This is the accepted version of the journal article: Escobar Rivera, Dalilis; Casadesús Fa, Martí; Simon, Alexandra. «Delightful tourism experiences : a cognitive or affective matter?». Tourism Management Perspectives, Vol. 32 (2019), p. 100569. DOI 10.1016/j.tmp.2019.100569 This version is available at https://ddd.uab.cat/record/288408 under the terms of the license This is the accepted version of the journal article: Escobar Rivera, Dalilis; Casadesús Fa, Martí; Simon, Alexandra. «Delightful tourism experiences : a cognitive or affective matter?». Tourism Management Perspectives, Vol. 32 (2019), p. 100569. DOI 10.1016/j.tmp.2019.100569 This version is available at https://ddd.uab.cat/record/288408 under the terms of the license Delightful tourism experiences: A cognitive or affective matter? Abstract There are many times that we, as tourist consumers, have been more than satisfied. We have been delighted with the product or service received. However, do we know the factors which affect our delight as customers? Are they affective or cognitive factors? The objective of this article is to empirically identify delighted consumers and the factors associated with these memorable tourist experiences when the threshold of satisfaction is exceeded. To this end, the authors used structural equation modelling to test a model based on a sample of 400 tourists obtained through a survey. The findings revealed that a consumer’s positive affective state of delight seems to be affected largely by cognitive-affective antecedents. Specifically, the cognitive dimension encompasses three main factors of the tourist service: the customer-service interaction, the staff and the availability of the service. However, the tourist´s happiness, as the main affective dimension, is also emphasised. Keywords: Customer delight, customer behaviour, tourist experiences, cognitive factors, affective factors 1. Introduction The subject of consumer delight (CD) has increasingly drawn researchers’ attention (Ahrholdt, Gudergan, & Ringle, 2017; Roberts-Lombard & Petzer, 2018; Torres & Ronzoni, 2018) whose studies are focused mainly in the tourism industry (Torres & Ronzoni, 2018). Despite different authors’ attempts to explain CD, the dimensions and variables that enable us to distinguish this behaviour from others, such as customer satisfaction (CS), remain unclear. Finn (2012) sees quite a clear difference between CS and CD, affirming that CD is a higher positive response in relation to CS with a higher impact on post-consumption behaviours, such as loyalty (Ahrholdt et al., 2017; Ali, Kim, Li, & Jeon, 2016; Kim, Vogt, & Knutson, 2013; Loureiro & Kastenholz, 2011; Ou & Verhoef, 2017). While Torres and Ronzoni (2018) argued that the difference is between the more affective nature of delight vs. the more attitudinal nature of satisfaction. Despite consensus, CD is limited by a lack of conceptual clarity, explained by a diverse set of theories and divergent empirical conclusions, e.g. in the case of joy and surprise. In addition, recent research focuses on how to apply basic emotions i.e. happiness (Nordhorn, Scuttari, & Pechlaner, 2018) by hierarchy to the appraisals associated with each one of them, as a possible better alternative to explain behaviours such as CD (Laros & Steenkamp, 2005) in remarkable experiences. Cognitive appraisals cover interpretations of an event’s characteristics combined to cause particular emotions (Watson & Spence, 2007). Besides, despite contributions, empirically, there is less research focus on providing a consistent measure of the construct (Torres & Ronzoni, 2018). This is also as a consequence of the lack of conceptual clarity. Studies explain CD as behaviour resulting from managing a customer’s needs or exceeding the consumer’s expectation positively to a surprising degree (Liu & Keh, 2015; Oliver, Trust, & Varki, 1996). Insights emerging from these first overviews could be grouped as cognitive approaches, understanding the processes of customer assessment by attitudinal and motivational theories (Torres & Ronzoni, 2018). In fact, Barnes, Ponder, and Dugar (2011) previously identified a cognitive route of delight. The second group of insights emerged from emotional approaches, as per Oliver et al. (1996) relating to emotions such as joy and surprise, as well as from several theories. At this point, one of the most interesting issues is that surprise is a key factor in achieving delight with opposing criteria among authors (Ali et al., 2016; Kumar, Olshavsky, & King, 2001; Torres & Kline, 2006). In order to explain emerging sources of consumer experience, positive relationships have been identified between emotional factors (i.e. happiness, exhilaration), cognitive factors (i.e. staff attitude, unusual ambience) and CD. Consequently, Kwong and Yau (2002) and Liu and Keh (2015) suggested a cognitive dimension to delight, while Liu and Keh (2015) and Barnes, Meyer, and Kinard (2016) also considered the emotional dimension. Once a point of maximum CS has been reached, an organisation’s resources should be directed to trying to delight the customer and not just on maintaining a certain standard level of quality within the touristic experiences. Authors assume that not all tourism experiences are memorable experiences (Kim & Ritchie, 2014) since memorable experiences are selectively reconstructed based on the individuals’ assessment (Kim, Ritchie, & McCormick, 2012) and individuals remember positive emotional events better than ordinary events that occurred equally long ago (Kim & Ritchie, 2014). Thus, given its importance for the future of companies worldwide and especially in tourism, the study aimed at examining the role of cognitive and affective attributes in memorable tourism experiences of customer delight, as well as its determinants. In turn, an exploratory model for high-quality service memories was developed and validated in order to deem the touristic experiences delightful. Consequently, the study contributes to better understanding how customers appraise unique experiences and what factors are most determinant. 2. Literature review and research framework 2.1 The cognitive approach to customer delight (CP) Tourism products are in essence experiential (Kim & Perdue, 2013) for both brand and competitive differentiation (Zhao, Yan, & Tat-Keh, 2018). The picture of what exactly makes certain experiences special is explored by Holbrook and Hirschman´s (1982) and Pine and Gilmore´s (1998). These studies emphasize the transition from service delivery to experience design, arguing that satisfaction is one component of experience (Tung & Ritchie, 2011). An individual who is engaged emotionally, physically, spiritually, and intellectually interprets the memorable experience subjectively (Tung & Ritchie, 2011). According to Kim and Perdue (2013:246), this experience is created through the interplay of cognitive, affective and sensory choice. Thus, the tourism experience is memorable when it is positively remembered and recalled by the tourist (Kim et al., 2012). Kim and Perdue (2013:247) define the cognitive attributes as "physical dimensions or non- physical qualities that satisfy utilitarian needs and provide functional benefits or value", i.e. the price, value for money or interpersonal service. Cognitive attributes are instrumental in fulfilling functional goals (Kim & Perdue, 2013) through rational and practical attributes that are purchased for their utility based on consumers’ needs. As such, the cognitive paradigm (CP) of CD is based on two extended theories: the expectation disconfirmation theory and the needs theory. As Kim and Perdue (2013) highlight, the cognitive attributes are the “foundation” of customer experience because the rational system of the cognitive-experiential self-theory (CEST) (Epstein, 2003) is basically cognitive in nature. The rational system emphasises the practical and functional meaning of the touristic experiences with less emotional attachment. Thus, the cognitive attributes in CD reflect an attitudinal or motivational trend. The attitudinal trend is based on expectation measures understood through the tolerance zone’s interpretation. The tolerance zone concept was adopted to identify CD beyond CS by Keiningham, Goddard, Vavra, and Iaci (1999) and explain the difference between the desired service and the level of service considered adequate (Parasuraman, Zeithaml, Valarie, & Berry, 1988). Applying this theory to delight, when the consumer’s expectations are exceeded and they are pleasantly surprised then this is delight (Ali et al., 2016; Torres & Kline, 2006). Consequently, most of the conceptualisations following this logic equate high indices of CS with CD (Arnold, Reynolds, Ponder, & Lueg, 2005; Shneider & Bowen, 1999), therefore not providing sufficient empirical evidence for the identification of delight. Likewise, recognised cognitive factors in CD come from models such as SERVQUAL (Parasuraman et al., 1988). Liu and Keh (2015), for instance, include justice expectations in the measurement scale they propose. While Barnes et al. (2011) emphasise the cognitive route to CD, summarising four avenues to cognitive delight centers: employee effort, employee skills, the core product (CD occurs because of the value inherent in the product) and service recovery (because of the dominance of disconfirmation theories). Thus, explorative studies focusing on discovering the factors that create delight in the customer´s mind and positively correlating independent cognitive attributes, such as unusual ambiance, problem solving gestures, caring behaviour, feedback and VIP treatment with CD, place emphasis on the importance of the human factor in being delighted (Kumar et al., 2001; Preko, Kwami, & Feglo, 2014). Some of these variables have also been studied before as tangible assets of the organisation (Hasan, Raheem, & Subhani, 2011; Kwong & Yau, 2002; Liu & Keh, 2015; McNeilly & Feldman, 2006; Preko et al., 2014; Torres, Fu, & Lehto, 2014). CD being based on customers’ needs is the second theory and comprises a basically motivational meaning (Torres & Ronzoni, 2018), influenced by previous experiences. This theory suggests that the extent of fulfilment of three basic needs: security, justice, and self- esteem (Shneider & Bowen, 1999) gives rise to CD. However, recent studies (Kim, Kim, & Kim, 2017) introduce a set of complementary needs including hedonic needs (i.e. pleasure attainment and pain avoidance), eudemonic need (i.e. personal growth, and self-actualisation) and social need (i.e. feeling connected with others). Kim et al. (2017) for instance, present important findings precisely arguing the positive impact of hedonic needs fulfilment in pleasurable events (Kim et al., 2017; Ma, Gao, Scott, & Ding, 2013) because hedonic needs are linked to power and affiliation recognition behaviours in society. In this regard in particular, previous studies seem to be focused on identifying cognitive attributes in relation to delight in memorable experiences, rather than examining whether cognitive attributes allow identification or affect the evocation of delight. Similarly, prior research examined the extent to which expectations or needs influenced CD and was mostly associated with the affective route of CD (Barnes et al., 2011), rather than verifying whether cognitive attributes have a more significant effect on consumers’ delight appraisal compared to emotional attributes. The lack of consensus about what theory better explained CD affected the development of operational measures and is considered a critical issue in the field (Torres & Ranzoni, 2018). Additionally, previous studies did not investigate the combined effect of cognitive attributes generated in customers’ delightful experiences in order to further verify the nature and interaction of such relationships, especially as applied to the field of tourism. Thus, the following hypothesis is formulated: H1: Cognitive factors will positively influence a consumer’s delight in their most memorable tourism experiences 2.2 The affective approach to customer delight (AP) Since the first definitions of CD, its relationship with emotional factors, such as joy and surprise, has been evident. Given this basis, the relationship between joy and surprise has been one of the most studied because while the rational system is cognitive in nature, the experimental system is affective. This picture is essential as it sheds light on the main conceptualisations of CD in tourism experiences. The experimental system can be interpreted as hedonic consumption to satisfy emotional and sensory needs based on what customers expect will be pleasurable (Kim et al., 2017). Affective reaction is not static but has its own temporal course depending on whether or not a stimulus is present and when it began (Watson & Spence, 2007), which explains why when an emotional instigator like surprise occurs frequently, the organism displays habitual affective reactions and so greater stimuli are required to produce equivalent emotions in terms of intensity. Thus, Kumar et al. (2001) concluded that if customers feel excitement and pleasure, they may not need to feel a high degree of surprise to be delighted. On another hand, conceptualisations of CD according to Liu and Keh (2015) talk about the single or multi-item emotional definition of delight. Among these conceptualisations are: delight as the combination of delight, elation and glee (Ali et al., 2016; Barnes, Ponder, & Hopkins, 2015; Finn, 2012); delight as thrill, exhilaration and joy (Kumar et al., 2001) and, delight as a combination of enchantment, enthusiasm, excitation, cheer, astonishment and joy (Ball & Barnes, 2017; Ma, Scott, Gao, & Ding, 2016; Loureiro, 2010). Apparently, literature is focused on a few emotions considered as direct antecedents of CD (Torres & Ranzoni, 2018). According to Laros and Steenkamp (2005:1437), whereby the main issue with using emotional theories to explain customer behaviors such as delight, is “the structure and content of emotions”. The two-dimensional character of emotions justifies research where authors study both sides, such as Liu and Khen (2015), studying delight and outrage. The two-dimensional character comes from the definition of two elements common to affects: activation level (high/low) and affective valence (Watson & Spence, 2007). However, even when the literature recognises opposing states of affectivity, it is not effective in measuring them at the same time (Laros & Steenkamp, 2005). To illustrate, the so-called appraisals theory explains an individual’s emotional response for each one through different and opposing appraisal assessments (Liu, Sparks, & Coghlan, 2016). Appraisals include a range of dimensions, such as goal relevance, goal congruence, coping potential and future expectancy (Liu et al., 2016). Therefore, to study only part of these emotions to explain a complex concept like delight is not sufficient. Therefore, Laros and Steenkamp (2005) integrate emotions in Relational Marketing (RM) studies in a model with three levels, considering a large number of contributions in the literature, including Plutchik (1980) and rethinking how to evaluate feelings based on basic emotions. Basic emotions such as fear or happiness are universal (Nordhorn et al., 2018) and have been related to a dimensional approach, however there are different ways to conceive emotions and there is also disagreement about which emotions are basic (Laros & Steenkamp, 2005). Laros and Steenkamp´s study concludes that the best way to obtain information from customers about what they feel during a consumer experience is through a first level of emotive generalisation (based on three levels), which implies distinguishing between positive and negative affectivity. In the second level, the model considers four basic emotions included in positive affectivity (contentment, happiness, pride and love), grouping the appraisals that describe each of them. In this line, Nordhorn et al. (2018) also explores customers’ emotions (happiness, sadness, surprise and anger), finding that happiness is achieved if service and relationship quality are at higher levels according to customers. Despite individual emotional response expressed in basic emotions based on the appraisals theory, Diener and Seligman (2002) conclude that very happy people do have a functioning emotion system that can react appropriately to life events, i.e. in a really good service encounter, they tend to externalise those emotions. Evidences also points to emotions related to hedonic consumption and characterised by aroused positive affect (Ma et al., 2013). This author explains CD as being related to appraisal of personal well-being or special needs (Collier & Barnes, 2015). Well-being, wellness, or qualities of life in tourism studies are treated as terms similar to the universal emotion of happiness, receiving little attention in touristic experience studies (Filep & Deery, 2010). In fact, Filep, Laing, and Csikszentmihalyi (2016) propose to develop a new sub-field entitled “positive tourism” to study human emotions in the tourism context beyond the interpretation of positive psychology as pleasure, and to study human well-being. Well- being can be defined and interpreted based on hedonic terms, which means that it is temporary or, as Filep and Deery (2010) propose, based on positive psychology the positive emotions (i.e. love, interest, joy, contentment), engagement relationships, meaning and achievement, which in turn evoke the authentic happiness state in tourists. Thus, considering these theories, the following hypothesis is proposed: H2: Positive emotions will positively affect the customer’s delight in their most memorable tourism experiences 2.3 The link between the cognitive and affective paradigms. The authors suggest that delight should be explained from two basic approaches, cognitive and affective. The cognition-emotion mix is underexplored in delight literature, although the concept has been understood as high emotion and positive cognition (Kwong & Yau, 2002). Thus, both will be considered as essential dimensions in identifying delight in consumers. However, are the cognitive and affective dimensions related in customers’ delightful touristic experiences? RM reflects how consumers assess experiences with emphasises on affective and cognitive aspects of the service in memorable experience creation (Alnawas & Hemsley-Brown, 2018). Holbrook and Hirschman (1982:137) define consumer experiences as whole events experienced by a person, where goods and services consumed often affect emotions. From here, many authors attribute the customer’s distinct appreciation to cultural, environmental, sensorial-emotional, pragmatic, lifestyle and relational dimensions (Gentile, Spiller, & Noci, 2007) and other previous experiences (Andajani, 2015; Liu et al., 2016). After all, the final appraisal of a service experience is related to well-being based on needs and emotions (Torres & Kline, 2006). In the existing tourism literature, researchers have identified a set of experiential components, for instance, hedonism, happiness, pleasure and stimulation, giving further attention to those dimensions likely to be powerful in tourists’ memory formation (Kim & Ritchie, 2014). In fact, Kim and Ritchie (2014) highlight the impact of positive emotions and feelings associated with these experiences, such as happiness and excitement, in the so called most memorable experiences, because, according to previous studies development by Tung and Ritchie (2011), people primarily seek enjoyment (i.e. hedonism, pleasure) in tourism experiences. Each situation will be appraised using dimensions related to that particular situation and the discrete emotional reaction will elicit an emotion (Ma et al., 2013). In discussing delight studies, researchers point to considering delight as a combination of arousal and pleasure, primarily consumed for hedonic purposes (Hosany, Prayag, Deesilatham, Cauševic, & Odeh, 2015) meaning that tourism studies recognise that tourism experiences have predominantly hedonic components (Kim & Ritchie, 2014). In general, RM literature demonstrates that delighted consumers are emotional decision makers who are less affected by utility when going through a decision path process or successive stages, in which the post-consumption behaviour is but the final stage. However, delightful experiences have also been related to cognitive attributes, meaning that the consumer is a rational decision maker too when it comes to service enjoyment, selecting the one that meets their needs. Given this background, this article investigates the role of cognitive factors guiding affective reactions in order to evoke delightful tourism memories. That is why, according to Ludwig, Barnes, and Gouthier (2017), the most important consequence of delight is the creation of exceptional value. Thus, the cognitive and experiential nature of CD has not been well represented and identified in previous conceptualisations. In light of the fact that both dimensions can provide information about the tourist´s experiences and that their appraisal will depend on the interaction of both dimensions, the next hypothesis states: H3: Cognitive factors will activate positive emotions in the affective dimension that influence the consumer’s delight’ in their most memorable tourism experiences The general hypotheses are represented in the research model in Figure 1: --------Insert Figure 1 about here--------- The three hypotheses are depicted diagrammatically in the conceptual model (see Fig. 1). These relationships will be tested in the following empirical study. The literature review serves to identify and group the most significant variables and factors that make up the model (see Appendix I). Appendix I summarises factors, variables and items in each component of the model, specifying origins. 3. Methodology 3.1 Survey instrument Studies of operational measures for delight are much fewer. However, a few measures have been detected (Hasan et al., 2011; Kwong & Yau, 2002; Liu & Keh, 2015; Torres & Ronzoni, 2018). The most concrete proposal is from Liu and Keh (2015), although two other proposals have not been empirically tested to date: Kwong and Yau (2002) and Torres and Ronzoni (2018). Liu and Keh´s (2015) study serves as reference for other scales, although studies like Laros and Steenkamp (2005) suggest not assessing opposite behaviours like delight and outrage at the same time in order to avoid the lack of sufficient empirical evidence. An extensive review of both cognitive and affective literature enabled identification of the main dimensions and the items that can be included in the study (see Appendix I). All the variables and factors were initially gathered. Next, these factors were integrated and reduced in number, retaining the most significant that best corresponded to the conceptualisations of CD. A panel of experts assessed the appropriateness of removing or rewriting some of the items in each interaction in order to adapt them to the context. Hence, the survey with a view to identifying CD was then developed. Once the initial group of items had been obtained, a small pilot test was undertaken using a sample of 34 individuals to test all the items in the survey. The pilot-test aimed to refine the measurements and check the results in the objective sample (i.e. degree of difficulty, extent of the survey, writing the declarations, answer rate etc.) (Cristóbal-Fransi, Hernández-Soriano, & Marimon, 2017). This analysis permitted a further reduction which was definitive. Besides, in relation to the items, the degree of dispersion and the correlations in the results of the measurements were analysed as recommended by Cristóbal-Fransi et al., (2017). The final operational measure groups attributes recognised in the literature into two main dimensions: cognitive and affective. The cognitive dimension encompasses eight factors: tangibles, customisation, organisational staff, justice, trust, esteem/self-esteem, the culture of CD and feedback. While, the affective dimension identifies three factors from which customers elicit CD: joy, happiness and love from basic emotions, by adapting the distribution of Laros and Steenkamp´s (2005) model to facilitate integrating and reducing items. Following the recommendations of Laros and Steenkamp (2005), pride will be tested in future studies. Each basic emotion is related to a set of appraisals recognised in previous literature. The affective dimension also includes three factors explaining hedonic consumption, social value and appraisal of the experience previously identified in the literature (see Appendix I). The survey was designed as a 35-item survey. Sixteen of these items assessed the cognitive dimension through the eight factors and fourteen of them assessed the affective dimension in five factors. Delight was directly measured by 5 items, leaving the emotions in a different dimension and considering the consumer experience. One item was a joint assessment of both aspects entitled “appraisal within the affective dimension with three variants: (1) satisfaction, (2) extreme satisfaction and (3) delight”. Lastly, three demographic information items were included. 98% of the scales were Likert attitude scales (1-5) and the remaining 2% were items where one of three possible variants had to be selected. There were three variants of the specific scales for each item: (1) the cognitive attributes were assessed on a scale from 1 to 5 (1-Not good, 2- Acceptable, 3- Good, 4- Excellent and 5-Extraordinary); (2) the affective attributes were also assessed from 1 to 5 (1-Nothing, 2-A Little, 3- Moderately, 4- A Lot and 5-Absolutely) and; (3) CD was assessed from 1-Completely disagree to 5-Completely agree. From collected data, the exploratory factorial analysis was performed on the items, using the principal component method, with Varimax rotation to see their correspondence with the number and distribution of the dimensions of the proposed model and the number of factors obtained (Cristóbal-Fransi, et al., 2017). Having obtained and studied these subscales, a reliability analysis was performed as prior step to the confirmatory factor analysis of the dimensions and observable variables (items) obtained. Finally, authors proceed with the validation and the estimation of the proposed model and the causal relationships using Structural Equation Modelling (SEM). 3.2 Data collection In this research, the sample frame includes tourists who visited Girona (Spain) during their trip. The data collection process was developed by means of a self-administered survey in a website created using the tools Survey Monkey and Google Site, and administered by well- trained field researchers. Interviews were conducted face-to-face at the Tourist Information Office in the city of Girona. The authors asked respondents to recall a memorable experience they had had during the previous two years, suggesting some options based on previous studies, but also allowing them to choose other experiences. Once they had selected the experience, the tourist had to keep it in mind while they responded to the survey. The interviews and the online-survey were conducted in a structured manner utilising a bilingual (English and Spanish) survey instrument. To avoid the negative impact of any language barriers, the survey was developed in English, considering that the questions previously made in the literature were then translated into Spanish, and then revised by two native speakers. The respondents were tourists in Spain who had experiences in hotels, restaurants, spas, water parks, natural parks, shopping in stores or malls, among others in the last two years. The sample was comprised of n= 400 individuals, even though 410 questionnaires across the two variants were completed, only 97.56% of these were valid. The high response index was due to the online questionnaires, which represented 24% of the total, and the face-to-face surveys conducted in a comfortable environment by the research team itself. In the sample, 54.5% (218) were female, and 45.5% (182) were male, 64.25% were 18-49 years of age, and 35.75% were 50 or more years of age. Most of them (72.50%) were from Europe, 20% from the Americas, and the 6.8% from Asia, Africa and Oceania. Table 1 and 2 show the frequency distribution of the demographic variables. --------Insert Table 1about here--------- --------Insert Table 2 about here--------- 4. Results 4.1 Assessment of scales. The principal components analysis (PCA) using the Varimax orthogonal rotation method was separated by paradigm according to the research model. The PCA is one of the most widely used methods to reduce the dimensionality of a dataset, while preserving as much statistical information as possible, finding new variables defined by the dataset at hand, not a priori (Jollife & Cadima, 2016). Based on this, and on the correlations matrix, two tests were conducted: Bartlett’s test of sphericity and the sampling adequacy Kaiser–Meyer–Olkin (KMO) index (KMO). The measure of the KMO index was 0.895, which was greater than the recommended value of 0.7. Bartlett’s test of sphericity was 2463.198 (df = 210) with a significance of 0.000. These results confirmed a linear dependence between the variables and supported our view that the results were sound. Based on all of the above, the authors proceeded to remove those items that showed high secondary loads and whose items were quite similar as suggest Cristóbal-Fransi et al. (2017). Thus, CDT1, CDT2, CDJ2, ADH5, ADL1, ADC1, ADSV1 and ADSV2 present only high secondary loads, as well as CDP1, CDC2, CDQ1, and CDF1 despair, considering one dimension or another. After the three exploratory analyses by subscales, variables were grouped in six factors: three factors in the cognitive dimension (staff, interaction with the service and service availability) assessed by 9 items, two factors in the affective dimension (happiness and hedonic consumption) assessed by 6 items, and the delight factor assessed by 3 items, explaining the 67.84% of the total variance. Table 3 shows the load of each factor by dimension. In all cases, the values were shown to be significant, i.e. either greater than or equal to 0.50, as recommended by Bernardo, Marimon, and Alonso-Almeida (2012). Four basic criteria were considered to eliminate the items by dimension: (1) the items that loaded below 0.60 in each of the factors; (2) the communalities with values below 0.50; (3) inter-item correlations below 0.30; and (4) item- dimension correlations below 0.40, as considered in Bearden, Hardesty, and Rose (2001). In all cases, the inter-element correlations were within the range of 0.30-0.50 and the correlations were greater than 0.30. Even though the KMO value in the component that directly measured CD was within the recommended limit (Bernardo et al., 2012), only the three items that measured this component explained 70.28% of the variance and they presented the highest inter-item correlation of 0.53% and the highest inter-dimension correlation of 0.55. --------Insert Table 3 about here--------- The three factors in the cognitive dimension (see Table 3) explain 64.83% of the variance of the sample. The first (16.36%), which was labelled staff includes five items; two items from staff, one item from customisation, one from trust and the item designed to measure the appreciation of esteem/self-esteem needs (see Appendix I). The second factor, which consisted of two items related to how the service responded to customer request, was labelled interaction with the service and built on one item from justice and one item from feedback (see Appendix I). The third factor in the cognitive dimension was labelled availability and includes two items from tangibles describing a service as ‘accessible as consumer need’. The three factors in the affective dimension (see Table 3) explain 73.62% of the variance of the sample. The first factor (15.42%) was labelled as happiness and includes the first four items from happiness (see Appendix I). Variables from the remaining basic emotions identified (love and contentment) were not significant, thus we kept the label of happiness. Happiness is shorthand for an individual’s personal appraisal of full enjoyment in the experience. The second factor was kept labelled as hedonic consumption, describing reactions to the externalisation of the experience with existing and future consumers. The differentiation of these two factors suggests that happiness and hedonic consumption are independent constructs. Finally, the delight factor (13.04%), referenced in Table 3 as CD, includes three of the five items initially identified from delight. Three factors: staff, happiness and CD explained 44.82% of the variance of the sample, while the other three factors: accessibility, interaction with the service and hedonic consumption explained the remaining 23.03% of the total variance. Reliability was assessed using the Cronbach’s alpha (0.877) test and composite reliability in all cases exceeded the recommended value of 0.7 for internal consistency. The same coefficient for the six factors ranged from 0.78 to 0.91, which is also indicative of good reliability among the variables within each dimension. The general RHO correlation coefficient was 0.898 and in almost all cases, the extracted variance was within the recommended limit of 0.50. Table 4 shows the results of the confirmatory factor analyses. --------Insert Table 4 about here--------- Convergent validity was determined by observing the t-values (t > 2.58), which had a high weighting in all cases. Discriminant validity was analysed by comparing the linear correlations and standardised covariances among the latent factors, checking that the correlations were less than the square root of the extracted variance (AVE). Table 5 shows that the square root values for each AVE were greater than the elements of the diagonal. In this case, the extracted variance test did not completely guarantee the discriminate validity, even though the values were all close to the recommended value of 0.50. This may be because this is the first time a model with these characteristics has been used to measure the construct and also, because the sample included customers who described experiences in very different services which may have been affected by different variables. However, this diversity of experiences is essential to determine the overall contingency factors in tourism experiences. --------Insert Table 5 about here--------- Despite low AVE coefficient, the authors agree with Borsboom, Mellenbergh, and Van Heerden’s (2004) criterion about the concept of validity giving much more importance to goodness of fit and composite reliability measures, which in our case showed acceptable values. In this case, following Borsboom et al.’s (2004) criterion, this issue will be mentioned in the limitations. Nonetheless, taking into account values for the extracted variance, the authors tested two CFA nested models to determine the discriminate validity of each dimension in relation to the others. The test consisted of fixing the correlation at 1 in one of the models, while the other calculated the correlation among the dimensions. Validity was verified by means of the Chi-square test. If the variation of the Chi-square was significant (p ˂ 0.05), as it was for almost all the factors in the model, this would indicate that the models were not equivalent; in other words, where the correlation coincided, the model would not adequately explain the distribution of the data. The same items, therefore, cannot be used in the two models to measure the dimension. This test allows for determining the validity of the scale, as recommended by Beltran-Martín, Roca-Puig, Escrig-Tena, and Bou-Llusar (2009). The only distinction in the proposed model was between the factors happiness and hedonic consumption. According to the results (p ˃ 0.05), they could be a single factor. However, the authors consider that the variables included in each of these factors are different in terms of content and so they must be separated. Discriminant validity was determined using three methods: the confidence interval of the correlations (Anderson et al., 1988), the comparison of two nested CFA models for each pair of factors (Beltran-Martin et al., 2009) and the extracted variance test against the correlations (Fornell & Larcker, 1981). Even though the last was not sufficiently consistent in the case of the percentage of variance explained by factor, the other two tests demonstrated its validity. On the other hand, CR indicated strong construct reliability in all cases except for availability and external happiness; all values were above or around 0.7 as recommended by Fornell and Larcker (1981). Indices established convergent validity and all the t-values were > 2.58, thus fits are acceptable, especially given the multi-service experiences within the sample. 4.2 Causal model The external model is comprised of six latent first-order reflective factors and eighteen variables, where the indicators are a reflection of the observed construct and they condition its variation (See Figure 2). This measuring model was tested using the robust maximum likelihood method from the asymptotic variance-covariance matrix using the EQS 6.3 software package. The fit indices obtained indicated that the goodness-of-fit of the variables in the factors determined in the CFA was consistent (see Table 4) in accordance with the propositions made by Hu and Bentler (1999). The x2 Satorra–Bentler scaled chi-square was 231.43 with 123 degrees of freedom and the p-value was 0.000. X2/df = 1.89, which was somewhat below the accepted limit of five. The RMSA was 0.057 and the CFI was 0.919. The result of the Cronbach's Alpha test (0.877) showed that the scale was reliable and, in all cases, the coefficient for composite reliability exceeded the recommended value of 0.70. The general RHO correlation coefficient was 0.898. Table 6 shows the results of the model. Considering the general indices and approaching the specific analyses of the dimensions with caution, the model is sufficiently consistent. However, some low factor loadings (less than 0.7) were observed. In light of that, Hair, Anderson, Tatham, and William (1998) posit that factor loadings as low as 0.3 are acceptable provided that the sample is more than 350 cases, and that in our case the sample covers 400 cases, we kept the items that loaded above 0.50. --------Insert Table 6 about here--------- Appraisals were included in the study variables as controls and to describe three different consumer assessments. Variant A (see Appendix I) could be considered the appraisal of a satisfied consumer; variant B as the appraisal of a much more than satisfied consumer; and variant C as the appraisal of a delighted consumer. Furthermore, considering that the appraisal is based on emotional theories (Watson & Spence, 2007), a relationship was established between this variable, CD and the affective dimension. Table 7 shows the results of the model after introducing the control variable. The model in Table 7 has similar goodness-of-fit in the general indices as the previous one, although some points were lost in the CFI (while remaining within the recommended values) and some degrees of freedom were gained (x2 Satorra–Bentler chi-square test= 302.899/df = 1.92). Both the measuring model and the measuring model including the control variable gave acceptable results and confirmed the applicability of the measurements. According to the demographic distribution of the control variables in Table 2 and their variance, the authors decide to test the power of the coefficient of determination of regression (R2) following the recommendations of Fornell and Larcker (1981). Appendix II shows the results of the analysis. --------Insert Table 7 about here--------- Based on examination of a wide range of service experiences, the resulting model (see Figure 2) suggested that both cognitive-affective dimensions significantly affect CD. The direct impact of the cognitive dimension on CD was not significant (p ˃ 0.05); however, there is a relationship between both components through the affective dimension. Regarding the affective dimension, the correlation coefficient (0.999, p < 0.01) demonstrated how important this factor is in identifying delighted consumers. The affective dimension of the model presents a considerably high (0.975) and significant (p ˂ 0.05) correlation with CD. The correlation of both factors in the affective dimension was significant and relatively similar, even though happiness’s construct had a greater impact on the variability of the affective dimension. Meanwhile, the latent factors in the cognitive dimension suggested a high value index among the consumers with two basic factors: interaction with staff (0.802, p ˂ 0.01) and service availability (0.661, p ˂ 0.01). The results in the causal model and in the causal model after introducing the control variable (appraisal) demonstrate the significant positive relationship (0.351, p < 0.01) between CD and the affective dimension (0.419, p < 0.01). These results suggest that this component is an essential factor in a memorable experience of delight. Appraisals can differentiate between groups of customers by means of a gradual evaluation of their experiences, including both dimensions of the model in their description. The model was also tested by analysing the mediation effect of the affective dimension using the EQS software. The model includes three direct effects and one indirect effect among variables according to the classical mediation distribution with one variable. Despite this, the mediation analysis is not the main objective of the study; results are in Appendix III, as well as the values for the Sobel test application. The results in the causal model and in the causal model with the control variable suggest that there are three variables which are useful in distinguishing CD from other behaviours (exceptional value, excellence and positive emotions). In this case, the three variables presented medium-high loads. Excellent service is the item with the greatest load (0.803), indicating that consumers are willing to pay for what they consider as benefits or more convenience from a service. Furthermore, an item was included to see what the exceptional value during the service experience was. Even though the extracted variance test showed that the CD factor did not extract the recommended values of 0.50, the correlation of the individual and group variance was acceptable. The nested CFA models test demonstrated the difference between the chi-square of this factor and the remainders in the model. --------Insert Figure 4 about here--------- A Lagrange test was applied in the latent factors model, which suggested determining two correlations to improve the goodness-of-fit indices. The first of these was a significant correlation between the errors of the positive affect variables and exceptional value (p ˂ 0.001, value=0.399) and the second was between the variable appreciation of customers’ needs (extremely concerned about my specific needs, in the survey) and CD (p ˂ 0.001, value=0.278), so this item could correspond to both factors. 5. Conclusions The research helps to increase the understanding of the path of CD in tourist’s memorable experiences. The study identifies that a combination of cognitive-affective dimensions had a positive influence on the experience of consumer delight; however, both paths do not have the same effect. Findings evidence that the chain from cognitive factors to affective factors and then to delight evocation is significant in the context of memorable tourism experiences. Findings show that this path includes engaging happiness and hedonic consumption aspects, presented in positive tourism theory in order to extend existing research on the identification of delightful and most memorable experiences and the effect of positive emotions in tourism experiences. It should also be noted, that the study obtained three factors in the cognitive dimension which represent characteristics of excellent service delivery in interaction with consumers. In this line, findings are in accordance with recent studies like Nordhorn et al. (2018), whose results reveal that consumers respond more positively in emotional terms to a very good service, giving as much importance to the intangibles attributes as to the tangible attributes in the elicitation of emotional reactions (Nordhorn et al., 2018). The tested model results do not entirely support the hypothesis of a direct relationship between the cognitive dimension and delight. However, the authors consider that the hypothesis is partially supported because findings confirm that different attitudes of consumers toward cognitive factors can affect the evocation of the affective dimension, in turn affecting the elicitation of delight. The affective dimension had a greater impact on consumers’ exceptional experiences. In the affective dimension, the relationship and composition of its two constructs are interesting since the study obtained only two factors taken to represent constructs of happiness and hedonic consumption, which describe all variables related to happiness in Laros and Steempak’s model (2005).The literature presents two dominant happiness theories from positive psychology to explain tourist happiness: the subjective well-being (hedonic conception) and the authentic happiness theory. In this regard, Kim et al. (2012) highlight that the memorable tourism experience components of hedonism are more likely to increase behavioural intentions in tourists. Nevertheless, the present study suggests that both factors (hedonic consumption and happiness), but especially happiness in terms of positive emotions, make the experience delightful. Additional insights are also gained in identifying incremental states (satisfaction, more than satisfaction and delight) emerging simultaneously from the applied measures. The study reinforces Torres’s criteria for the affective nature of CD in response to the attitudinal nature of CS (Torres, 2014), although it is also apparent that both cognitive and affective dimensions contribute to evoked delight. It should be emphasised, however, that it remains unknown whether the perceived cognitive effect of the event will last over the affective one, or if other interactions between both dimensions occur. From a theoretical point of view, the affective dimension was firstly used with a different perspective, analysing the basic emotions in consumers’ appraisal formation, which implied removing traditional emotions leading to customer delight to a separate dimension and including other variables to measure delight directly. This study relies on the fact that the most memorable experiences of delighted consumers tend to be more service-oriented, where long-term and emotional bonds are maintained with customers in order to co-create memorable experiences (Zhang, Lu, Torres, & Chen, 2018) and features of the service experience related to delight (Torres & Ronzoni, 2018). Secondly, this study recognises the importance of the consumer experience concept (Harrington, Hammond, Ottenbacher, Chathoth, & Marlowe, 2018) and positive tourism for delight identification based on the individual attitude of consumers. This statement could mean that basic emotions can provide more information and can distinguish between ascending consumer states. Based on the findings, it is proposed that happiness as a basic emotion is a major mechanism in the identification of delight. The results indicate that service providers, as well as costumers should pay attention to customising exceptional experiences for future tourism success, thinking in terms of individual consumers’ happiness. Therefore, service providers should focus on customising the service carefully. Additionally, these findings demonstrate that service suppliers should be aware that, with the objective of achieving the CD, they can affect the current appraisal and future appraisals based on created memories. This study has some limitations; firstly, it includes a group of service experiences that customers can evaluate very differently. The results suggest that appraisal is a very important variable and it is considered a limitation to only use one item in this variable, even though it was not the aim of this study to prove this theory. Another limitation is the results of the extracted variance test for the factors, even though discriminant validity was proven by two different methods and some fit indices that can be interpreted as not sufficiently strong, however the model is consistent. The study strengthens the idea that tourist happiness is more than just pleasurable events, even if more research is needed to better understand all the components of the conceptualisation of tourist happiness. In relation to this aspect, this study only measured the delightful state of the most memorable touristic experiences perceived by the participants. Besides, the authors consider that more research is needed to examine the antecedents and interactions between excellent service attributes in the service delivery process and CD. Declarations of interest The authors declared no potential conflict of interest. References Ahrholdt, D. C., Gudergan, S. P., & Ringle, C. M. (2017). Enhancing service loyalty: The roles of delight, satisfaction, and service quality. Journal of Travel Research, 56(4), 436 –450. Ali, F., Kim, W. G., Li, J., & Jeon, H.M. (2016). Make it delightful: Customers’ experience, satisfaction and loyalty in Malaysian theme parks. Journal of Destination Marketing & Management, 7, 1–11. Alnawas, I., & Hemsley-Brown, J. (2018). The differential effect of cognitive and emotional elements of experience quality on the customer-service provider’s relationship. International Journal of Retail and Distribution Management, 46(2), 125–147. Andajani, E. (2015). Understanding customer experience management in retailing. Procedia - Social and Behavioral Sciences, 211, 629–633. Anderson, J. L., Kellogg, J. C., Gerbing, D. W., Brett, J., Fornell, C., Larcker, D., Perreault, W., & Steiger, J. (1988). Structural Equation Modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411–423. Ariffin, A. A. M., & Yahaya, M. F. (2013). The relationship between airport image, national identity and passengers delight: A case study of the Malaysian low cost carrier terminal (LCCT). Journal of Air Transport Management, 31, 33–36. Arnold, M. J., Reynolds, K. E., Ponder, N., & Lueg, J. E. (2005). Customer delight in a retail context: Investigating delightful and terrible shopping experiences. Journal of Business Research, 58, 1132–1145. Ball, J., & Barnes, D. C. (2017). Delight and the grateful customer: beyond joy and surprise. Journal of Service Theory and Practice, 27(1), 250–269. Barnes, D. C., Meyer, T., & Kinard, B. R. (2016). Implementing a delight strategy in a restaurant setting: The power of unsolicited recommendations. Cornell Hospitality Quarterly, 57(3), 329–342. Barnes, D. C., Ponder, N., & Dugar, K. (2011). Investigating the key routes to customer delight. Journal of Marketing Theory and Practice, 19(4), 359–376. Barnes, D. C., Ponder, N., & Hopkins, C. D. (2015). The impact of perceived customer delight on the frontline employee. Journal of Business Research 68(2), 433–441. Bearden, W. O., Hardesty, D. M., & Rose, R. L. (2001). Consumer self-confidence: Refinements in conceptualization and measurement. Journal of Consumer Research, 28(1), 121–134. Beltran-Martín, I., Roca-Puig, V., Escrig-Tena, A., & Bou-Llusar, J. C. (2009). Internal labour flexibility from a resource-based view approach: Definition and proposal of a measurement scale. International Journal of Human Resource Management, 20(7), 1576–1598. Bernardo, M., Marimon, F., & Alonso-Almeida, M. D. M. (2012). Functional quality and hedonic quality: A study of the dimensions of e-service quality in online travel agencies. Information and Management, 49, 342–347. Borsboom, D., Mellenbergh, G. J., & Van Heerden, J. (2004). The concept of validity. Psychological Review, 111(4), 1061–1071. Chena, S.C., & Lin, C.P. (2015). The impact of customer experience and perceived value on sustainable social relationship in blogs: An empirical study. Technological Forecasting & Social Change, 96, 40–50. Collier, J. E., & Barnes, D. C. (2015). Self-service delight: Exploring the hedonic aspects of self-service. Journal of Business Research, 68(5), 986–993. Collier, J. E., Barnes, D. C., Abney, A. K., & Pelletier, M. J. (2018). Idiosyncratic service experiences: When customers desire the extraordinary in a service encounter. Journal of Business Research, 84, 150–161. Cristóbal-Fransi, E., Hernández-Soriano, F., & Marimon, F. (2017). Critical factors in the evaluation of online media: creation and implementation of a measurement scale (e-SQ- Media). Universal Access in the Information Society, 16(1), 235–246. Diener, E., & Seligman, M. E. P. (2002). Very happy people. Psychological Science, 13(1), 81–84. Ekinci, Y., Dawes, P. L., Massey, G. R., & Marketing, I. H. (2008). An extended model of the antecedents and consequences of consumer satisfaction for hospitality services. European Journal of Marketing, 42(1/2), 35–68. Epstein, S. (2003). Handbook of Psychology (1st ed.). New Jersey: Wiley & Sons (Chapter 7). Filep, S., & Deery, M. (2010). Towards a picture of tourists’ happiness. Tourism Analysis, 15(4), 399–410. Filep, S., Laing, J., & Csikszentmihalyi, M. (2016). Positive Tourism (1st ed.). New York: Routledge, (Chapter 1). Finn, A. (2012). Customer delight: Distinct construct or zone of nonlinear response to customer satisfaction? Journal of Service Research, 15(1), 99–110. Fornell, C., & Larcker, D. F. (1981). Structural Equation Models with unobservable variables and measurement error: Algebra and Statistics. Journal of Marketing Research, 18(3), 382–388. Gentile, C., Spiller, N., & Noci, G. (2007). How to sustain the customer experience: An overview of experience components that co-create value with the customer. European Management Journal, 25(5), 395–410. Hair, J. F., Anderson, R. E., Tatham, R. L., & William, C. (1998). Black. Multivariate Data Analysis, 5, 87–135. Harrington, R. J., Hammond, R. K., Ottenbacher, M. C., Chathoth, P. K., & Marlowe, B. (2018). From goods-service logic to a memory-dominant logic: Business logic evolution and application in hospitality, International Journal of Hospitality Management, 76, 252–260. Hasan, S. A., Raheem, S., & Subhani, M. I. (2011). Measuring customer delight: A model for banking industry. European Journal of Social Sciences, 22(4), 510–518. Holbrook, M. B., & Hirschman, E. C. (1982). The experiential aspects of consumption: Consumer fantasies, feelings, and fun. Journal of Consumer Research, 9(2), 132–140. Hosany, S., Prayag, G., Deesilatham, S., Cauševic, S., & Odeh, K. (2015). Measuring tourists’ emotional experiences: Further validation of the Destination Emotion Scale. Journal of Travel Research, 54(4), 482 –495. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. Jollife, I. T., & Cadima, J. (2016). Principal component analysis: A review and recent developments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 374(2065), 1–16. Keiningham, T. L., Goddard, M. K. M., Vavra, T. G., & Iaci, A. J. (1999). Customer delight and the bottom line. Marketing Management, 8(3), 57–63. Kim, D., & Perdue, R. R. (2013). The effects of cognitive, affective, and sensory attributes on hotel choice. International Journal of Hospitality Management, 35, 246–257. Kim, J. H., & Ritchie, J. R. B. (2014). Cross-cultural validation of a Memorable Tourism Experience Scale (MTES). Journal of Travel Research, 53(3), 323–335. Kim, J. H., Ritchie, J. R., & McCormick, B. (2012). Development of a scale to measure memorable tourism experiences. Journal of Travel Research, 51(1), 12–25. Kim, J., Kim, Y., & Kim, D. (2017). Improving well-being through hedonic, eudaimonic, and social needs fulfillment in sport media consumption. Sport Management Review, 20, 309–321. Kim, M., Vogt, C. A., & Knutson, B. J. (2013). Relationships among customer satisfaction, delight, and loyalty in the hospitality industry. Journal of Hospitality & Tourism Research, 39(2), 170–197. Kumar, A., Olshavsky, R., & King, M. (2001). Exploring alternative antecedents of customer delight. Journal of Customer Satisfaction, Dissatisfaction and Complaining Behavior, 14, 14–26. Kwong, K. K., & Yau, O. H. M. (2002). The Conceptualization of customer delight: A research framework. Asia Pacific Management Review, 7(2), 255–266. Laros, F. J. M., & Steenkamp, J. B. E. M. (2005). Emotions in consumer behavior: A hierarchical approach. Journal of Business Research, 58, 1437–1445. Liu, M. W., & Keh, H. T. (2015). Consumer delight and outrage: scale development and validation. Journal of Service Theory and Practice, 25(6), 680–699. Liu, W., Sparks, B., & Coghlan, A. (2016). Measuring customer experience in situ: The link between appraisals, emotions and overall assessments. International Journal of Hospitality Management, 59, 42–49. Loureiro, S. M. C. (2010). Satisfying and delighting the rural tourists. Journal of Travel and Tourism Marketing, 27, 396–408. Loureiro, S. M. C., & Kastenholz, E. (2011). Corporate reputation, satisfaction, delight, and loyalty towards rural lodging units in Portugal. International Journal of Hospitality Management, 30(3), 575–583. Loureiro, S. M. C., Miranda, F. J., & Breazeale, M. (2013). Who needs delight? The greater impact of value, trust and satisfaction in utilitarian, frequent-use retail. Journal of Service Management, 25(1), 1757–5818. Ludwig, N. L., Barnes, D. C., & Gouthier, M. (2017). Observing delightful experiences of other customers: the double-edged sword of jealousy and joy. Journal of Service Theory and Practice, 27(1), 145-163. Ma, J., Gao, J., Scott, N., & Ding, P. (2013). Customer delight from theme park experiences. The antecedents of delight based on Cognitive Appraisal Theory. Annals of Tourism Research, 42, 359–381. Ma, J., Scott, N., Gao, J., & Ding, P. (2016). Delighted or satisfied? Positive emotional responses derived from theme park experiences. Journal of Travel & Tourism Marketing, 34(1), 1–19. McNeilly, K. M., & Feldman Barr, T. (2006). I love my accountants – they’re wonderful: understanding customer delight in the professional services arena. Journal of Services Marketing, 20(3), 52–159. Nordhorn, C., Scuttari, A., & Pechlaner, H. (2018). Travel motivation of Indonesian seniors in choosing destination overseas. International Journal of Culture, Tourism and Hospitality Research, 12(2), 185–197. Oliver, R. L., Trust, R. T., & Varki, S. (1996). Customer delight: Foundations, findings, and managerial irnsight. Journal of Retailing, 73(3), 311–336. Ou, Y. C., & Verhoef, P. C. (2017). The impact of positive and negative emotions on loyalty intentions and their interactions with customer equity drivers. Journal of Business Research, 80, 106–115. Parasuraman A, Zeithaml A, Valarie, L., & Berry, L. (1988). SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64, 12–40. Pine, J., & Gilmore, J. H. (1998). Welcome to the Experience Economy. Harvard Business Review, July–August, 97–105. Plutchik, R. (1980). A general psychoevolutionary theory of emotion. Theories of Emotion, 1, 3–31. Preko, A., Kwami A. S., & Feglo, M. (2014). Service delivery, customer satisfaction and customer delight in the real estate business. Evidence from Elite Kingdom Investment and Consulting Company Ghana. European Journal of Business and Management, 6(3), 71–83. Roberts-Lombard, M., & Petzer, D. J. (2018). Customer satisfaction/delight and behavioural intentions of cell phone network customers-an emerging market perspective. European Business Review, 30(4), 427–445. Shneider, B., & Bowen, D. (1999). Undestending customer delight and outrage. Sloan Management Review, 41(1), 35–45. Torres, E. N. (2014). Deconstructing service quality and customer satisfaction: Challenges and directions for future research. Journal of Hospitality Marketing & Management, 23(6), 652–677. Torres, E. N., Fu, X., & Lehto, X. (2014). Are there gender differences in what drives customer delight? Tourism Review, 69(4), 297–309. Torres, E. N., & Kline, S. (2006). From satisfaction to delight: a model for the hotel industry. International Journal of Contemporary Hospitality Management, 18(4), 290–301. Torres, E. N., & Ronzoni, G. (2018). The evolution of the customer delight construct: Prior research, current measurement, and directions for future research. International Journal of Contemporary Hospitality Management, 30(1), 57-75. Tung, V. W. S., & Ritchie, J. R. B. (2011). Exploring the essence of memorable tourism experiences. Annals of Tourism Research, 38(4), 1367–1386. Vanhamme, J. (2008). The surprise-delight relationship revisited in the management of experience. Recherche et Applications En Marketing, 23(3), 113–138. Watson, L., & Spence, M. T. (2007). Causes and consequences of emotions on consumer behaviour. European Journal of Marketing, 41(5/6), 487–511. Zhang, T., Lu, C., Torres, E., & Chen, P. J. (2018). Engaging customers in value co-creation or co-destruction online. Journal of Services Marketing, 32(1), 57-69. Zhao, Y., Yan, L., & Tat-Keh, H. (2018). The effects of employee behaviours on customer participation in the service encounter: The mediating role of customer emotions. European Journal of Marketing, 52(5/6), 1203–1222. Appendix I Research dimensions, variables, items and sources Factor Survey Variable/Item Origin code Cognitive dimension Tangibles CDT1 Ambiance/interior appearance of the place (Ekinci, Dawes, Massey, & Marketing, 2008) CDT2 Accessibility (Ahrholdt et al., 2017) CDT3 Flexibility and accommodation of schedules (Loureiro & Kastenholz, 2011) CDT4 Communication facilities (Ariffin & Yahaya, 2013) Personalisation CDP1 Offers according to individual consumer´s (Collier, Barnes, Abney, & Pelletier, 2018) /customisation preferences/needs CDP2 Unsolicited recommendations provided by the staff (Barnes et al., 2016) Organisational staff CDS1 Making me feel like a VIP consumer (Hasan et al., 2011) CDS2 Staff attitude (Liu & Keh, 2015) Justice CDJ1 Awards and discount (Loureiro, Miranda, & Breazeale, 2013; CDJ2 The price relative to the quality of the service Hasan et al., 2011; Kwong & Yau, 2002) Trust CDC1 I trust / rely on their knowledge to meet my needs CDC2 Refund Culture of delight CDQ1 The service is adapted to consumers of all kinds including people with special needs Feedback CDF1 Information availability CDF2 Behaviours in response to my complaints and suggestions Esteem/Self-esteem CDN1 Extremely concerned about meeting my specific (Vanhamme, 2008) needs Affective dimension Happiness ADH1 Enthusiastic (Laros & Steenkamp 2005) ADH2 Optimistic ADH3 Hopeful ADH4 Stimulated ADH5 Surprise Love ADL1 Tenderness Contentment ADC1 Joy ADC2 Peace Hedonic consumption ADHC1 The sensation of no time passing during this (Bernardo et al., 2012) experience ADHC2 Sharing the experience with other people Social value ADSV1 Knowledge of service policies (Chena & Lin, 2015) ADSV2 Other customer’s opinion Appraisal APPD1 The experience makes me: The authors Option A: Satisfied Option B: More than satisfied Option C: Delighted Delight ADD1 Excellence (Vanhamme, 2008) ADD2 Overprice The authors ADD3 Positive affect ADD4 Best offers available ADD5 Exceptional value Appendix II The exploratory power of the overall model. Standardised solution of the mediation model with control variables Fornell and Larcker (1981) developed an evaluating structural model with unobservable variables and measurement errors. The authors explain that if the value for the extracted variance is less than 0.50, the validity of the individual indicators as well as the construct should be questioned. However, the authors explain that to assess the exploratory power of the overall model researchers must consider both measurement and theory. Since the EQS tool does not provide sufficient values in the output to develop a canonical correlation analysis as posit Fornell and Larcker (1981), we tested the model by considering the control variables. The introduction of the control variables in the model through SEM is conditioned by the nature of the variables. Only the "age" variable is of scale, the other control variables (nationality, gender and sector) are "nominal". In SEM, it is only possible to introduce qualitative variables to the analysis, not nominal. Therefore, the sample was separated into two sub-samples, considering males and females to test the model. In addition, the age variable was introduced to test the power of the structural model. The Table 1 shows the coefficient of determination of regression (R2) for each tested model in order to show their variation. We only include the paths corresponding to the second-order factors and their R2. It is important to clarify that the coefficient of determination is the proportion of the total variance of the variable explained by the regression. The coefficient of determination reflects the goodness of the fit of the model to the variable that it tries to explain. However, the problem of the coefficient of determination is that it does not penalize the inclusion of non-significant explanatory variables. That means that it is very likely that the coefficient will increase if we include more explicative variables. In this case, it must also be taken into account that model 1 and 2 correspond to sub-samples within the whole sample. Table 1 R-Squared behavior R-Squared behaviour Path Base model Model 1 Model 2 Model 3 CP→AP 0.951 0.351 0.278 0.477 AP→ Delight 0.643 0.949 0.986 0.947 Sample n=400 n=182 n=218 n=400 Fit statistics x2 Satorra–Bentler (df = 123) = x2 Satorra–Bentler (df = x2 Satorra–Bentler (df = 153) = x2 Satorra–Bentler (df = 139) 231.4349 (p-value = 0.000); 153) = 319.201 (p-value = 137.2127 (p-value = 0.000); = 291.4865 (p-value = RMSEA = 0.057; 0.000); RMSEA = 0.048; RMSEA = 0.057; CFI = 0.913; 0.000); RMSEA = 0.054; CFI CFI = 0.949; BB-NFI = BB-NFI = 0.915; BB-NNFI = = 0.901; BB-NFI = 0.902; CFI = 0.919; BB-NFI = 0.921; BB- 0.951; BB-NNFI = 0.870.* 0.806.* All significant at p ˂ BB-NNFI = 0.806.* All NNFI = 0.819.* All significant at p All significant at p ˂ 0.05 0.05 significant at p ˂ 0.05 ˂ 0.05 In the Table: Base model (Causal model in the article); Model 1 (Causal model for male sub-sample); Model 2 (Causal model for female sub-sample); Model 3 (Causal with age group as control variable in relation to the delight factor). Appendix III Mediation model. Indirect and directs effects. Sobel test. In development, the mediation effect among the variables in the study was conducted via the direct and indirect effects analysis using the EQS software. In the research model, the effect of the cognitive attributes (CP) on the delight state (customer delight) is mediated through the affective factors (AP). The table below also shows the coefficient for the Sobel test for determining the significance of mediation and the probability density value. The reported p- values are drawn from the unit normal distribution, under the assumption of +/- 1.96, as critical values of the test ratio which contain the central 95% of the unit normal distribution. Path Total effect Significance Partial indirect Direct effect effect CP→ customer 0.697 (3.19) p ˃ 0.05 a*b= 0.677 0.678 (3.19) (c') delight (c) p < 0.01 CP→ AP (a) 0.707 (6.22) p < 0.01 AP→ customer 0.958 (4.39) p < 0.01 delight (b) Note: Standardized parameters (t-value); n = 400 Sobel test: Z= 3.61065282 p-value = 0.0003 According to the analysis, the mean indirect effect from the analysis is positive and significant, with a 95% confidence interval excluding zero. The direct effect c=0.69 is not significant (p ˃ 0.05). However, the other direct effects are all positive and significant (p < 0.01). The coefficient and significance for the Sobel test are also acceptable. Figures caption Figure 1 Cognitive-Affective-Delight Model (CAD) Figure 2 Coefficients in the mediation model: *p ˂ .05; ** p ˂ .01 *** p ˂ .001 Table 1 Frequency distribution according to the service selected by customers Frequency % 77 19.3 Hotel 109 27.3 Restaurant Spa 18 4.5 Water Park 9 2.3 Natural Park 91 22.8 Mall/Store 34 8.5 Other 62 15.5 Total 400 100 Table 2 Demographic information of the sample Variable Description % Age Between 18 and 29 years old 37.00 Between 30 and 49 years old 27.25 Between 50 and 59 years old 18.75 60 or more 17.00 Gender Male 45.50 Female 54.50 Nationality Europe 72.50 America 20.00 Asia 5.00 Oceania 1.70 Africa 0.80 Sample size: n=400 Table 3 Measurement model (validity of scales) Factor Item Mean ± SD Factor Communality Variance Loading explained Cognitive component Staff CDP2 3,99 (0.83) 0,688 0,550 64.83% CDS1 4,22 (0.82) 0,846 0,740 CDS2 4,16 (0.88) 0,760 0,600 CDC1 3,97 (0.81) 0,702 0,548 CDN1 4,05 (0.79) 0,706 0,551 Interaction CDJ1 2,87 (1.08) 0,795 0,127 with the 3,51 (0.98) 0,649 0,581 CDF2 service AvailabilityCDT3 3,73 (0.82) 0,796 0,684 CDT4 3,80 (0.88) 0,617 0,570 KMO =0.80. Bartlett’s test of sphericity= (x =963.456, df =65, p=0.000) 2 Affective component Happiness ADH1 3,89 (0.96) 0,833 0,723 73.62% ADH2 4,04 (0.86) 0,869 0,792 ADH3 3,82 (0.98) 0,842 0,729 ADH4 4,08 (0.89) 0,790 0,709 Hedonic ADHC1 4,00 (0.90) 0,606 0,577 consumption ADHC2 3,98 (1.07) 0,621 0,501 KMO =0.87. Bartlett’s test of sphericity= (x2=1085.447, df =21, p=0.000) Customer Delight ADD1 3,97 (0.80) 0,787 0,619 70.28% ADD3 4,02 (0.88) 0,864 0,746 ADD5 3,87 (0.93) 0,862 0,743 KMO = 0.687. Bartlett’s test of sphericity= (x2=347.721, df =3, p=0.000) Sample size: n=400 Table 4 Measurement model (reliability and validity) Construct CFA Standardized r2 t-values loadings Staff CDP2: Staff recommendations 0,636 0,341 10,89 AC= 0.813; RACIR= 0,927-0,931; RCBI= 0,952- CDS1: VIP treatment 0,869 0,645 10,90 0,773; CR=0,816; AVE= 0,48 CDS2: Staff attitude 0,785 0,405 10,34 CDC1: Trust 0,608 0,755 8,50 CDN1: Needs’ appreciation 0,497 0,616 7,78 Interaction with the service CDJ1: Awards/off 0,511 0,260 3,54 AC=0.906; RACIR= 0,887-0,899; RCBI = 0,644- CDF2: Complaints/suggestions 0,822 0,369 3,53 0,799; CR= 0,626; AVE= 0,47 Availability CDT3: Schedule 0,584 0,675 5,74 AC = 0.915; RACIR = 0,881-0,908; RCBIS = CDT4: Communication 0,809 0,561 5.74 0,723-0,848; CR = 0,659; AVE=0,50 Happiness ADH1: Enthusiastic 0,498 0,645 7,63 AC= 0.864; RACIR = 0,850-0,918; RCBIS = ADH2: Optimistic 0,716 0,561 7,63 0,771-0,852; CR= 0,725; AVE= 0,41 ADH3: Hopeful 0,513 0,645 10,27 ADH4: Stimulated 0,777 0,393 7,91 Hedonic consumption ADHC1: Time in the place 0,691 0,376 8,09 AC= 0.915; RACIR =0,881-0,908; RCBIS ADHC2: Share the experience 0,648 0,248 8,09 =0,723-0,848; CR=0,619; AVE=0,45 Customer Delight ADD1: Excellent experience 0,803 0,512 3,13 AC= 0.788; RACIR = 0,884-887; RCBIS = 0,681- ADD3: Positive affect 0.627 0,545 3,57 0,727; CR=0,735; AVE= 0,42 ADD5: Exceptional value 0.613 0,393 3,67 In the last column: Alpha Cronbach (AC); Range of Cronbach’s alpha if one item is removed (RACIR); Range of correlations between items and total corrected scale (RCBIS); Composite Reliability (CR); Average Variance Extracted (AVE); n=400. Table 5 Correlation matrix of latent factors 1 2 3 4 5 6 1. Delight 0,692 2. Staff 0,541 0,638 3. Interaction/service 0,314 0,373 0,645 4. Availability 0,447 0,530 0,307 0,684 5. Happiness 0,974 0,554 0,321 0,457 0,706 6. Hedonic consumption 0,811 0,462 0,268 0,381 0,491 0,670 In the table: Values of the correlations among latent factors Table 6 Standardised solution of the mediation model Path Coefficient* R-squared t-value CP→AP 0.692 0.951 4,40 AP→ Delight 0.975 0.643 7,57 CP→ Staff 0.802 0.216 4,99 CP→ Interaction/service 0.465 0.437 3,01 CP→ Availability 0.661 0.998 4,28 AP→ Happiness 0.832 0.692 6,42 AP→ Hedonic consumption 0.999 0.479 6,43 Fit statistics: x2 Satorra–Bentler (df = 123) = 231.4349 (p-value = 0.000); RMSEA = 0.057; CFI = 0.919; BB-NFI = 0.921; BB-NNFI = 0.819.* All significant at p ˂ 0.05 Table 7 Standardised solution of the mediation model with control variable Path Coefficient* R-Squared t-value CP→AP 0.642 0.413 6,14 AP→ Delight 0.905 0.943 7,59 CP→ Staff 0.785 0.616 7,82 CP→ Interaction/service 0.446 0.218 3,30 CP→ Availability 0.672 0.451 5,47 AP→ Happiness 0.902 0.989 5,71 AP→ Hedonic consumption 0.805 0.649 5,70 Relationships incorporated and equations Appraisal→ Delight 0.351 0.943 6,13 AP →Appraisal 0.419 0.989 5,61 Fit statistics: x2 Satorra–Bentler (df = 157) = 302.899 (p-value = 0.000); RMSEA = 0.059; CFI = 0.901; BB-NFI = 0.904; BB-NNFI = 0.767.* All significant at p ˂ 0.05; n.s nonsignificant