Self-congruity theory in tourism research: A systematic review and future research directions

This study aims to generate future research directions by discovering how self-congruity theory has been applied in tourism research so far. The systematic literature review method was employed and 53 articles applying selfcongruity theory published in 20 top-tier tourism journals between 1992-2019 were content analysed. Findings revealed that the most prominent journal in publishing self-congruity research was the Journal of Travel and Tourism Marketing, most of the relevant articles followed a quantitative research design and utilized the new method of measuring self-congruity which obviates the discrepancy scores problem of the traditional method by measuring self-congruity directly, actual self-congruity was the most considered self-congruity dimension and the theory was investigated mostly with regards to destinations. Future directions for self-congruity researchers were generated according to these parameters.


Introduction
Self-congruity theory defines the psychological process and outcome in which consumers compare their perception of a brand personality or brand-user image with their own actual, ideal, social and/or ideal social self-concepts (Sirgy, 2018). The theory suggests that the greater match between the brand image and consumer's self-concept positively influences the consumption behaviour and behavioural outcomes such as customer loyalty, brand trust, positive word-of-mouth communication, etc (Sirgy, 2019). Because consumers regard the products/brands they purchased as an expression of their own selves (Graeff, 1996;Sirgy, 1985), self-congruity also plays an important role in value creation (Sirgy et al., 2018).
For more than 50 years, self-congruity theory has been applied extensively in research related to consumer behaviour and marketing (Sop, 2017). Previous studies revealed that consumers' images of automobile brands were the first and most examined research question in developing this theory. Birdwell (1968), for example, investigated the perceived congruence between automobile owners' selfimage and brand image of the cars they bought. He concluded that "an automobile owner's perception of his car is essentially congruent with his perception of himself" (Birdwell, 1968: 78). Meanwhile, the theory was also being applied and tested in studies regarding cigarettes, alcoholic beverages, cosmetics, toothpaste, shoes and stores to understand if consumers choose brands which reflect who they are. In 1990, Kye-Sung Chon extended these research contexts by defending his dissertation related to destination satisfaction in which self-congruity theory was applied for the first time within the tourism discipline (Chon, 1990). Chon & Olsen (1991) and Chon (1992) published the pioneering studies applying self-congruity theory in the destination context. Sirgy et al. (1997) followed them in their research by discussing the traditional method of measuring self-congruity, which suggests calculating the gap score between the image of the destination and the self-image of the tourist, and developed a new method that allows measuring self-congruity directly instead of relying on discrepancy scores. Sirgy & Su's (2000) conceptual paper introducing an integrative model that consists of self-congruity, functional congruity and travel behaviour also stimulated researchers to apply self-congruity theory in tourism (Kılıç & Sop, 2012). Sirgy & Su's (2000) self-congruity model proposes that the greater match between the self-concept (or self-image) and the destination image, the greater the possibility that tourists feel motivated to visit the destination. Besides self-congruity, functional congruity, described as the level of match between tourists' expectation toward the functional attributes of a destination (cultural heritage, leisure activities, amenities, etc.) and their perceived performance outcome, also motivates tourists to travel to a destination (Chon & Olsen, 1991;Sirgy & Su, 2000). Within this context, Sirgy and Su (2000) assert that self-congruity has a biasing effect on functional congruity. In other words, tourists who perceive a high level of self-congruity will have positive first impression toward the destination and this may also tend to impact the functional attributes of the destination positively (Sirgy, 2014;Sirgy & Su, 2000). Hence, self-congruity increases the value perception of tourists, which in turn increases their positive attitude, preference, choice, satisfaction, loyalty and word-of-mouth communications . Since self-congruity is effective in explaining tourists' pre-visit and or post-visit behaviours regarding destinations (Xu & Pratt, 2018), hotels (Sop & Kozak, 2019), restaurants (Jeong & Jang, 2018), museums (Wang & Lu, 2011), etc., it can be understood as an important theory for travel and tourism marketing.
The purpose of this study is to discover how self-congruity theory has been applied in tourism research after the pioneering self-congruity studies. Although there are some previous relevant studies reviewing the consumer behaviour and marketing literature (Aguirre-Rodriguez et al., 2012;Sirgy et al., 2000), a systematic literature review of self-congruity related tourism research is needed to reveal the current interest of the researchers and suggest future research directions. Twenty top-tier tourism journals ranked by Scimago were examined and 53 articles applying the theory were content analysed for this purpose. It is believed that the results of the current study will contribute to the self-congruity theory literature by suggesting new directions for future tourism research.
In the following section of the study, self-congruity theory is introduced briefly within different perspectives to generate research questions which are later discussed as the research parameters. Then, the research methodology of the study is explained and findings are presented. Finally, future directions for self-congruity researchers are discussed.

Self-congruity (SC) theory Self, self-concept and SC
The self is basically expounded as the sum total of all that an individual can call his/hers, including his/her physical characteristics, house, wife/husband, friends, reputation at work, etc. (James, 1950). According to Grubb & Grathwohl (1967), the self also consists of the individual's perceptions, attitudes and evaluations of himself/herself as an object. These approaches clearly indicate that the self is a strong cognitive element directing human behaviour. For this reason, a notion called self-concept (self-image) was established by theorists considering the discussions toward what the self is (Han, 2006). Consequently, Rosenberg (1989) defined self-concept as the "totality of the individual's thoughts and feelings with reference to [the] self as an object" (Rosenberg, 1989: 34). More simply, Mehta (1999) described this notion as a person's perception of himself/herself.
Theorists proposed that the self-concept is a multi-dimensional construct (Landon, 1974;Sirgy, 1980;1982;Markus & Nurius, 1986;Malhotra, 1988). Supporting that paradigm, Sirgy (1980Sirgy ( , 1982 developed a four-dimensional approach to describe self-concept (or self-image) in consumer behaviour: "Actual self-image [how consumers see themselves], ideal self-image [how consumers would like to see themselves], social self-image [how consumers believe they are seen by significant others] and ideal social self-image [how consumers would like to be seen by significant others]" (Sirgy & Su, 2000: 343). In accordance with his multi-dimensional self-concept approach, Sirgy (1982;1985) regarded SC (or selfimage congruity, self-congruence, self-image congruence) under four dimensions after he defined the term as the similarity between self-image of the consumer and the image of the brand or product (Sirgy, 1982;1985): "Actual SC [the congruence between the actual self-image and the brand image], ideal SC [the congruence between the ideal self-image and the brand image], social SC [the congruence between the social self-image and the brand image] and ideal social SC [the congruence between the ideal social self-image and the brand image]" (Sirgy, 1985: 195). In this regard, SC theory proposes that consumers tend to prefer the products/brands which they think have a similar image to their self-concepts because they consider the products/brands they purchased as an expression of their own selves (Graeff, 1996;Sirgy, 1985). Therefore, a consumer may be inclined to refuse to purchase a product/brand because it does not match his/her self-image or purchase the one that reflects himself/herself (Beerli et al., 2007;Sirgy, 1985). In this sense, the first research question (RQ) was developed as follows:

RQ1: Which dimension of SC has mostly been investigated in tourism research so far?
Measuring SC Sirgy et al. (1997) introduced two primary methods for measuring SC. The first one was called the traditional method which is grounded on tapping the individual's perception of the product/brand image and his/her self-image (Sirgy et al., 1997). In this method, the ratio or discrepancy score with each image dimension is mathematically computed [∑ | − | =1 that Pi is the rating of a product/brand image along image dimension i, and Si is the rating of self-image along dimension i] and the discrepancy scores are summed across all dimensions (Sirgy & Su, 2000). An example of the application of the traditional method is presented below. Chon (1990) asked the participants to imagine a typical tourist who visits Destination X and indicate their agreement or disagreement [-2= Strongly Disagree, +2= Strongly Agree] with the following statements (a short version): The traditional method can be implemented via Likert type scales as Chon (1990) or Sirgy et al. (1997) performed. However, the most widely used measurement method has been Osgood's semantic differential scale (Birdwell, 1968;Dolich, 1969;Malhotra, 1981;Mehta, 1999;Beerli et al., 2007;Hosany & Martin, 2012;Hung & Patrick, 2012). An example of this approach can be demonstrated briefly from Dolich (1969: 82): Applying the traditional method, Dolich (1969) calculated the congruence score for impulsive as [3 (selfimage) -1 (brand image) = 2 (difference/gap)]. The difference between the self-image and the brand image equals zero for simple, which indicates that the brand and the consumer are similar in terms of being simple. Dolich (1969) performed this formula for each scale item and reached the total difference score for SC. Although the traditional method has been widely accepted, Sirgy et al. (1997) criticized this method and identified three key problems methodologically (Sirgy et al., 1997): The use of discrepancy scores, predetermined images and a compensatory decision rule. To overcome the problems of the traditional method, Sirgy et al. (1997) asserted the new method of measuring SC. According to Sirgy et al. (1997), this method obviates the discrepancy scores problem by measuring SC directly. Furthermore, the new method does not use predetermined images and participants feel free while judging the brand image. The new method also copes with the compensatory decision rule issue by presenting a short scenario to participants that guides them to evaluate SC through the brand's userimagery instead of an image dimension of the brand. The new method for measuring four dimensions of SC can be exemplified as follows (Sirgy & Su, 2000: 350).
Take a moment to think about [destination x]. Think about the kind of person who typically visits [destination x]. Imagine this tourist in your mind and then describe this person using one or more personal adjectives such as classy, poor, stylish, masculine, sexy, old, athletic, or whatever personal adjectives you can use to describe the typical visitor of [destination x]. Once you have done this, indicate your agreement or disagreement to the following statement (1-Strongly Disagreement, 5-Strongly Agreement): Since measuring SC has been discussed for more than 20 years, the following RQs were developed to discover the current position of tourism researchers toward these measurement approaches.

RQ2: Which SC measuring method has been used predominantly in tourism research? RQ3: It is obvious that both methods of measuring SC were developed only for quantitative research design. Then, what has been done in qualitative research?
Application of SC in consumer behaviour and marketing SC theory has been referenced in marketing and consumer behaviour research since the 1960s. The theory was investigated with regards to automobiles (Birdwell, 1968;Grubb & Hupp, 1968), cigarette, beer, soap and toothpaste brands (Dolich, 1969), color TV, adult games, coffee, country club membership and dress shirt (Landon, 1974), perfume (Mehta, 1999), precious jewellery (Jamal & Goode, 2001), supermarkets (Ibrahim & Najjar, 2008;Willems & Swinnen, 2011), retail stores (Das, 2014), handbags (Zhu et al., 2019), fast-food restaurants (Islam et al., 2018), newspapers (Kim, 2018) and numerous research contexts. The research variables linked to SC in those studies have generally been brand personality, functional congruity, brand attitude, brand preference, satisfaction, behavioural intentions and loyalty. These studies, which have been carried out on various samples (students, gamers, automobile owners, etc.) and published in a variety of top-tier journals, have revealed that SC theory is highly effective in explaining consumer behaviours. Therefore, the following RQs were generated to discover how these parameters of SC seen in consumer behaviour and marketing literature have been reflected in tourism research:

Research methodology
This study aims to reveal how SC theory was applied in tourism research, and to generate future directions for tourism researchers. For this purpose, the systematic literature review (SLR) method was applied based on previous studies (Bueno et al., 2019;Gómez et al., 2019;Mustak, 2014;Prayag et al., 2019). Following the stages of the SLR method, as Mustak (2014: 152) suggested, "extant publications were located, analysed, and evaluated in relation to the research questions posed, and findings have been presented in such a manner that enables those questions to be answered in a comprehensible and concise manner". In Stage 1, publication databases were selected and an initial search was performed. Scimago Journal Rank (SJR) was used in this stage to list the top-tier tourism journals which have higher scores in terms of various bibliometric indicators compared to others, such as journal impact factor, h-index or SJR score. SJR score [a measure of journal's impact, influence or prestige. It expresses the average number of weighted citations received in the selected year by the documents published in the journal in the three previous years (Scimago, 2019)] was considered in determining and ranking the top-tier journals of Tourism, Leisure and Hospitality Management. This stage of the SLR method was performed in March 2019, and top-tier tourism journals were limited by the first 20 (see Table 1). In Stage 2, each journal's publisher website (Elsevier, Emerald, Sage, Wiley, Taylor & Francis, etc.) was visited to specify the articles applying SC theory. First, advanced searches were performed for each journal in "title", "keywords" and "abstract" by entering the terms "self-image congruency, self-image congruence, selfimage congruity, self-congruence, self-congruity" and some other variations like "self, self-image and self-concept". These terms have been selected by investigating the relevant literature. The literature indicates that some researchers use different terms instead of self-congruity, such as self-image congruency (Sparks et al., 2011), self-image congruity (Jeong & Jang, 2018), self-image congruence (Hosany & Martin, 2012;Jamal & Goode, 2001) or self-congruence (Ahn et., 2013). The advanced search resulted in 57 articles, but four of them were identified as irrelevant articles and eliminated in the content analysis process. Consequently, the sample of the study consisted of 53 articles published in 20 different top-tier journals. In Stage 3, the research sample (53 articles) were content analysed in terms of seven parameters highlighted as the RQs of the current study, which were determined after an indepth literature review both in marketing and tourism fields: 1) considered SC dimensions, 2) applied methods for measuring SC, 3) research designs (qualitative, quantitative or conceptual research), 4) research contexts (destination, hotel, etc.), 5) sampling methods and sample size, 6) linked research variables, 7) the prominent journal in publishing SC research. The content analysis was performed on each article's abstract and method section. 53 articles were carefully reviewed by the author and the keywords related to the RQs extracted to a table (see the Appendix for Table 2). Finally, in Stage 4, the RQs were answered to present the current state of SC theory in tourism research and discuss guidelines for future research.

Findings and discussion
Top-tier tourism journals were ranked according to SJR scores in  Chon (1992) who had already defended his dissertation related to this theory (Chon, 1990). However, the figure also exhibits a seven-year period when SC theory was not applied in any tourism research. Sirgy and Su's (2000) study is the first article published after Chon (1992), but there are some years (2005, 2006, 2008 and 2009) that SC related articles were not published. Between 2010 and 2019, at least two articles applying SC theory were published every year in top-tier tourism journals. It is also obvious that SC theory has been widely applied in tourism research in the last five years because 56.6% of the SC related articles were published between 2015-2019. It proves the fact that SC theory has been recently re-discovered by tourism researchers although it could be seen as an outmoded research topic due to the fact that its background dates back to the 1960s in marketing literature. By considering the parameters mentioned in the previous section as RQs, 53 articles published in those journals were content analysed, one by one. The results of the content analysis were presented in Table  2 (see the Appendix). Findings reveal that only quantitative research methods have been applied in SC studies so far. Nevertheless, there are three exceptional studies related to this discussion. Correia et al. (2016) used qualitative and quantitative approaches together in their research, but they utilized SC theory to strengthen their theoretical background in explaining the conspicuous consumption of elite individuals, instead of measuring the congruence between self-image and destination image. Sirgy & Su (2000) and Sirgy (2019) are the other exceptions because these studies are conceptual, not empirical. In this sense, Sirgy & Su (2000) who broke the seven-year-silence after Chon (1992) (see Figure 1) developed an integrative model including SC, functional congruity and travel behaviour. Their study intensified the position of SC theory in tourism research. In his latest article, Sirgy (2019) discussed SC, selfexpressiveness, need hierarchy, goal, leisure benefits, broaden-and-build and bottom-up spillover theories to promote quality-of-life and well-being research in hospitality and tourism.

Figure 1. Distribution of SC Articles in Top-Tier Journals by Years
In addition to Table 2, Figure 2 focuses on the research contexts that were studied in the 53 articles and reveals that SC studies were conducted mostly on destinations (n= 24, 45%) and hotels (n= 7, 13%). Across the 53 articles, only Andreassen & Lanseng (2010) employed SC theory within the organizational behaviour perspective to investigate the image congruence between company and employee. These results prove that more SC research in different contexts (museums, theme parks, festivals, etc.) is needed to consolidate the theory. Furthermore, SC theory can be extended with more research on the organizational behaviour context by following Andreassen & Lanseng (2010).  Table 2 also reveals an interesting finding related to the sampling methods applied in SC research. According to the content analysis results, only 17% (n= 9) of the empirical SC studies in tourism employed probability sampling methods, such as random, systematic, cluster and stratified random sampling. On the other hand, convenience sampling was the most utilized sampling method in the respective tourism studies (n= 19, 37%). Moreover, Zhang et al.'s (2017) study stood out amongst the other articles with its huge sample size that was 15,377. However, the authors discussed SC theory only for developing the theoretical framework of their research, which explored how international tourist flows are influenced by Hofstede's cultural dimensions. Accordingly, the average sample size of the empirical tourism studies applying SC theory is calculated as 892 and the median sample size is 403. When the articles in which SC theory was not directly utilized are excluded (Palmer et al., 2013;Correia et al., 2016;Zhang et al., 2017;Tung et al., 2017), the average sample size decreases to 639, while the median sample size increases to 410.

. Distribution of Research Contexts in Relevant Articles
Since measurement methods and dimensions of SC have been discussed for years, these two parameters were also analysed in the current study. The findings also reveal that actual self-congruity was the most discussed dimension of SC theory in tourism, within 42 empirical studies (82%). Secondly, ideal SC was considered in 29 studies (57%). Although social SC (n= 13, 25%) and ideal social SC (n= 12, 24%) were investigated in some studies, most of them did not directly discuss these two dimensions. Finally, the research variables linked to SC in empirical tourism studies were analysed and findings were also presented in Table 2. In addition, Figure 3 was

Conclusion and future directions
In the current study, 53 articles applying self-congruity theory published in 20 top-tier tourism journals ranked by Scimago were examined by employing the SLR method. The findings illustrated that most of the self-congruity related articles were published between 2015-2019, which demonstrates that SC theory has recently been re-discovered by tourism researchers. In order to explore the current position of SC theory in tourism research and to generate future research directions, the research parameters which were previously listed as RQs were answered in Table 3 and discussed as follows. The convenience sampling method, the average sample size is 892 and the median sample size is 403.

RQ6: Which research variables have been linked to SC in tourism research?
Behavioural intentions, brand personality, consumer/brand attitude, functional congruity, satisfaction and loyalty.

RQ7:
Which top-tier tourism journal has published more articles applying SC theory?

Journal of Travel and Tourism Marketing
In regard to RQ1 it was found that although four dimensions of SC have been investigated in tourism research so far, actual SC has been the most discussed dimension of SC theory in tourism, which is followed by ideal SC. This finding proves that the social aspect of SC theory has been neglected by tourism researchers so far. Therefore, tourism researchers should investigate the social and the ideal social dimensions of SC for understanding the broader effects of image congruity between brands and consumers/tourists. Thus, their behaviours toward the product/brand can be analysed in terms of social consistency and social approval motives as Sirgy (2014) proposes.
The result of RQ2 revealed that the new method has been the predominant measurement method in empirical tourism research applying SC theory. On the other hand, there are numerous studies which were published after the 2010s in top-tier journals using the traditional method (Boksberger et al., 2011;Kastenholz, 2010;Khoo-Lattimore & Prayag, 2016;Hung & Petrick, 2011;Shu et al., 2015;Xu & Pratt, 2018), although Sirgy et al. (1997) and Sirgy and Su (2000) proposed that the traditional method has three key problems (the use of discrepancy scores, predetermined images and a compensatory decision rule). Thus, it is believed that comparative research should further re-explore the advantages and disadvantages of these methods within the context of tourism.
In regard to RQ3, it was discovered that only quantitative research methods have been applied in SC researches so far, except for Correia et al. (2016). Obviously, there is a huge lack of qualitative research applying SC theory in tourism. To find out the unrevealed aspects of SC on tourist behaviours and support Correia et al.'s (2016) insight, more qualitative studies are needed in the tourism field of research.
The result of RQ4 demonstrated that SC has mostly been investigated with respect to destinations, followed by hotels. This suggests that SC theory should also be applied to various tourism research contexts besides destinations or hotels; such as museums, theme parks or events. In addition, the applications of the theory beyond the consumer behaviour context should be considered. For example, employee behaviours in organizations can be investigated via SC theory as the pioneering study of Andreassen & Lanseng (2010) proposes.
Concerning RQ5, the average sample size of the empirical tourism research applying SC theory is 892 and the median sample size is 403. Convenience sampling has been the most used sampling method in relevant tourism research so far. Statisticians claim that probability sampling methods should be followed in empirical studies to produce more powerful findings. For example, Blalock (1972: 510) states that "with nonprobability sampling, we may obtain a very representative sample, but we shall not be in a position to evaluate the risks of error involved". For that reason, implementing probability sampling methods (such as simple, systematic, cluster and stratified random sampling) instead of nonprobability ones are suggested for future SC related studies. Conducting more research on samples that represent the whole research population may strengthen the theory itself. For this purpose, online surveys can be implemented to customers who are selected via probability methods from the databases of hotels', restaurants', etc.
About RQ6, it was found that SC has been investigated by linking mostly with behavioural intentions, brand personality, consumer/brand attitude, functional congruity, satisfaction and loyalty. This supports the integrative model of Sirgy & Su (2000) consisting of self-congruity, functional congruity and travel behaviour. To broaden these variables, the role of SC in tourist-tourist, tourist-resident and tourist-employee interactions should be investigated in future research. The residents' attitudes toward tourism should be examined by applying SC theory, as well. In addition, considering SC theory in an organizational behaviour perspective as Andreassen & Lanseng (2010) implemented for the first time, more research variables can be investigated such as the intention to leave the organization, organizational commitment or employee attitudes toward the organization. The result of RQ7 indicated that the Journal of Travel and Tourism Marketing has published most of the SC articles in tourism so far. Tourism Management takes the second place in the relevant journal list. Furthermore, all of the top-tier tourism journals except for TG, CHQ, Cities, LS, AG and JST have published at least one SC research study. This may indicate that a special issue on SC theory should be considered by the editors of the top-tier tourism journals in order to maintain the current interest toward SC theory which was started by Chon, Olsen and Sirgy in the early 1990s.
Besides these contributions to the existing literature, this study has also some limitations which need to be highlighted. First, the sample consisted of 20 top-tier journals with 53 articles which may be enhanced by future studies. Second, the research variables linked to SC in the 53 articles were reported in this study to explore the frequently considered variables in most of the top-tier articles. Future research may also more deeply analyse those studies by reporting which variables significantly affect SC, or are significantly affected by SC. Although this study reveals how SC was applied in tourism, a quantitative meta-analysis advancing the findings of the current study (by considering the moderator or mediator variables of the SC models tested in previous studies, etc.) may enable a closer and deeper look at SC theory in tourism research Co= Conceptual, Ql= Qualitative, Qn= Quantitative, Tr= Traditional, Nw= New, SC= Self-Congruity, ASC= Actual Self-Congruity, ISC= Ideal Self-Congruity, SSC= Social Self-Congruity, ISSC= Ideal Social Self-Congruity, BP= Brand Personality, FC= Functional Congruity. *Because it was not directly mentioned in the related study, the sampling method was interpreted by the author himself. **In cases that SC dimensions were not mentioned directly, the author interpreted SC dimensions by considering scale items used in the related study.