Examining the Relationships of Cognitive , Affective , and Conative Destination Image : A Research on Safranbolu , Turkey

Destination image is formed by three distinctly different but hierarchically interrelated components called cognitive, affective, and conative (Gartner, 1993:193). In this context, the main purpose of this research is to confirm the relationships between the cognitive, affective, and conative components of destination image. It also aims to reveal the multidimensional nature of cognitive destination image and determine the dimensions that compose it. Data for the sample was collected from 446 tourists who visited Safranbolu, Turkey. Exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and hierarchical regression analysis were conducted to test the hypotheses. The results show that the cognitive destination image is a multidimensional construct. Also it is confirmed that destination image is a hierarchical structure within the cognitive, affective, and conative components. The assessment of both the cognitive and the affective components of destination image can be used as a predictor of tourists’ behavioral intentions toward destinations, such as intention to revisit, recommend, and spread positive word of mouth. Moreover, it is revealed that the affective component is influenced by the cognitive component and the affective component also mediates the relationship between the cognitive and conative components. These results provide some theoretical and managerial implications.


Introduction
The tourism sector is one of the major sectors around the world with an important source of revenues for the economies.Destinations may be considered as factors that make a significant contributions to countries' tourism receipts as well.Then, it can be commented that destinations are competitive elements that need to be marketed in the correct way for the countries.If the destinations within the country are the places that are known and preferred by domestic and foreign tourists, it will have a great deal of influence on regional development and also development countrywide.Branding and in this way increasing the perceived image and perceived value of destinations is an important manner in terms of tourism competitiveness between countries.
The aim of this study is to determine whether the cognitive destination image is a multidimensional structure and also specify the aspects composing cognitive destination image perceived by tourists.Furthermore, this study aims to provide empirical evidence in order to reveal the hierarchical nature of the cognitive-affective-conative model of destination image proposed by Gartner (1993); Dann (1996); Pike and Ryan (2004); Konecnik and Gartner (2007); Tasci, Gartner, & Cavusgil (2007); Pena et al. (2012); Agapito et al. (2013).The study model is designed to propose relationships among the different levels of evaluations within destination image structure composed of cognitive, affective, and conative.In the study the mediation effect of affective destination image in the relationship between cognitive and conative destination image is also investigated.
This study is expected to contribute to the literature by analyzing the destination image within the relations between cognitive, affective, and conative components.Also, it is crucial to determine the attributes that is formed the cognitive evaluations of tourists about the destination.Moreover, it is anticipated to be important in terms of practice by leading the way to local authorities, tourism establishments, and destination management organizations that want to improve perceived destination image and, therefore to increase the tourists' willingness to behave positively toward the destination.

Literature Review
Cognitive destination image refers to individual's own knowledge and beliefs about the destination (an evaluation of the perceived attributes of the destination) (Beerli & Martin, 2004a).In other words, cognitive destination image should be composed of perceptions of individual attributes (Echtner & Ritchie, 1993).Within the studies in the literature, cognitive destination image is measured with several attributes and dimensions (Calantone, et al., 1989;Fakeye & Crompton, 1991;Baloglu & McCleary, 1999a;Baloglu & McCleary, 1999b, Baloglu & Mangaloglu, 2001;Beerli & Martin, 2004a;Chen & Tsai, 2007;Chi & Qu, 2008;Qu, Kim, & Im, 2011;Assaker, 2014;Quintal et al., 2014;Kim & Park, 2015).Calantone et al., (1989) used 13 attributes called good shopping facilities, warm and friendly people, safety, varied and good food, unusual cultural experiences, many tourist attractions, good tourist facilities, value for money, good transportation facilities, exciting night life and entertainment, beautiful scenery, relaxing places to visit, beaches and water sports in their research for measuring the destination perceptions.Fakeye and Crompton (1991) suggested 32 items that were used to measure the Rio Grande Valley's destination image.After factor analysis, 23 items within the five factors (social opportunities and attractions; natural and cultural amenities; accommodations and transportation; infrastructure, foods and friendly people; bars and evening entertainment) were revealed.(Fakeye & Crompton, 1999).Baloglu andMcCleary (1999a, 1999b); Baloglu and Mangaloglu (2001) measured perceptional/cognitive destination image within 14 items (good value for money, beautiful scenery/natural attractions, good climate, interesting cultural attractions, suitable accommodations, appealing local food (cuisine), great beaches/water sports, quality of infrastructure, personal safety, interesting historical attractions, unpolluted/unspoiled environment, good nightlife and entertainment, standard hygiene and cleanliness, interesting and friendly people) generated from the literature review and a content analysis of the four destinations' (Turkey, Greece, Italy, and Egypt) guidebooks and brochures.Beerli and Martin (2004a:666) revealed 21 items classified into 5 dimensions (natural and cultural resources; general, tourist and leisure infrastructures; atmosphere; social setting and environment; sun and sand) for destination image of Lanzorate, Spain.Chen and Tsai (2007) carried out an empirical study in Kengtin Region, an important and famous coastal destination in Southern Taiwan.After the factor analyses 14 items and 4 factors of "destination brand", "entertainment", "nature and culture", "sun and sand" were served as the measurement variables of destination image (Chen & Tsai, 2007:1119).Chi and Qu (2008) found out nine-factor solution with 37 items.The nine factors were labeled based on the core variables that constituted them: travel environment, natural attractions, entertainment and events, historic attractions, travel infrastructure, accessibility, relaxation, outdoor activities, and price and value (Chi & Qu, 2008).Qu et al. (2011) designated that underlying dimensions of cognitive destination image of Oklahoma consist of quality of experiences, touristic attractions, environment and infrastructure, entertainment/outdoor activities and cultural traditions.Assaker (2014) tested an empirical model in the context of Australia as a tourism destination.Research confirmed that destination image is operationalized as a second-order factor model, which is formed by six first order factors (18 destination attributes) of natural and well-known attractions; variety of tourist services and culture; quality of general tourist atmosphere; entertainment and recreation; general environment; and accessibility (Assaker, 2014).Quintal et al., (2014) derived 5 factors (infrastructure, activities/services, perceived financial risk, local produce, climate) within 15 items from the brand image of Western Australia's South-West region.Kim and Park (2015) explored 4 factors which are labeled as quality of experience; infrastructure and attractions; value and environment; and excitement and comfort within 16 items for cognitive destination image of Weh Island, Indonesia.Hence, Hypothesis 1 is developed as follows.
H1: Cognitive destination image is a multidimensional construct.
H2: Cognitive destination image significantly influences affective destination image.
Conative destination image is the action step: how one acts on the information and how they feel about a destination.The actions comprise individual's actual conduct or intention to revisit, recommend the destination to others or spread positive word of mouth (Gartner, 1993;Bigne, Sanchez, & Sanchez, 2001;Pike & Ryan, 2004;Konecnik & Gartner, 2007;Tasci & Gartner, 2007;Tasci et al., 2007;Chi & Qu, 2008;Kim et al., 2013).There are several empirical evidences that support the influence of the destination image on behavioral intentions or loyalty (Bigne et al., 2001;Chen & Tsai, 2007;Alcaniz, Garcia, & Blas, 2009;Qu et al., 2011;Moon, Ko, Connaughton, & Lee, 2013;Song et al., 2013).Bigne et al. (2001) found out that the overall image determines people's recommendation of the destination and the intention to return.According to Chen and Tsai (2007)'s research, destination image composed of destination brand; entertainment; nature and culture; and sun and sand has a significantly positive effect on behavioral intentions comprised of likeliness to revisit and willingness to recommend.The results of Alcaniz et al. (2009)'s research revealed that the functional component of cognitive destination image, based on more tangible or measurable perceptions, such as scenery, accommodation or price levels, significantly affects the revisit intention.Besides, research demonstrated that the psychological component of cognitive destination image contains more abstract and intangible characteristics, such as atmosphere or friendliness; shows influence on the intention to recommend.Moon et al. (2013) stated that destination image, included both cognitive (opportunity for adventure, ease of communication, hospitality/friendliness/receptiveness, tourist sites/activities, and nightlife/entertainment) and affective components (relaxing/distressing, friendly/unfriendly, arousing/sleepy, interesting/boring, pleasant/unpleasant, and exciting/gloomy) have positive influences on behavioral intention.Song et al. (2013) determined that destination image that consists of cognitive (people, life and customs; infrastructure and superstructure; indoor and outdoor resources) and affective dimensions have statistically significant and positive influence on destination loyalty intention.
On the other hand, there is little empirical evidence regarding the separate effects of the cognitive destination image (Baloglu, 1999;Alcaniz et al., 2009;Agapito et al., 2013;Li & Yang, 2015) and the affective destination image (Baloglu, 1999;Li et al., 2010;Agapito et al., 2013) on the conative destination image or behavioral intentions.Thus, Hypotheses 3 and 4 are developed as follows.Some theoretical evidences exist that affective destination image might mediate the relation between the cognitive destination image and the conative destination image or behavior (Gartner, 1993;Pike & Ryan, 2004;Konecnik & Gartner, 2007;Tasci et al., 2007).From an empirical point of view, there is a lack of evidence that the mediation effect of the affective destination image is in existence (Agapito et al., 2013).So, Hypothesis 5 is offered as follows.

Sample and Procedure
The survey method was used to collect data in the study.The questionnaire has been prepared in Turkish and English and consists of three parts.In the first part, the statements are used for measuring the cognitive image perceptions of tourists.In the second part, there are statements determining the affective and conative destination image.In the third part, there are questions about the number of visits, the length and the purpose of the visit and the demographic characteristics of the respondents.Data were collected between April and June, 2015 from the tourists who visited Safranbolu, Turkey.
The population of the study consists of all domestic and foreign tourists who were visited Safranbolu.According to the information received from Safranbolu Tourism Information Office, the number of tourists who stayed in accommodations in Safranbolu was 263.444 in 2014.It is also thought to be that the number of tourists who visited Safranbolu was around 700.000 in 2014.The minimum sample size was calculated as 384 participants (Sekaran, 2013).The questionnaire was performed on 500 visitors who selected by convenience sampling method.As a result of the examination, 54 questionnaires were regarded as invalid.Consequently, 446 questionnaires were evaluated in the study.
Primarily, the validity and reliability analyses were performed for the studied constructs.The construct validity of the constructs was tested by factor analysis.In general terms, factor analysis is a modeling approach for studying hypothetical constructs by using a variety of observable proxies or indicators of them that can be directly measured (Raykov & Marcoulides, 2006:117).There are two types of factor analysis including exploratory and confirmatory.An exploratory factor analysis (EFA) in which there is no prior specification of the number of factors is exclusively exploratory (Anderson & Gerbing, 1988:412).In contrast to EFA, confirmatory factor analysis (CFA) is appropriately used when the researcher has some knowledge of the underlying latent variable structure (Byrne, 2010).
Cognitive destination image scale was composed from the several studies in the literature (Baloglu & McCleary, 1999a;Beerli & Martin, 2004a;Chi & Qu, 2008;Pena et al., 2012;Agapito et al., 2013;Song et al., 2013;Assaker, 2014).Therefore, the construct validity of this construct was tested by EFA first.Afterwards, a measurement model was created that covers cognitive, affective, and conative destination image and CFA was performed to examine the construct validity of all studied constructs.
In addition to CFA, reliability analyses were conducted for all constructs.Cronbach's alpha (α) statistics were used in order to determine the internal consistency of the scales within reliability analyses.Cronbach's alpha (α) statistic indicates whether the statements in the scale create a whole to describe a homogeneous structure.Generally, reliability coefficients around 0.90 are considered "excellent", values around 0.80 are "very good", and values around 0.70 are "adequate" (Nunnally & Bernstein, 1994;Kline, 2011).
Hierarchical regression analysis was used to test the relationships between the components of destination imagecognitive, affective, and conative.Exploratory factor analysis, reliability analyses and hierarchical regression analysis were conducted using SPSS, confirmatory factor analysis was performed using AMOS.

Measures
The variables and the sources of scale items used for the measurement are reported in Table 1.Cognitive destination image variable consists of 30 items.In order to measure the respondents' level of agreement with these attributes, individuals were asked to rate the attributes are how much offered in Safranbolu according to the scale ranging from 1 (offers very little) to 5 (offers very much).Affective destination image variable consists of 4 items.These four items are Russell and Pratt's (1980) four bipolar scales of affect (unpleasant-pleasant; gloomy-exciting; sleepy-arousing; distressing-relaxing).In order to measure the respondents' level of agreement with these affects, individuals were asked to rate the feelings toward Safranbolu according to the scale ranging from 1 to 7 (the higher the score, the more positive the affect).
Conative destination image variable consists of 3 questions.These questions involves intention to revisit (Would you return to Safranbolu in the next 12 months?),intention to recommend (Would you recommend Safranbolu to your family and friends?), and intention to spread positive word of mouth (Would you say positive things about Safranbolu to other people?).The respondents were asked to answer these questions according to the scale ranging from 1 (definitively no) to 5 (definitively yes).

Results
Within the demographic characteristics of domestic and foreign tourists participated in the study, gender, marital status, age, education, home city, home country, nationality, number of visits, length of the visit and purpose of the visit variables are evaluated (Table 2).EFA was conducted first to cognitive destination image scale consists of 30 items to reveal the construct validity of the measure.As a result of factor analysis, factor loading of the item -"Interesting cultural activities and festivals" -was determined below 0.50.The specified expression has been deleted (Brown, 2006).EFA results retained with 29 statements are presented in Table 3.The first factor of the scale is "natural attractions" (4 item), second factor is "cultural attractions" (6 items), third factor is "social setting and environment" (5 items), fourth factor is "infrastructure and facilities" (6 items), fifth factor is "accessibility" (4 items) and sixth factor is "price and value" (4 items).Therefore, Hypothesis 1 was supported.
For convergent validity, the composite reliability (CR) and the average variance extracted (AVE) were calculated in order to determine whether the measurement variable was representative of the related construct.As seen in Table 4, all AVEs exceed the cutoff value of 0.50 (Fornell & Larcker, 1981;Hair, Anderson, Tatham, & Black, 1998), and all CRs exceed the cutoff value of 0.70 (Fornell & Larcker, 1981;Hair et al., 1998).These results provided evidence for convergent validity of each of the constructs involved in the research model of the study.
The means, standard deviations, and reliability estimates (Cronbach's alpha) for the measures used in the study are also reported in Table 5.As seen in Table 5, the internal consistency (α) for each measure was above the 0.70 as recommended by Nunnally and Bernstein (1994).Hypotheses 2 and 3 predicted that cognitive destination image would be related to affective and conative destination image.As shown in Table 6, cognitive destination image was significantly and positively associated with affective destination image (β=0.692,p<0.01) and conative destination image (β=0.757,p<0.01).Thus, Hypotheses 2 and 3 were supported.
Hypotheses 4 as well predicted that affective destination image would be related to conative destination image.As shown in Table 4, affective destination image was significantly and positively associated with conative destination image (β=0.276,p<0.01).Therefore, Hypotheses 4 was supported.
Three-step mediation regression procedure outlined by Baron and Kenny (1986) was followed to test the mediating role of affective destination image (Hypotheses 5).Mediation represents a model in which an independent variable causes an intervening variable, which in turn causes the dependent variable (Baron & Kenny, 1986:1176;MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002;Preacher & Hayes, 2004).Baron and Kenny (1986) stipulate three requirements for establishing mediation effect.
1.The independent variable (cognitive destination image) must be significantly related to mediator variable (affective destination image).
2. The independent variable (cognitive destination image) must be significantly related to dependent variable (conative destination image).
3. When the mediator variable (affective destination image) is included in the second step, the effect of independent variable (cognitive destination image) on dependent variable (conative destination image) should be non-significant (full mediation) or significantly weaker (partially mediation).
The first and second requirements were met as indicated by the support for Hypotheses 2 and 3 (Table 6, Step 1 and 2).The third requirement for establishing mediation was examined by including affective destination image (Table 6, Step 3) in the regression equation for conative destination image.When affective destination image was included in the regression equation as a mediator variable, cognitive destination image still remained as a significant predictor, but its estimation weight decreased significantly (β=0.565,p<0.01).Finally, the Sobel test revealed that the indirect path from cognitive destination image to conative destination image through affective destination image was significant (Z=6.23,p<0.01).The indirect estimate of cognitive destination image on conative destination image, through affective destination image, is calculated as 0.192 (i.e., 0.692 × 0.276).Thus, affective destination image partially mediated the relationship between cognitive destination image and conative destination image, providing support for Hypotheses 5.
The summary of the overall model and the results of the hypotheses testing of the study are demonstrated in Figure 1.

Discussion
This research highlights three different points about the destination image.Firstly it is determined that cognitive destination image is a multidimensional construct.In the research, it is found out that the cognitive destination image that perceived by tourists who visited Safranbolu, consists of six dimensions labeled natural attractions, cultural attractions, social setting and environment, infrastructure and facilities, accessibility, and price and value.Accordingly, several researches supported that cognitive destination image is composed of various atrributes or dimensions (Calantone, et al., 1989;Fakeye & Crompton, 1991;Baloglu & McCleary, 1999a;Baloglu & McCleary, 1999b, Baloglu & Mangaloglu, 2001;Beerli & Martin, 2004a;Chen & Tsai, 2007;Chi & Qu, 2008;Qu et al., 2011;Assaker, 2014;Quintal et al., 2014;Kim & Park, 2015).Understanding perspectives of tourists' mental evaluations about the destination, may help decision makers to build strong and long-term marketing strategies and raise the awareness about the destination.
Within the other hypotheses for testing the hierarchical structure of the destination image, the effects of cognitive and affective destination image on conative destination image is revealed.The purpose of analyzing these relationships is to determine which one is more effective on tourists' behaviors between cognitive and affective destination image.Within tourists behaviors or indicators of loyalty; intention to revisit, and recommend the destination and spread positive word of mouth about the destination are included as conative destination image.Study findings show that both cognitive and affective components have an impact on tourists behavioral intentions.However, the direct effect of cognitive destination image is greater than affective destination image.Besides, it was determined that affective destination image mediates the relationship between cognitive and conative destination image.These findings support other studies which underline the importance of affect in destination image, and indicate that the willingness to react positively to the destination might be higher when the visitors associate positive feelings to the destination in addition to positive knowledge and beliefs about the destination (Baloglu, 1999;Li et al., 2010;Agapito et al., 2013).Therefore, thirdly it is revealed that affective destination image is an important predictor of the conative destination image, in addition to cognitive destination image, and affect is also crucial for increasing loyalty, and consequently for the development of a strong relationship between tourists and the destination.
The research contributes to the literature in two ways.First, in the study, cognitive destination image is analyzed in a multidimensional manner.This allows to determine the dimensions which are effective in the formation of the cognitive destination image of a place that is visited mostly for cultural tourism.Second, this study investigates the destination image components separately in a hierarchical structure as cognitive, affective, and conative, in order to better understand how people evaluate and act toward places.It is generally admitted that all mental activity has three aspects, cognitive, affective and, conative; and when we apply one of these three adjectives to any phase of mental process, we mean merely that the aspect named is the most prominent of the three at that moment (McDougall, 1928:266;2001:31).
In this context, the triple classificatory scheme is useful in the assessment of metal evaluations, since responses that express evaluation and so reveal attitudes should be divided into three classes (cognition, affect, and conation) and therefore each dimension should study and understand separately (Lavidge & Steiner, 1961;Insko & Schopler, 1967;Ostrom, 1969;Bagozzi, 1978;Holbrook, 1978;Hilgard, 1980;Breckler, 1984).In this regard, the three components also proposed for destination image by Gartner (1993); Dann (1996); Konecnik and Gartner (2007); Tasci et al. (2007) can be very useful for destination marketing organizations which are responsible for positioning, branding, communicating of destinations, and also building image, and developing marketing strategies for destinations.
The study results provide some practical implications.First, the findings show that cognitive destination image is consist of various attributes offered by the destination.Natural beauties such as, mountains, forests, valleys, lakes, rivers, and caves, etc., and make them attractive for tourists is an important manner.Cultural and historical sites, heritage, traditions, in other words the individuality and novelty of the destination and the promotion of the local culture is also significant.In addition, the urban infrastructure, tourism facilities and social facilities, easy access and transportation facilities contribute to the perception of cognitive image of destination.The attributes such as, safe and secure environment, cleanliness, convenience climate conditions, value for money that is offered by the destination should be also emphasized.In this way, it is proposed that emphasizing and highlighting the various attributes of the destination is important for destination marketing organizations, in order to determine the perceived destination image accurately and improve it within the intellectual, perceptual or cognitive component.
Second, the findings suggest that the attributes of the destination perceived by tourists affect their feelings against the destination.Additionally, the assessment of the cognitive and the affective components of destination image can be both used as a predictor of tourists' behavioral intentions toward destinations, such as intention to revisit the destination, intention to recommend the destination and intention to say positive things about the destination.Results also indicate that affective component partially mediated the relationship between cognitive and conative destination image.That is, tourists' perception of cognitive attributes offered by destination may have more powerful impact on willingness to response positively to the destination in the case of the tourist associates positive feelings to the destination, such as pleasure, excitement, arousal, and relaxation.
Based on these findings, it is suggested for destination marketing organizations that communication and promotion of the natural and cultural richness, and social envionment of the destination, providing improvements in general infrastructure, tourism facilities, and accessibility, making reasonable adjustments on prices for products and services are crucial for enhancing the cognitive destination image.In this way, it can be provided that visitors have positive feelings about the destination.However, destination marketing organizations need to take the cognitive and affective destination image together into consideration for developing strategies properly and therefore increasing the competitiveness of the destination.Thereby, positive behavioral intentions of visitors toward the destination can be increased.
There are some limitations related to the study which needs to be acknowledged.First, the research includes only one destination.Therefore, other studies should be carried out in different destinations, in order to validate the results and contribute to the generalization of the findings.Second, the study was conducted in the spring period, thus findings were limited to spring visitors.Tourists who travel in different seasons may constitute different beliefs and opinions about the destination or develop different feelings toward the destination.Futher researches should be performed in different seasons, in order to compare the results.Third, in the study the relationships between cognitive, affective, and conative destination image were analyzed merely.For future research, it is proposed that different variables such as information sources, perceived risks, or personal factors like travel motivations and socio-demographic characteristics that could be related to cognitive and affective components of destination image could be inserted into the proposed model for extending the study area.In addition, it is suggested that the effects of cognitive and affective components on overall image of the destination should be investigated for increasing the contribution of the study to the literature.Besides, testing the proposed model separately regarding to first-time and repeated visitors, or domestic and foreign tourists, since the destination images may differ within the different groups of tourists, may enhance the practical implications for the destination marketing organizations.

Figure
Figure 1.Model summary

Table 1 .
Variables and sources of scale items used for measurement

Table 2 .
Demographic characteristics of the participants (N=446)

Table 3 .
Results of exploratory factor analysis for cognitive destination image scale

Table 4 .
Results of confirmatory factor analysis for cognitive, affective, and conative destination image scales

Table 5 .
Reliability estimates of scales and descriptive statisticsHypothesis 1 was supported as a result of EFA and CFA.Hierarchical regression analysis was conducted to test the other hypotheses (Table6).Because they may affect the variables and relationships of interest, gender, marital status, age, and education were included as control variables.