Countries’ positioning on personality traits

Abstract

This article discusses a change in the positioning strategies by countries as tourism destinations from cognitive components to affective ones. The tourism literature recognizes destination personality as an affective evaluation. While previous literature has established that place personality traits are important for Destination Marketing Organizations to consider, little empirical work assessing the personality metaphor in destination branding has been undertaken. This study uses a multistage methodology using a combination of content analysis and correspondence analysis to analyze the use of the personality metaphor in the official English tourism websites of South American countries. Results reveal that South American countries can be classified in four main clusters of different personality profiles: Ecuador is somehow associated with agreeableness; Brazil and Paraguay with openness to experience and conscientiousness; Peru, Chile, Argentina, and Uruguay as strong in extraversion and emotional stability; and Bolivia, Venezuela, and Colombia do not display any dominant personality characteristics. Overall the results indicate that South American countries are not yet using in a substantial way the affective cues of personality traits to create a clear positioning among international tourists.

Keywords

Big Five content and correspondence analysis destination branding destination personality DMO positioning South American tourism websites

Introduction

The advent of the 21st century has given rise to the study of place personality. Since 2000, there have been approximately 100 articles published that are related to place personality, among which most are specifically focused on destinations. Recent research has stated that place personality traits can convey rich meaning (i.e. symbolic and affective cues) to potential travelers to tourism destinations (Ekinci and Hosany, 2006; Zouganeli et al., 2012). Similar to individual personality traits, place personality traits reflect enduring characteristics of a location and its associated culture. In this context, Hankinson (2004) developed the “relational network brand model” where personality is listed as one of the three elements of the core brand. Place personality traits can be used to address the criticism that positioning purely based on place functional attributes (such as climate and geography) often yields no advantage because the functional attributes are not necessarily unique across different destinations (Prentice, 2006). Given that place personality traits positively and significantly impact the development of attitudes and behavior about a destination (Ekinci and Hosany, 2006; Rojas-Mendez et al., 2013a), we propose that a country’s Destination Marketing Organization (DMO) should expand their focus beyond a destination’s functional attributes and also include destination personality traits to achieve more effective differentiation and positioning. Therefore, we propose that personality traits are a means to position countries as tourism destinations, and as such countries should use some of the vast personality inventory to position themselves.

While there are many mechanisms for countries to position themselves, a country’s online presence is becoming increasingly important and is critical to disseminate information and drive tourism (Pike, 2009). Consumers who are selecting travel destinations rely on the Internet, and in particular tourism websites, to help inform their decision on where to travel and spend their vacation dollars (Oliveira and Panyik, 2015). How a country represents itself online through the conscious choice of particular text reflects a certain image and personality that is interpreted by the consumer. We have chosen to focus our study on South America given that its tourism industry is growing at a rate 65% faster than the worldwide average (UNWTO, 2013) and that its countries provide diverse and distinct tourist opportunities.

While previous literature has established that place personality traits are important for DMOs to consider, little empirical work assessing the personality metaphor in DMO tourism materials has been undertaken. The limited work that has been done (De Moya and Jain, 2013; George and Victor Anandkumar, 2014; Pitt et al., 2007) is based on Aaker’s brand personality research (1997) which has been criticized for its lack of cultural neutrality (George and Victor Anandkumar, 2014) and that it is actually assessing the much broader construct of brand identity rather than brand personality (Azoulay and Kapferer, 2003). Assessing the extent to which National DMOs are currently active in using the personality metaphor (as measured by culturally neutral and valid personality traits) in their online marketing materials is a critical and necessary first step in moving this nascent research area forward.

This exploratory study’s main research question is: To what extent are South American National DMOs using destination personality traits in their tourism websites?

This study adds insight into how a country chooses to represent itself online and may have potential implications for tourism destination choices. This article employs a text mining methodology to detect place personality within tourism websites. As the amount of unstructured text continues to grow text mining is becoming an increasingly popular and important methodology for social scientists to employ. Critical to the text mining approach to analyzing data is the construction of a dictionary that reduces search space so the large volumes of data become manageable. By employing a customized, culturally neutral personality trait dictionary within the text mining methodology, this article contributes to the literature by providing a unique approach to the detection of place personality that has been neglected in previous studies. Results are visualized and analyzed using correspondence analysis.

This manuscript is structured as follows. First a comprehensive literature review including destination branding, tourism positioning, place personality, information and communication technologies (ICT), and destination branding is presented. Next the text mining and correspondence analysis methodology is discussed followed by the results of analyzing the national tourism websites of the 10 largest and most traveled to South American countries (UNWTO, 2013). These countries include Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Paraguay, Peru, Uruguay, and Venezuela. Finally, conclusions, limitations, and suggestions for future research are provided.

Literature review

Destination branding

Destination branding has emerged as an important marketing tool available for country DMOs due to tough international competition and high levels of substitutability of destinations (Kumar and Nayak, 2014; Oliveira and Panyik, 2015; Pike and Ryan, 2004). Destination branding has been defined as a way to communicate a destination’s unique identity by differentiating a destination from its competitors. This construct has also been explained as:

the set of marketing activities that (1) support the creation of a name, symbol, logo, word mark or other graphic that readily identifies and differentiates a destination; that (2) consistently convey the expectation of a memorable travel experience that is uniquely associated with the destination; that (3) serve to consolidate and reinforce the emotional connection between the visitor and the destination; and that (4) reduce consumer search costs and perceived risk. (Blain et al., 2005: 337)

Thus, destination branding activities help DMOs to project an intended destination brand identity, which is expected to contribute to the creation of a corresponding unique positive and differential destination image (Qu et al., 2011). In addition, destination image has been identified as a key construct in destination positioning (Pike and Ryan, 2004).

Positioning a tourism destination

Countries compete with each other as tourism destinations to attract a larger portion of the finite number of international tourists (Beverland and Lindgreen, 2002). As a result, positioning has become a key strategy for countries as tourism destinations worldwide compete for tourists’ attention (Pike and Ryan, 2004). The term positioning was initially applied to products (Ries and Trout, 1981), but it has become increasingly popular and therefore well accepted as a strategic tool for other entities including tourism destinations (Li et al., 2015). Positioning of a tourism destination is understood as a process of establishing and maintaining a distinctive place for a destination in the minds of travelers in the targeted market in reference to competing destinations (Pike and Ryan, 2004).

Qu and Qu (2015) argue that positioning studies have mainly followed a utilitarian perspective by focusing on identifying and accentuating the more distinguishing and reputable functional attributes. Theoretical support for this trend relies on the information processing theory (IPT) that sees both marketers and consumers as logical thinkers. However, IPT serves to explain decision-making process only in terms of cognitive operations (Tybout et al., 1981). Therefore, in the context of tourism destinations, all functional attributes such as climate, geography, and nature have been systematically used by many DMOs as a way of positioning arguments for consumers who are assumed to be motivated by utilitarian needs. This approach has been criticized because it is easily substitutable (Usakli and Baloglu, 2011) and does not often result in distinctive positioning (Prentice, 2006).

An alternative way to examine decision-making process is by means of attachment theory (Bowlby, 1958). Attachment theory was initially focused on the long-term relationships and emotional bonds between individuals. However, research in marketing suggests that attachments can be extended beyond the person-to-person relationship context without changing the fundamental conceptual properties and behavioral effects (Belk, 1988; Kleine et al., 1989; Mehta and Belk, 1991). For example, a destination can emotionally bind a person and evoke a sense of self that can be maintained over time (Hill and Stamey, 1990).

Consumers’ nonutilitarian needs allow marketers to apply the conceptual properties of attachment (Qu and Qu, 2015). These consist of all needs incompatible with functional utility and may include symbolic, emotional, and experiential needs. In the context of tourism destinations, the use of the personality metaphor has been suggested to establish a distinctive brand identity that possesses strong symbolic meanings for potential tourists (Hosany et al., 2006). The perception of the personality of a place or destination is symbolic and affective (Hosany et al., 2006; Hultman et al., 2015; Ibrahim and Gill, 2005), and destinations are rich in terms of these values (Zouganeli et al., 2012). Recent research has demonstrated the positive link between affective destination images such as place personality and salient behavior variables such as visit and revisit intentions (Jeong et al., 2009; Murphy et al., 2007a; Rojas-Mendez et al., 2013a). An individual will be inclined to engage into and attached to a relationship with a destination that he considers close to his personality (as part of his self-concept). Consequently, congruence between own personality and perceived brand personality can foster identification (Rojas-Mendez et al., 2015; Stokburger-Sauer, 2011). Destination personality serves to differentiate and create sustainable competitive advantage (Aaker, 1996), develop an emotional relationship with a destination (Park and Jung, 2010), and thus can be considered as a relevant component of effective branding when it comes to positioning destinations with nonfunctional values (Kim and Lehto, 2013). Recently, Kumar and Nayak (2014) have identified the creation of a personality for a destination as a critical and effective component for its positioning process.

The place personality construct

The first destination personality definition appearing in the literature was proposed by Hosany et al. (2006). They adapted Aaker’s product brand personality definition (1997) and define destination personality as “the set of human characteristics associated to a tourism destination” (p. 639).

In the context of country personality, d’Astous and Boujbel (2007) were the first to offer an ad hoc definition for the concept. They define country personality as “the mental representation of a country on dimensions that typically capture an individual’s personality” (p. 233). More recently, Rojas-Méndez et al. (2013a) have presented a more comprehensive definition for the concept of nation brand personality. They define it as

the set of positive and/or negative human personality traits comprising specific dimensions that internal and external audiences associate to a country name, based on previous experiences and perceptions as a consequence of the actions, intentions, and opinions of that country’s government, companies and institutions, and society at large. (p. 1029)

This definition recognizes the complexity of the concept, the existence of not only positive but also negative traits, and that personality can be used to position nations in the eyes of diverse audiences such as citizens and foreigners (i.e. tourists, consumers, investors, students, etc.).

From the literature, it is evident that the application of the personality metaphor to places is still in its embryonic stage (Kumar and Nayak, 2014). Nevertheless, there are seven general accepted conclusions that deserve to be highlighted. First, personality dimensions are appropriate and easily used to describe places. Several studies report no problems for respondents to associate personality traits to destinations (e.g. d’Astous and Boujbel, 2007; d’Astous and Li, 2009; Murphy et al., 2007b; Rojas-Méndez and Papadopoulos, 2012; Rojas-Méndez et al., 2013a, Zouganeli et al., 2012). Ekinci and Hosany (2006) assert that personality traits can be associated with a destination through both direct and indirect ways. A direct way includes citizens, hotel employees, restaurants, and tourist attractions or simply through the tourist’s imagery. An indirect way includes marketing programs and media construction of destinations.

Second, place personality is a multidimensional construct. All the scales published in the last decade for countries, regions, cities, and destinations in general have shown that the construct has between three and six dimensions (d’Astous and Boujbel, 2007; Rojas-Méndez et al., 2013a; Rojas-Méndez and Papadopoulos, 2012; Kaplan et al., 2010). In addition, several of the place personality dimensions have been found to have several facets or sub-dimensions (Kaplan et al, 2010; Rojas-Méndez and Papadopoulos, 2012; Rojas-Méndez et al., 2013a). Several dimensions in the context of place personality resemble human personality as represented by the Big Five. This scale has been developed over many years by several researchers including Allport and Odbert, Norman, Fiske, Goldberg, McCrae, and Costa (John and Srivastava, 1999). The five aggregated domains of human traits, with indications of the typical concepts associated with each in parentheses, include (Costa and McCrae, 1995; John and Srivastava, 1999) neuroticism (anxiety, anger hostility, self-consciousness, and impulsiveness); extraversion (being sociable, talkative, humorous, optimistic, or jolly); openness to experience (imagination, curiosity, originality, and broad-mindedness); agreeableness (kindness, affection, soft-heartedness and tolerance, trust, and cooperativeness); and conscientiousness (dependability, prudence, conformity, thoughtfulness, and being organized). Third, the concept of place personality is culturally bounded. In her seminal study, Aaker (1997) pointed out that while there are culture-free brand personality etic dimensions, some emic dimensions need to be identified and adjusted in different cultures or countries. Since cultures are different, sometimes exhibiting variation within the same country, they should exert significant impact on the emergence of place personality dimensions (Rojas-Méndez and Papadopoulos, 2012).

Fourth, the personality metaphor has been used to differentiate and position places because personality traits provide some rich meanings in terms of symbolic and affective cues (d’Astous and Boujbel, 2007; d’Astous and Li, 2009; Ekinci and Hosany, 2006; Kim and Lehto, 2013; Murphy et al., 2007a, 2007b; Rojas-Méndez et al., 2013b). This idea is consistent with the Haigood’s proposition (1999) for product brand personality which puts forth that a strong positive brand personality should lead to relatively higher product evaluations in comparison to claims that are based on a product’s features and benefits.

Fifth, place brand personality is more related to symbolic or affective attributes of brand image (Hosany et al., 2006; Kneesel et al., 2010). According to Aaker (1996), brand personality connects the brand’s emotional and self-expressive benefits and forms the basis for customer–brand relationships. Thus, destination personality can be used to develop and maintain close ties with different stakeholders such as tourists, local citizens, potential investors, and so on.

Sixth, the concept of place personality provides a context to establish good relationships among stakeholders (Blain et al., 2005). Ekinci and Hosany (2006) claim that the sincerity dimension emphasizes the importance of good relationships between tourists and hosts in evaluating holiday experiences. It is also suggested in the consumer psychology literature that when a brand has a well-defined personality, people can grow a better understanding of it and hence relate to it easier (Sung and Tinkham, 2005).

Finally, place personality dimensions affect stakeholders’ attitudes and behaviors. It has been already acknowledged that brand personality influences consumer preferences and patronage (Sirgy, 1982). Ekinci and Hosany (2006) contend that a strong destination personality leverages the effect of destination image on intention to recommend, and d’Astous and Boujbel (2007) demonstrate that all six personality dimensions derived from their study (i.e. agreeableness, wickedness, snobbism, assiduousness, conformity, and unobtrusiveness) have a statistically significant impact on attitudes toward specific countries with “agreeableness” being the most important dimension. Similar results have been suggested by Zeugner-Roth and Zabkar (2015) and Rojas-Méndez et al. (2013a) in the context of country personality.

ICT and destination branding

Recent literature has identified the duality of destination branding that is enabled by traditional push ICT platforms, such as tourism websites, and by more recent social media tools such as blogs, Twitter, YouTtube, and Facebook, among others (Oliveira and Panyik, 2015). Social media tools have reduced the “barriers to entry” for web content creation, and thus travelers, tourists, and journalists have a hand in socially constructing a destination’s brand (Oliveira and Panyik, 2015). While there is little doubt the role of social media is playing within the tourism industry, data indicates that tourism websites are the second most influential source of information for tourists making a travel destination decision (European Commission, 2013).

There has been some initial work completed on extracting destination branding information from tourism websites (Govers and Go, 2005; Lee et al., 2006; Tang et al., 2009). Govers and Go found that there are differences between the perceived destination identity of public and private representations of Dubai based on an analysis of photos and text scraped from the web (2005). Lee et al. examined 50 US state tourism websites and identified and compared unique selling propositions and positioning strategies (2006). In a very interesting study, Tang et al. compared the tourism image between the English and Chinese version of the Macau tourism website and concluded that portrayed images were quite different (2009).

While the aforementioned studies all report about destination branding on tourism websites, none of them report on the place personality construct. Several recent articles have made inroads into the detection and reporting of place personality on web-based tourism artifacts (De Moya and Jain, 2013; George and Victor Anandkumar, 2014; Pitt et al., 2007). Pitt et al. were among the first to attempt to detect place personality within tourism websites (2007). They performed their analyses across a subset of African nations but clearly state their main contribution is the method rather than the comparison of African nations. In fact, several articles’ stated contribution are detailed content analysis processes for structuring and subsequently analyzing digital tourism information (Koltringer and Dickinger, 2015; Oliveira and Panyik, 2015). Koltringer and Dickinger provide a high-level web content mining approach to extract destination identity and image (2015). While their work is not specific to place personality, their method could be adapted to extract such information. Similarly, Oliveira and Panyik’s proposed methodology is fairly generic but does not include place personality as a construct of interest (2015).

Using the methodology prescribed by Pitt et al. (2007), De Moya and Jain compared the place personality of Mexico and Brazil as projected through their Facebook sites. In addition, they analyzed how consistently the projected personality traits from Facebook are transferred and reflected in each countries’ Facebook “friends” postings (2013). Similarly, George and Victor Anandkumar operationalized the method described by Pitt et al. (2007) to elicit the place personality of several tropical island destinations via their tourism websites and compared said personality to the perceived destination personality of the various islands as assessed via an online survey (2014). All of the studies reviewed that assess place personality expressed in digital tourism materials uses Aaker’s brand personality (1997) as the foundation of the dictionary that enables place personality identification. There have been several important criticisms of Aaker’s approach, including the fact that the words used to detect personality lack cross-cultural validity (George and Victor Anandkumar, 2014) and secondly that the operationalization of the brand personality construct is too broad and is actually assessing brand identity of which personality is only a single facet (Azoulay and Kapferer, 2003).

To address the criticisms of Aaker’s approach of detecting brand personality, we ensure that (a) we based our place personality dictionary on a culturally neutral set of place personality traits and (b) the traits chosen for analysis are validated as personality traits and not some other facet of brand identity. A description of the derivation of the traits used for the study is provided in the methodology section that follows.

Methodology

For our sample we included all South American countries with a population of over one million people. This yielded 10 countries. For each country, we identified the English version of their official government-endorsed tourism website and scraped all the text content from all web pages of all websites. No images, videos, or content from external sites were included. The collection took place in November and December 2013. All the collected text was preprocessed and organized such that the content from a single web page was represented as a record of data that was subsequently entered into the computer aided qualitative data analysis (QDA) software Miner. From within QDA Miner, an associated software product called WordStat was enabled and used to do some initial descriptive analysis. Table 1 lists all the identified tourism websites and the number of words that exist within each website.

Table 1.

South American tourist websites.

Country	Website	No. of pages	No. of words	Words/page
Argentina	http://www.turismo.gov.ar	198	46,694	236
Bolivia	http://www.bolivia.travel/home_en.aspx?IDI=2	504	56,646	112
Brazil	http://www.braziltour.com	2925	1,35,986	47
Chile	http://chile.travel/whynot/?	85	29,445	346
Colombia	http://www.colombia.travel/en	235	1,21,784	518
Ecuador	http://en.ecuador.travel/	89	16,321	183
Paraguay	http://country.paraguay.com/tourism/touristic_attractions.php	56	14,383	257
Peru	http://www.peru.travel/en-us	633	1,00,616	159
Uruguay	http://www.turismo.gub.uy/	81	16,712	206
Venezuela	http://venezuela-us.org/tourism-10/	8	3065	383

A customized country personality categorization dictionary was used to determine the number of words in each country’s website that could be used for the analysis. A preliminary version of the dictionary used in this research was developed, based on an exploratory study aimed to develop a list of culturally neutral personality traits used by people worldwide to describe place personality (Rojas-Méndez et al., 2013b). They used a multi-country qualitative exploratory study to generate a list of nation brand personality traits. Their approach was designed to avoid or reduce biases that may occur when traits developed in one country are used by subject consumers in another country (a problem identified by Aaker et al., 2001; d’Astous and Li, 2009). There were initially 588 different personality traits identified. This list includes positive, neutral, and negative valence traits. Even though destinations can take on personalities and appear as positive, negative, or even neutral to potential travelers (Ryu et al., 2013), we assumed no country would consciously use negative personality traits to position itself and thus only classified positive traits.

The next step consisted in grouping all the positive traits in a manageable number of dimensions. Since there is no agreement yet on well-accepted scales to measure nation personality, we used as a proxy for further analysis the NEO-Five-Factor model, also known as the Big Five, which is well accepted within the psychology community (McCrae and Costa, 1987). While neuroticism is typically included in the Big Five, we have chosen to use its opposite pole, emotional stability. This choice was based on the context of our study; we do not expect tourism websites to display traits consistent with neuroticism while we may expect to see traits consistent with emotional stability.

We classified each of the positive nation personality traits within the corresponding Big Five dimensions of agreeableness, extraversion, emotional stability, conscientiousness, and openness to experience. This process was completed by studying and cross-referencing our nation personality traits with already accepted and previously mapped words to the Big Five dimensions that existed in the extant social psychology literature. Because the categorization process was strictly an objective matching exercise of our dictionary words with existing and accepted Big Five categories of words, there was no interpretation required and thus the categorization was performed by a single researcher. Where appropriate, some words were categorized into more than one Big Five dimension. This resulted in a customized dictionary that had 33 positive personality traits classified in extraversion, 72 in agreeableness, 77 in conscientiousness, 29 in emotional stability, and 50 in openness to experience (the remaining words were either not possible to classify or they were negative expressions not apt for this study). The customized dictionary was expanded to include the common inflected forms, derivations, and where appropriate roots of the original words. The initial list of additional word forms were generated by WordStat, and the resulting word list was vetted by an English linguistics expert to ensure that only words form with similar meanings in the context of the study were included.

Results

Analysis based on absolute values

As shown in Table 2, our initial analysis focused on the frequencies of personality traits existing in each country’s tourism website.

Table 2.

Personality words in South American tourism websites.

Country	Agreeableness	Conscientiousness	Emotional stability	Extraversion	Openness to experience	Traits per country
Argentina	77	218	31	215	194	735
Bolivia	124	227	53	220	252	876
Brazil	235	1375	223	792	2012	4637
Chile	31	45	33	82	94	285
Colombia	298	367	87	447	519	1718
Ecuador	120	122	30	88	118	478
Paraguay	28	71	13	55	144	311
Peru	113	259	68	323	462	1225
Uruguay	29	37	11	59	42	178
Venezuela	10	22	7	18	19	76
Traits per Dimension	1065	2743	556	2299	3856	10519

Brazil, by far, is the country that in absolute terms uses the most the personality traits. Colombia and Peru follow Brazil but at a considerable distance. At the other extreme, Venezuela, Uruguay, Chile, Paraguay, and Ecuador use the least amount of personality traits in their tourism websites. Thus, DMOs’ discourses embedded with personality traits that could capture the emotions and attachment of potential tourists are not significantly present in the South American tourism websites.

Even though all countries exhibit some amount of personality traits for all personality dimensions, there is a clear tendency for the majority of them to use openness to experience, and to a lesser extent conscientiousness, as the driver to attract tourists to their countries. For example, Brazil, Colombia, Peru, Bolivia, Paraguay, and Chile, all have openness to experience as their dominant personality dimension. None of the countries have agreeableness as a dominant personality dimension. This is surprising as agreeableness is typically a desirable quality in a tourist destination.

Analysis based on relative values

The previous analysis indicates only how many personality traits the DMOs are using at a country level but says nothing about how those traits are being used in a holistic and relative manner to position a country as destination on the personality dimensions. Therefore, we explored the relative amount of each personality dimension present in each country’s tourism website. Figure 1 below is a bubble plot of relative percentages of trait occurrences per personality dimension per country. The size of the bubble reflects the frequency occurrence data in Table 2 scaled by the number of total words in the respective website (available in Table 1). This allows us to roughly compare bubble size not only within-country but also across country by country personality dimension. A common problem with bubble plots is the scaling of the size of the bubbles. We need to exercise caution assessing bubble sizes beyond broad statements (e.g. largest, smallest, the same, smaller/larger, much smaller/much larger), as the relative size of the bubble does not necessarily reflect the same numerical differential (i.e. a bubble with twice as many relative occurrences of a particular country brand personality dimension will not have a bubble that is exactly twice as large as its comparator).

Figure 1.

Bubble plot of relative percentages of trait occurrences per personality dimension per country.

The most salient macro-results from above are the following. First, few countries have an extreme dominant personality profile when the data is scaled for the size of the website. Brazil clearly is dominated by openness to experience and then by conscientiousness. After that the bubbles of the other three dimensions are much smaller. Ecuador has quite a bit of agreeableness and conscientiousness present as compared to other dimensions. Similar to Brazil, Paraguay’s tourism website has prominent features of openness to experience and conscientiousness relative to other dimensions. While other countries’ websites do display some variance in country personality profile dimensions, by eye, the differences are not as pronounced as the aforementioned countries.

The next step was to perform correspondence analysis (CA) on the data. CA is a method that allows the graphical representation of contingency table data in low dimensional space (Greenacre, 2007) and consequently provides a relatively easy way to graphically represent larger contingency tables in low dimensional space. CA has been successfully used in a variety of domains including marketing (Inman et al., 2004), tourism management (Opoku, 2009; Pitt et al., 2007), teaching and learning (Askell-Williams and Lawson, 2004), among others. In addition, CA has been identified as a powerful approach recommended for use in psychological research (Doey and Kurta, 2011).

The first step in CA is to test the “homogeneity assumption” (Greenacre, 2007) about whether significant differences exist between the countries’ website content in terms of their expressed personality words. This assumption is tested using the χ ² statistic and is reported in Table 3. Given the high χ ² value of 702.185, we can reject the hypothesis and conclude that real differences exist between the countries website content in terms of their expressed personality words. Stated another way we can say that there is a statistical dependence between the rows and columns of the contingency table shown in Table 3.

Table 3.

Summary statistics.

Dimensional representation	Eigenvalues/inertia	χ ²	Percentage of inertia	Cumulative percentage
1	0.052		0.780	0.780
2	0.008		0.119	0.899
3	0.005		0.069	0.968
4	0.002		0.032	1.000
Total	0.067	702.185^a	1.000	1.000

^a p < 0.0001; df 36.

The total variance explained by the model is 6.7%. While the association may be considered weak, it is still significant. Note that there are four dimensions listed in the summary table. The number of dimensions in CA will be (y − 1), where y is the minimum number of columns or rows in the contingency table. In our model, the first dimension explains 78% of the total inertia in the model. Similarly, dimensions 2, 3, and 4 explain 11.9%, 6.9%, and 3.2% of the model’s inertia, respectively. Because the first two dimensions explain 90% of the total inertia, we decided to drop the third and fourth dimensions from the analysis. The reduction of dimension for purposes of creating the perceptual map is a standard step in CA. This allows us to represent the data in a two-dimensional space which aids in interpretation. There is no standard accepted numeric threshold for keeping or rejecting dimensions, rather it is at the researcher’s discretion based on the topic of inquiry (Doey and Kurta, 2011).

There are different types of CA maps available including asymmetric, symmetric, and standard CA biplots. Each of these approaches has benefits and drawbacks, and the choice of map to present is often influenced by the type of inquiry of interest, the overall quality of the display, and personal preference. The main issue identified with interpreting symmetric CA maps has to do with drawing inferences regarding the relationship of row and column data. This is because in symmetric maps, the row and column data are actually being displayed in different planes (as opposed to asymmetric maps and standard CA biplots, which display the data in a single plane). While much discussion has taken place in the literature regarding this issue, it has been found that the distortion in interpreting the association between row and column data in the symmetric map as it were a “true” asymmetric map is minimal (Gabriel, 2002). Taking a cautionary approach, we produced all three types of maps, studied those using identical techniques, and formulated results based on the various displays. The results were all consistent (i.e. we would report the same inferences no matter which display we chose to use). To this end, we are choosing to report a symmetric CA map in Figure 2 which follows the recommendation of Greenacre who states that “the symmetric map is the best default map to use” (2007: 267). The symmetric map typically provides a “nicer-looking” representation than the asymmetric approach which often compresses the primary coordinates of the row profiles toward the center of the map to allow the display of the extreme vertices of the column profiles (obviously, this would be reversed if we were primarily interested in the column profiles, i.e. column profiles would be mapped in primary coordinates and possibly be concentrated in the center of the map to allow the display of the row vertices). The symmetric map is an overlaid joint display of both row and column profiles and is the most popular type of map to use in the CA literature (Greenacre, 2007). We are thus displaying the principal coordinates for both the countries and personality dimensions. The aforementioned coordinates are presented in Tables 4 and 5, respectively.

Figure 2.

CA map of South American tourism websites in relation to personality traits.

Table 4.

Principal coordinates for countries.

Country	Axis 1	Axis 2
Argentina	0.127	−0.091
Bolivia	0.188	−0.001
Brazil	−0.229	0.028
Chile	0.172	−0.222
Colombia	0.276	0.015
Ecuador	0.443	0.279
Paraguay	−0.126	0.045
Peru	0.028	−0.122
Uruguay	0.366	−0.137
Venezuela	0.183	−0.003

Table 5.

Principal coordinates for personality.

Dimension	Axis 1	Axis 2
Agreeableness	0.539	0.148
Conscientiousness	−0.107	0.062
Emotional stability	0.087	−0.069
Extraversion	0.171	−0.143
Openness to experience	−0.187	0.010

The point at which the axes cross represents the average country profile with respect to the destination personality traits. If we look primarily at the horizontal axis, which explains more of the variance than the vertical axis, we see that Brazil and Ecuador have the most different profiles as the horizontal distance between the two countries is the greatest. Brazil’s profile is relatively similar to Paraguay while Ecuador’s is horizontally close to Uruguay. Still focusing on the horizontal dimension there is not a lot of difference between Peru, Argentina, Chile, Bolivia, Venezuela, and Colombia. However, when we take into consideration the vertical dimension, we see more separation in the countries and four clusters are represented. Brazil and Paraguay remain close together while Ecuador is vertically separated from Uruguay and is distinct from any of the other countries. Peru, Argentina, Chile, and Uruguay form a grouping in the lower right corner of the map, while Bolivia, Venezuela, and Colombia are clustered together along the horizontal dimension.

To interpret the similarities and differences in the country profiles, we must view their positions relative to the personality mappings. If the quality of the display is good, then the closer a country is to a personality profile, the higher its profile is to that personality dimension. Quality of display measured by the percentage of cumulative inertia explained by the map is good (i.e. 90%). Countries that cluster near a personality dimension are proportionally using more words in their websites that map to that particular personality dimension.

Ecuador’s website is clearly using proportionally more traits expressing agreeableness than other countries under study. Dominant characteristics of agreeableness include being good natured, soft hearted, selfless, sympathetic, trusting, acquiescent, lenient, forgiving, and flexible. Similarly, both Brazil and Paraguay map closely to the openness to experience dimension which is characterized by being original, imaginative, having broad interests, and being daring and to the conscientious dimension which is characterized by being dutiful, scrupulous, moralistic, hardworking, ambitious, energetic, and persevering. Argentina, Peru, Chile, and Uruguay display to both the extraversion dimension which is reflective of a country being sociable, fun loving, affectionate, and friendly and to the emotional stability dimension characterized by being calm, quiet, relaxed, stable, easygoing, and peaceful. Finally, compared to the other countries, Bolivia, Venezuela, and Colombia do not display any dominant personality characteristics.

To further our exploratory inquiry, we were interested in two additional measures which we label as country personality density (CPD) and country personality consistency (CPC). CPD is defined as the degree to which personality traits were used within each country’s tourism website. It was calculated by the total number of traits detected divided by the total number of words on the website. CPC is defined as the degree of consistency of country personality across web pages within a country’s tourism website. For CPC, we calculated the percentage of individual web pages within a given country’s website that displayed evidence of country personality traits. Table 6 below shows the CPD and CPC measures, respectively.

Table 6.

CPD and CPC measures per country.

Country	Number pages	Total words	Personality words	CPD (%)	CPC (%)
Argentina	198	46,694	735	1.57	74.1
Bolivia	504	56,646	876	1.55	56.2
Brazil	2925	1,35,986	4637	3.41	55.0
Chile	85	29,445	285	0.97	79.5
Colombia	235	1,21,784	1718	1.41	65.4
Ecuador	89	16,321	478	2.93	99.0
Paraguay	56	14,383	311	2.16	76.4
Peru	633	1,00,616	1225	1.22	45.7
Uruguay	81	16,712	178	1.07	56.3
Venezuela	8	3065	76	2.48	42.9

CPC: country personality consistency; CPD: country personality density.

As shown in the table, CPD ranges from approximately 1% to 3.4%. The average for all the countries is 1.9 personality traits used per 100 words included in the website. This is a very low rate of use. The country with the lowest CPD is Chile with a score of 0.97%. This means that Chile is currently using the lowest percentage of country personality traits in its tourism website. Alternatively Brazil has the highest CPD measure at 3.41%. Other relatively high countries include Ecuador and Venezuela, while Uruguay, Peru, and Colombia join Chile as having relatively low CPDs. As far as consistency is concerned, almost every web page in Ecuador’s website displayed some form of country brand personality with a 99% CPC. Other high ranking countries included Chile, Paraguay, and Argentina with CPCs of 79.5%, 76.4%, and 74.1%. The lowest ranking countries were Venezuela and Peru, who had CPCs of less than 50% at 42.9% and 45.7%, respectively. Given CPC is highly influenced by website structure, we do need to be cautious about drawing any definitive conclusions from this measure.

Conclusions, limitations, and future work

This is the first article that uses text mining methodology of correspondence analysis with a dictionary explicitly developed for places in the context of nation brand personality. Previous studies used a dictionary originally developed for consumer products or taken from the human psychology literature in general.

Our research on South American destinations shows the emergence of four general clusters of countries with regard to the destination personality their websites are portraying. Specifically, Argentina, Chile, Peru, and Uruguay are most strongly associated with extraversion and emotional stability. Brazil and Paraguay most strongly associate with conscientiousness and openness to experience, while Ecuador is most strongly associated with agreeableness. Finally, Bolivia, Venezuela, and Colombia do not associate proportionally high with any of the personality dimensions.

While general country clusters have emerged, the results of this study, and the response to our main research question, are that South American countries’ DMOs are not using in a substantial way the personality metaphor to create strong positioning for international tourists. We therefore assume that the use of this new marketing approach is in its nascent stages. Out of the ten countries analyzed, only Brazil and Colombia stand out by using more than 300 unique personality traits to promote the country to potential tourists. At the other extreme, Paraguay, Chile, Uruguay, and Ecuador show a frequency of 50 or fewer personality traits used. Given the relative paucity of the use of personality traits in the websites, it is unlikely that there is an effect from the presence of the traits to create strong meaningful links between potential international tourists and the countries as destinations. This relationship should be empirically tested in some manner before any definitive conclusion can be reached.

Given that there is established theory from a variety of disciplines suggesting that personality traits should be used to create differentiation for tourism destinations, and given that the results of our study indicate that this is not occurring, a fundamental question that emerges is “Why aren’t South American DMOs using personality as a differentiation strategy?” While we cannot with any certainty answer this question without further inquiry, we feel comfortable prescribing the following advice for practitioners and researchers interested in this domain of inquiry. South American countries’ tourism brand managers should educate themselves and become more aware of destination personality traits and move toward strategically using them to represent their country’s offerings on their websites. A key artifact for operationalizing this strategy is determining what traits to use when describing a country on its tourism website. DMOs’ managers can use the results to see if they are consistent with their perceptions of their own country’s image to ensure that they are accurately reflecting themselves and the experience expected by future tourists. Similarly, they can use the result to reposition themselves relative to their competitors in the increasingly competitive South American tourism industry. In doing so, they should embrace a more holistic approach when developing an international positioning strategy which at the same time allows their country to differentiate from what competitors are offering. A more targeted approach could involve the inclusion of destination personality traits that have been shown to resonate with tourists from particular geographic areas. These region-customized trait lists could form the basis of a series of tailored websites one of which would display based on filtering the IP address of the HTTP request. For example, an HTTP request from North America could result in a slightly different website being served than an HTTP request that was sent from Europe. The advantage of this approach is that it allows fluidity in capitalizing on targeting nations that historically spend a lot on traveling and/or on countries with citizens with increasing disposable incomes. However, it is important that a country’s tourism websites display a relatively consistent personality position and said position should be consistent with the tourism experience that the country actually offers. DMOs could work with the academic community to validate that their tourism websites intended personality projection are being interpreted by potential tourism consumers in an unequivocal and consistent manner. This research could be executed using surveys, experiments, or a mixed-methods approach.

There are a few limitations to the current study. First, we have used the English language websites, even though the official language of all of them but one is Spanish. Second, at a general level, dictionary-based content analysis can be criticized for not capturing nuances in textual communication that could only be assessed by reading and interpreting text at a detailed level. While a valid concern, advances in software tools have made it much easier to construct dictionaries including associated rules that allow advanced processing of text. For example, WordStat provides mechanisms for searching for strings, patterns, and negations, among other linguistic structures to help build valid and useful dictionaries. The customized dictionary used in our current study is limited with regard to the number of words included and thus being detected in the assessment of the tourism websites. This limitation can also be viewed as strength as the dictionary leverages previous research that established a set of cross-culturally valid personality traits (Rojas-Mendez et al., 2013b) that resonate with potential tourism consumers. Future work may include the use of other types of dictionaries that may be relevant for DMOs’ management. For example, sentiment analysis could be performed which would give a more macro perspective on the tone (either positive or negative) of what tourism websites are communicating.

Additionally, more advanced inductive text analytic techniques could be employed to determine if any generalized country profile themes emerge that then could be compared to other tourist destination websites. While our current work is focused on text analytics, the majority of tourism websites we analyzed did include images and video which may also communicate place brand personality. Thus, there are opportunities for scholars to start assessing the impact of images and videos from tourism websites on potential visitors to countries.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Aaker

(1996) Branding Strong Brands. New York: Free Press.

Aaker

(1997) Dimensions of brand personality. Journal of Marketing Research 34: 342–352.

Aaker

Benet-Martinez

Garolera

(2001) Consumption of symbols as carriers of culture: a study of Japanese and Spanish brand personality constructs. Journal of Personality and Social Psychology 81(3): 492–508.

Askell-Williams

Lawson

(2004) A correspondence analysis of child-care students’ and medical students’ knowledge about teaching and learning. International Education Journal 5(2): 176–204.

Azoulay

Kapferer

(2003) Do brand personality scales really measure brand personality? Brand Management 11(2): 143–155.

Belk

(1988) Possessions and the extended self. Journal of Consumer Research 15: 139–168.

Beverland

Lindgreen

(2002) Using country-of-origin in strategy: the importance of context and strategic action. Journal of Brand Management 10 (2): 147–167.

Blain

Levy

Ritchie

JRB

(2005) Destination branding: insights and practices from destination management organizations. Journal of Travel Research 43(4): 328–338.

Bowlby

(1958) The nature of the childs tie to his mother. International Journal of Psychoanalysis 39: 350–371.

10.

Costa

McCrae

(1995) Domains and facets: hierarchical personality assessment using the revised NEO personality inventory. Journal of Personality Assessment 64(1): 21–50.

11.

d’Astous

(2009) Perceptions of countries based on personality traits: a study in China. Asia Pacific Journal of Marketing and Logistics 4(21): 475–488.

12.

d’Astous

Boujbel

(2007) Positioning countries on personality dimensions: scale development and implications for country marketing. Journal of Business Research 60(3): 231–239.

13.

De Moya

Jain

(2013) When tourists are your “friends”: exploring the brand personality of Mexico and Brazil on Facebook. Public Relations Review 39: 23–29.

14.

Doey

Kurta

(2011) Correspondence analysis applied to psychological research. Tutorials in Quantitative Methods for Psychology 7(1): 5–14.

15.

Ekinci

Hosany

(2006) Destination personality: an application of brand personality to tourism destinations. Journal of Travel Research 45(2): 127–139.

16.

European Commission (2013) Attitudes of Europeans towards tourism, 334. Available at: http://ec.europa.eu/public_opinion/flash/fl_370_en.pdf (accessed 9 December 2015).

17.

Gabriel

(2002) Goodness of fit of biplots and correspondence analysis. Biometrika 89(2): 423–436.

18.

George

Victor Anandkumar

(2014) Portrayed and perceived online destination personality of select island destinations. Anatolia – An International Journal of Tourism and Hospitality Research 25(2): 280–289.

19.

Govers

(2005) Projected destination image online: website content analysis of pictures and text. Information Technology and Tourism 7: 73–89.

20.

Greenacre

(2007) Correspondence Analysis in Practice. 2nd ed. Boca Raton: Chapman & Hall/CRC.

21.

Haigood

(1999) The brand personality effect: an empirical investigation, In Menon

Sharman

(Eds) Marketing theory and applications, pp. 149–150. Chicago: American Marketing Association.

22.

Hankinson

(2004) Relational network brands: towards a conceptual model of place brands. Journal of Vacation Marketing 10(2): 109–121.

23.

Hill

Stamey

(1990) The homeless in america: an examination of possessions and consumption behaviors. Journal of Consumer Research 17: 303–321.

24.

Hosany

Ekinci

Uysal

(2006) Destination image and destination personality: an application of branding theories to tourism places. Journal of Business Research 59(5): 638–642.

25.

Hultman

Skarmeas

Oghazi

. (2015) Achieving tourist loyalty through destination personality, satisfaction, and identification. Journal of Business Research 68(11): 2227–2231.

26.

Ibrahim

Gill

(2005) A positioning strategy for destination based on analysis of customers perceptions and satisfactions. Marketing Intelligence & Planning 23(2–3): 172–188.

27.

Inman

Shankar

Ferraro

(2004) The roles of channel-category associations and geodemographics in channel patronage. Journal of Marketing 68: 51–71.

28.

Jeong

Kim

. (2009) Horse racing image: re-examination of relations between image and intention to visit. Journal of Quality Assurance in Hospitality & Tourism 10(3): 194–217.

29.

John

Srivastava

(1999) The big five trait taxonomy: history, measurement, and theoretical perspectives, In Pervin

John

(eds) Handbook of Personality: Theory and Research. 2nd ed. New York: Guilford, pp. 102–138.

30.

Kaplan

Yurt

Guneri

. (2010) Branding places: applying brand personality concept to cities. European Journal of Marketing 44(9–10): 1286–1304.

31.

Kim

Lehto

(2013) Projected and perceived destination brand personalities: the case of South Korea. Journal of Travel Research 52(1): 117–130.

32.

Kleine

III Kleine

Kernan

(1989) These are a few of my favorite things - toward an explication of attachment as a consumer-behavior construct. Advances in Consumer Research 16: 359–366.

33.

Kneesel

Baloglu

Millar

(2010) Gaming destination images: implications for branding. Journal of Travel Research 49(1): 68–78.

34.

Koltringer

Dickinger

(2015) Analyzing destination branding and image from online sources: a web content mining approach. Journal of Business Research 68: 1836–1843.

35.

Kumar

Nayak

(2014) The measurement & conceptualization of destination personality. Tourism Management Perspectives 12: 88–93.

36.

Lee

Cai

O’Leary

(2006) WWW.branding.states.us: an analysis of brand-building elements in the US state tourism websites. Tourism Management 27: 815–828.

37.

Cheng

Kim

. (2015) Positioning USA in the chinese outbound travel market. Journal of Hospitality & Tourism Research 39(1): 75–104.

38.

McCrae

Costa

Jr (1987) Validation of the five-factor model of personality across instruments and observers. Journal of Personality and Social Psychology 52(1): 81–90.

39.

Mehta

Belk

(1991) Artifacts, identity, and transition: favorite possessions of indians and indian immigrants to the U.S. Journal of Consumer Research 17(March): 398–411.

40.

Murphy

Benckendorff

Moscardo

(2007a) Linking travel motivation, tourist self-image and destination brand personality. Journal of Travel & Tourism Marketing 22(2): 45–59.

41.

Murphy

Moscardo

Benckendorff

(2007b) Using brand personality to differentiate regional tourism destinations. Journal of Travel Research 46(1): 5–14.

42.

Oliveira

Panyik

(2015) Content, context and co-creation: digital challenges in destination branding with references to Portugal as a tourist destination. Journal of Vacation Marketing 21(1): 53–74.

43.

Opoku

(2009) Mapping destination personality in cyberspace: an evaluation of country web sites using correspondence analysis. Journal of Internet Commerce 8: 70–87.

44.

Park

Jung

(2010) A study on the impact of the congruence of store personality and self-image toward relationship strength and store loyalty. Journal of Marketing Studies 16: 45–68.

45.

Pike

(2009) Destination brand positions of a competitive set of near-home destinations. Tourism Management 30(6): 857–866.

46.

Pike

Ryan

(2004) Destination positioning analysis through a comparison of cognitive, affective, and conative perceptions. Journal of Travel Research 42(May): 333–342.

47.

Pitt

Opoku

Hultman

. (2007) What I say about myself: communication of brand personality by African countries. Tourism Management 28(3): 835–844.

48.

Prentice

(2006) Evocation and experiential seduction: updating choice-sets modelling. Tourism Management 27(6): 1153–1170.

49.

Hyunjung

(2011) A model of destination branding: integrating the concepts of the branding and destination image. Tourism Management 32(3): 465–476.

50.

(2015) Nonutilitarian tourism destination positioning: a case study in China. International Journal of Tourism Research 17: 388–398.

51.

Ries

Trout

(1981) Positioning: The Battle for Your Mind. New York: McGraw Hill.

52.

Rojas-Méndez

Papadopoulos

Alwan

(2015) Testing self-congruity theory in the context of nation brand personality. Journal of Product and Brand Management 24(1): 18–27.

53.

Rojas-Méndez

Murphy

Papadopoulos

(2013a) The U.S. brand personality: a Sino perspective. Journal of Business Research 66(8): 1028–1034.

54.

Rojas-Méndez

Papadopoulos

Murphy

(2013b) Measuring and positioning nation brands: a comparative brand personality approach. Corporate Reputation Review 16(1): 48–65.

55.

Rojas-Méndez

Papadopoulos

(2012) Argentine consumers’ perceptions of the U.S. brand personality. Latin American Business Review 13(4): 329–345.

56.

Ryu

Bordelon

Pearlman

(2013) Destination-image recovery process and visit intentions: lessons learned from hurricane katrina. Journal of Hospitality Marketing & Management 22: 183–203.

57.

Sirgy

(1982) Self-concept in consumer behaviour: a critical review. Journal of Consumer Research 9: 287–300.

58.

Stokburger-Sauer

(2011) The relevance of visitors’ nation brand embeddedness and personality congruence for nation brand identification, visit intentions and advocacy. Tourism Management 32(6): 1282–1289.

59.

Sung

Tinkham

(2005) Brand personality structures in the United States and Korea: common and culture-specific factors. Journal of Consumer Psychology 15: 334–350.

60.

Tang

Choi

Morrison

. (2009) The many faces of Macau: a correspondence analysis of the images communicated by online tourism information sources in English and Chinese. Journal of Vacation Marketing 15(1): 79–94.

61.

Tybout

Calder

Sternthal

(1981) Using information processing theory to design marketing strategies. Journal of Marketing Research 18: 73–79.

62.

UNWTO (2013) Tourism Highlights Available at: www.unwto.org/pub (accessed 6 February 2014).

63.

Usakli

Baloglu

(2011) Brand personality of tourist destinations: an application of self-congruity theory. Tourism Management 32(1): 114–127.

64.

Zeugner-Roth

Žabkar

(2015) Bridging the gap between country and destination image: assessing common facets and their predictive validity. Journal of Business Research 68(9): 1844–1853.

65.

Zouganeli

Trihas

Antonaki

. (2012) Aspects of sustainability in the destination branding process: a bottom-up approach. Journal of Hospitality Marketing & Management 21(7): 739–757.