therefore, when expertise is zero, still it is found Brand Personality has a value of 2.316 in statistic terms. Then expertise increases by one unit, the brand personality will increase by 0.41. According to B coefficient table, this is reported as a positive moderate relationship since it is between +0.3 and +0.7. Finally, P value is 0.000 which is less than 0.05 that allows to accept as statistical significance relationship between two variables and therefore hypothesis (H 2 ) is accepted (Tables 7-10).

H3: There is a significant relationship between perceived attractiveness of the celebrity endorsement and brand personality
The SPSS outcome tables have been given below with reference to the variable relationships of H 3 . The above table shows that there is a significant positive relationship between Perceived Attractiveness and Brand Personality by 39% of explanation. However, conclusion on the H 3 has been made by referring to ANOVA table statistics given below. According to foresaid ANOVA table, the R is the correlation between two variables which is the same as correlation coefficient found as 0.39. However, the value of R square which indicates that the attractiveness only has 39% of variation in Brand Personality whilst rest of 61% of variation in brand personality are explained by other factors. With reference to the significance level related to ANOVA, it has found a significance level in terms of the relationship between perceived attractiveness and brand personality since P value is less than 0.05. According to coefficient table, constant value is 2.493 saying when attractiveness is zero, still Brand personality has a value of 2.493 in statistical terms. In line with the said statistical results, when attractiveness increases by one unit, the brand personality will increase       by 0.442. According to B coefficient table, this is a sign of positive moderate relationship since it is between +0.3 and +0.7. Finally, since P value is 0.000 reporting less than 0.05, it could conclude that there is a statistical significance relationship between two variables and therefore hypothesis (H 3 ) is supported (Tables 11-14).

H4: There is a relationship between perceived respect related to celebrity endorsement and brand personality
The SPSS outcome tables referring to the variable relationship on H4 had been given below for the snapshot review. It has given the explanations to conclude H4 accordingly. The above table shows that there is a positive relationship between Perceived Respect and Brand Personality at the level of 15% of explanation. However, it does not statistically verify the relationship as statistically significant. However, conclusive explanation on H4 has been given below with reference to the outcome of ANOVA table. According to the statistical presentation, R is the correlation between two variables which is same as correlation analysis value of 0.15. According to the R square, it indicates that Perceived Respect only has 15% of variation in Brand Personality whilst the rest of 85% of variation in brand personality is explained by other factors related to brand personality. Meanwhile, it shows that the P value is more than 0.05 saying the relationship is statistically not significant as denotes in correlation coefficient table. According to coefficient table, constant is 3.432, when respect is zero, still Brand personality has a value of 3.432 in statistic terms. It refers that brand personality will increase by 0.116 when there is an incensement of perceived respect by one unit. However, as indicates in B coefficient table, this is a positive weak relationship since it is between +0.1 and +0.3. Finally, P value is also found as 0.053 which is more than 0.05 showing that there is no statistically significance relationship between two variables and concluded H 4 is not supported (Tables 15-18).

H5: There is a relationship between perceived similarities of the celebrity endorsement and brand personality
It has presented the SPSS outcome tables for the correlation analysis and ANOVA as follows, and the respective explanations have been made on H5 accordingly. The above table shows that there is a positive relationship between respect and brand personality with 34% of explanation, and the relationship is noted as statistically significant. Moreover, it has explained the outcome of ANOVA in the following content to conclude the H5. According to table 5.6.9.1 the R is the correlation between two variables which is same as correlation analysis; 0.34 and most important part in the table is that R square which indicates that the perceived similarity only has 34% of variation in brand personality and other 66% of variation in brand personality are explained by other factors related to brand personality. Meanwhile, in ANOVA, the main concern is about significance level. Since the significance level is less than 0.05, it is considered as statistically significance relationship between the said two variables. According to coefficient        similarity increases by one unit, the brand personality will increase by 0.342. According to B coefficient table, this is a positive moderate relationship since it is between +0.3 and +0.7. The indicated P value is 0.000 which is less than 0.05 that shows there is a statistical significance relationship between two variables, and therefore, H5is supported (Tables 19-22).

Conclusion of the Key Findings
As per the analysis and the discussion presented in the foresaid content, it has given the summery of the statistical analysis for a concluded overview in Table 23. From the results of the survey researcher can identify Trustworthiness of the endorser has a positive moderate relationship with the brand personality. Same scenario is relevant for the variables, Expertise and Attractiveness. However, reputation of the celebrity is having a weak positive relationship with the brand personality. Similarity is also having a positive moderate relationship with brand personality. According to the overview of this study, perceived effectiveness of celebrity endorsements in building perceived brand personality seems an effective approach besides the notion of "Perceived Reputation". As per the TEARS model [18,19] and the Source-Based Factors proposed by Seno and Lukas [8] perceived effectiveness of celebrity endorsement is determined by celebrity character-related factors. However, overall effectiveness of celebrity endorsement could also be determined by the combination mechanism of Management-Based Factors [7] that complement the effectiveness of Source-Based Factors. However, this study reveals Perceived Attractiveness, Perceived Trustworthiness and Perceived Expertise as major notions that determine the perceived effectiveness of celebrity endorsement resulting brand personality (Table 23).

Managerial implications and suggestions for further research
This study addresses the research gaps in Sri Lankan whilst focusing to practice related significance as to contribute new knowledge to practitioners. The application of celebrity endorsements into different consumer behavioral contexts had been referred in the Sri Lankan studies by emphasizing the extended notions for the future studies [2,12,28,45,46]. Accordingly, findings of this study could be used to different business sectors in planning and implementing celebrity endorsements to build brand related performance. Moreover, as it indicates in the findings, Trustworthiness, Attractiveness, Expertise and Similarity are found to be much effective qualities to be considered in selecting a celebrity to make the endorsement to build brand personality. It is strategically important to match the product personality and the personality of the celebrity character to make the endorsement an effective. Meanwhile, it proposes to manage the source-based factors alongside the management-based factors to make the holistic effectiveness of celebrity endorsement to build brand personality. Future research works are encouraged to measure the combined effect of source-based factors and multiplicity or over endorsement effect to determine perceived effectiveness of celebrity endorsement [47]. This is to refine the unconcluded argument on how over endorsement does influence to overall effectiveness of the endorsement [24]. Moreover, we suggest investigating the role celebrity-consumer psychological bond in patterning consumer's responses on endorsed message to evaluate brands leading to brandrelated behaviors. The celebrity worship motives postulated by         McCutcheon et al. [23] and Hung [5] could be hypothesized to examine their intermediary influence on brand evaluation leading to brandrelated behaviors. Further, practitioners could consider the notions of similarity to be matched with the target market profile in determining effective results for the endorsed brand. Since, Sri Lanka is a multicultural context noticed with socio-demographic dynamics, it needs to carefully manage celebrities in par with the profiles of target markets to make the endorsement much effective. Apart from the said, celebrity endorsements should be uniquely managed as per the industry or sector, for instance, services brand needs special attention in this regard depending on its nature and the unique trend noticed in Sri Lankan economy specially in the telecommunication, insurance and financial services [1]. As it denotes in Sri Lankan context, services sector brands intensively depend on celebrity endorsement practices but empirical researches are found as lag to contribute for the knowledge and practice related uses. As it suggests earlier, future research works are admired to examine the different facets of brand relationship behaviors including customer brand relationship (CBR), brand evangelism and brand attachment with celebrity endorsement to contribute for the empirical knowledge. We admire multiplicity [48][49][50] or over endorsement [24] as a noted case to be examined with brand evaluation and brand-related behaviors in the subcontinent market context. Sri Lanka, India, Bangladesh and Pakistan are intensive cases of celebrity endorsements with multiplicity. For an instance, India is a special case to study the multiple endorsement as it notices huge trend amongst cricketers and cinema starts endorsing many brands which may lead different perceived evaluations and brand relationship behaviors. The effectiveness of the endorsements attributed by local celebrity on multinational brands could be investigated as another special case to contribute for the empirical knowledge. Addressing to the life style dynamics, it suggest to extend the studies investigating how consumers respond to celebrity endorsements within the digital media context whilst admiring researches to compare the responses of X generation and Y generation within the context of celebrity endorsements [45][46][47][48][49][50][51][52][53][54][55].