Adapting and Validating Scale of Customer Engagement in Online Travel Communities

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Adapting and Validating Scale of Customer Engagement in Online Travel Communities © 2020 The Author(s) Published version Mkumbo, Peter J.; Ukpabi, Dandison C.; Karjaluoto, Heikki Mkumbo, P. J., Ukpabi, D. C., & Karjaluoto, H. (2020). Adapting and Validating Scale of Customer Engagement in Online Travel Communities. European Journal of Tourism Research, 25, Article 2501. https://doi.org/10.54055/ejtr.v25i.416 2020 © 2020 The Author(s) This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0). To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/ RESEARCH PAPER 1 Adapting and Validating Scale of Customer Engagement in Online Travel Communities Peter J. Mkumbo 1 , Dandison C. Ukpabi 2* and Heikki Karjaluoto 3 1 Department of Parks, Recreation and Tourism Management, Clemson University, USA. E-mail: [email protected] 2 Marketing Department at University of Jyväskylä, Finland. E-mail: [email protected] 3 Marketing Department at University of Jyväskylä, Finland. E-mail: heikki[email protected] * Corresponding author Abstract Increasing attention towards customer engagement has caused its measurement to remain a highly debated issue among scholars. This study adapts and validates measurement scales for customer engagement in Online Travel Communities. It builds on previous studies on scale development for customer engagement. Data were collected from 450 members of Online Travel Communities in eight countries: Australia, Canada, Hong Kong (Chinese territory), New Zealand, Singapore, South Africa, the United Kingdom and the United States of America. The process of adapting and validating the scale involved exploratory factor analysis, testing for differential item functioning, examination of item response theory, confirmatory factor analysis, testing for invariance and criterion validity. The results found that three dimensions (affection, absorption, and interaction) suitably and adequately measure customer engagement in Online Travel Communities. Keywords: Customer engagement; Online Travel Communities; Psychometric scale development and validation Citation: Mkumbo, P., C., Ukpabi, D. and Karjaluoto, H. (2020). Adapting and Validating Scale of Customer Engagement in Online Travel Communities. European Journal of Tourism Research 25, 2501 Adapting and Validating Scale of Customer Engagement in Online Travel Communities 2 Introduction The literature on consumer engagement (CE) is growing. Since its adoption into the consumer research literature, several studies have applied it in different contexts, such as brand community (Algesheimer, Dholakia, & Herrmann, 2005), Facebook (Cheung, Lee, & Jin, 2011), automobiles (Sarkar & Sreejesh, 2014) and mobile phones (Dwivedi, 2015), among others. The growing interest is due to scholarly evidence that highlights that CE is a driver of customer trust, value, affective commitment, satisfaction and loyalty (Bowden, 2009; Brodie, Hollebeek, Juric, & Ilic, 2011; Vivek, Beatty, Dalela, & Morgan, 2014). Additionally, other critical indicators of brand performance, such as profit, sales growth and return on investment, have been linked to CE (Harrigan, Evers, Miles, and Daly, 2017; Hollebeek, 2011). While some of these studies are contextualised offline (Moreau, 2011; Sarkar and Sreejesh, 2014), social media has been the context for the majority (Cheung et al., 2011; Dessart, Veloutsou, and Morganthomas, 2016; Hollebeek, Glynn, and Brodie, 2014). This is mainly due to social media’s fostering of CE through relationship creation and sustenance (Sashi, 2012), interaction and value co-creation and customer experience management (Brodie, Ilic, Juric, & Hollebeek, 2013). Similarly, social media aggregates brand enthusiasts into communities where they connect, share experiences of hospitality and travel services and offer common programmes that influence and advance a brand (Dessart, Veloutsou and Morgan-Thomas, 2016). Thus, customer engagement defined as ‘. . . the level of an individual customer’s motivational, brand-related and context-dependent state of mind characterised by specific levels of cognitive, emotional and behavioural activity in direct brand interactions’ (Hollebeek, 2011, p. 790) has continued to attract both scholarly and practitioner attention in recent times. Although some attempts have been made to measure CE, there is still no scholarly consensus on the most appropriate dimensions for which populations (Table 1). Furthermore, within the hospitality and tourism domain, So, King, and Sparks (2014) and Harrigan et al. (2017) have differed on their dimensions and scale for CE. While So et al. (2014) proposed a 25-item CE scale with 5 dimensions, Harrigan et al. (2017) contended that 3 dimensions with 11 items are sufficient. To this end, Harrigan et al. (2017, p. 607) recommended that ‘future research should validate the CE scale and model using random samples in countries with varying cultures.’ Consequently, by building on these two studies, the aim of this study is to contribute to this debate by adapting and validating the CE scale that suits OTCs with samples drawn from eight countries: Australia, Canada, Hong Kong (Chinese territory), New Zealand, Singapore, South Africa, the United Kingdom (UK) and the United States of America (USA). Thus, our study’s key contributions to the hospitality and tourism literature is that we adapt, examine and validate the CE scale with tourism brands as proposed by So et al. (2014) and applied by Harrigan et al. (2017) in online travel communities (OTCs), which constitute critical engagement platforms between hospitality and tourism service providers and customers. To this end, consumers’ interest in travel sites, such as TripAdvisor, Booking.com and Expedia, among others, have continued to grow, and they remain helpful in travel decisions (Xiang & Gretzel, 2010). This paper proceeds as follows: in the next section, we present the literature review; section three discusses the methodology; section four presents the results; and section five offers the discussion, implications and limitations. Literature review CE has enjoyed increasing attention in practitioner and consumer behaviour literature (Harrigan et al., 2017) because of its enduring benefits to firms in relation to other customer-centric activities, such Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 3 as advertising and loyalty programmes (So et al., 2014). Since the extension of engagement into consumer behaviour research began, it has been applied in different research domains (Table 1). The table (Table 1) also contains how different studies have measured CE including the dimensionality, number of items used and sample diversity. This is in a bid to demonstrate that there is no uniform measurement of CE in extant research. However, different customer touchpoints across a wide spectrum of a firm’s activities, such as advertising, product or service offerings or even an event, engender engagement platforms (Vivek, Beatty and Morgan, 2012), with brand communities accentuating critical engagement platforms in which closer interaction between the firm and customers occurs. Interestingly, social media has enjoyed the most attention in the CE body of knowledge because it enhances real-life and two-way communication between the firm and customers (Dessart, Veloutsou and Morgan-Thomas, 2016). Extant research has documented antecedents as well as consequences of CE. We refer to Van Doorn et al. (2010) for explicit discussion of antecedents and consequences of customer engagement behaviour. towards both the firm and the customer. Hospitality and tourism services are experience-based; hence, they are personal and memorable and often regarded as high-involvement consumption contexts (Hur et al. (2017). Similarly, participation in OTCs builds relationships with fellow customers and with the brand (Casaló, Flavián and Guinalíu, 2010a). CE measurement scale and its dimensions Despite the growth and increasing attention towards CE, its measurement remains one of the most debated topics in the general service literature (Hollebeek et al., 2014; Vivek et al., 2014) and tourism research domain (Harrigan et al., 2017; So et al., 2014). With contextual differences, these studies have either measured CE unidimensionally or multidimensionally (Table 1). However, most subsequent studies dwelled extensively on the affective, cognitive and behavioural dimensions (Dessart et al., 2016). As an extension of affective commitment, Bowden, (2009) argued that the application of affective CE implies an emotional state, such as enjoyment, passion and enthusiasm, towards the focal firm and/or brand. Additionally, cognitive CE embodies a customer’s attention, absorption and sustained active interests in the firm and/or its brand (Brodie et al., 2011; Hollebeek, 2013). Finally, behavioural CE has been conceptualised as the vigour and energy that encompass a customer’s interaction with a brand (Brodie et al., 2011). Interestingly, CE measurement within hospitality and tourism research has included five dimensions: enthusiasm, identification, attention, absorption and interaction (Harrigan et al., 2017; So et al., 2014). Enthusiasm A customer’s positive perception of a service or product leads to greater interest and provides fertile ground for CE (van Doorn et al., 2010), with further interaction leading to enthusiasm. Enthusiasm reflects one of the ways in which customers emotionally connect to a brand (Hollebeek, 2013), and it implies a ‘strong level of excitement’ regarding the firm or focal brand (So et al., 2014, p. 308). Consumers who join brand communities and/or recommend a brand to others are driven by enthusiasm for that brand. Although the literature is unclear regarding whether enthusiasm differs from passion (Hollebeek, 2011, 2013), the two are often regarded as critical parts of CE’s emotional component (Hollebeek, 2011). In the context of CE measurement with hospitality and tourism brands, So et al. (2014) maintained that there are five dimensions, including enthusiasm, while Harrigan et al. (2017) contended that enthusiasm can be merged with absorption. Adapting and Validating Scale of Customer Engagement in Online Travel Communities 4 Table 1. Dimensions and measurement of CE in extant research Author(s) Context Dimensionality Number of items Sample diversity (Number of countries) Unidimensional Multidimensional Algesheimer et al. (2005) Brand community Community engagement - 4 1 Cheung et al. (2011) Facebook users - Vigour, absorption, dedication 18 1 Enginkaya and Esen, (2014) Online - Trust, dedication, reputation 14 1 Hollebeek et al. (2014) Social media users - Cognitive, refers to “a consumer's degree of positive brand-related affect, activation 10 1 So et al. (2014) Travellers - Identification, enthusiasm, attention, absorption, interaction 25 1 Sarkar and Sreejesh, (2014) Car owners Active engagement - 4 1 Vivek et al. (2014) Students - Conscious attention, enthused participation, social connection 10 1 Dwivedi, (2015) Mobile phone users - Vigour, dedication, absorption 17 1 Dessart, Veloutsou, and Morgan- Thomas, (2016) Facebook users - Affective (enthusiasm, enjoyment); cognitive (attention, absorption); behavioural (sharing, learning, endorsing) 22 3 Harrigan et al. (2017) Amazon Mechanical Turk - Identification, absorption, interaction 11 1 Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 5 Identification Identification originated in social identity theory (So et al., 2014), which holds that an individual’s selfconcept represents his or her personal identity (the ‘I’) and the group with which he/she associates (Tajfel, 1982). In the marketing context, a consumer’s sense of identification with a firm and/or brand is based on the ability of the firm and/or brand to satisfy his or her self-definitional goal (So et al., 2014), which can occur through either cognitive or affective processes (Stokburger-Sauer, Ratneshwar, & Sen, 2012). The cognitive process comprises the consumer’s perception of his or her own personality traits in relation to the brand while the affective process comprises the memorable experiences pertaining to the brand (Stokburger-Sauer et al., 2012). Interestingly, both So et al. (2014) and Harrigan et al. (2017) found support for identification as a valid factor for measurement of CE in hospitality and tourism brands. Attention From the organizational behaviour literature, attention partly constitutes the cognitive component of CE (Hollebeek, 2013; Vivek et al., 2012). It is the customer’s focus and conscious participation in issues relating to a brand and/or firm (So et al., 2014; Vivek et al., 2012). Customers who are engaged with a brand learn more and spend more time thinking about the brand (So et al., 2014). The use of attention as a factor of CE has been measured through various methods in the general marketing literature. For instance, Vivek et al.'s (2014) measurement of conscious attention is similar to Hollebeek's (2011) dimensions of immersion and activation, supporting attention as a valid factor of CE. However, in the measurement of CE in hospitality and tourism brands, So et al. (2014) considered attention a key dimension, whereas a study by Harrigan et al. (2017) found no support for its inclusion. Absorption Absorption is an extension of flow theory (So et al., 2014), which describes the state of total immersion that an individual encounters when engrossed in a given activity (Nakamura & Csikszentmihalyi, 2014). In the context of CE, absorption is a state wherein the customer experiences intrinsic enjoyment and is fully concentrated, happy with and engrossed in the brand (So et al., 2014; Vivek et al., 2012). There is no scholarly consensus on how absorption is measured as a dimension of CE. For instance, while Cheung et al. (2011) and Dwivedi, (2015) have acknowledged absorption as a distinct dimension of CE, Dessart et al. (2016) consider it a component of the cognitive dimension. Notably, the two studies that examined scale measurement and validation of CE in hospitality and tourism brands found support for its inclusion (Harrigan et al., 2017; So et al., 2014). Interaction As a dimension of CE, interaction is critical because it constitutes the window through which engagement takes place, and it is practical because it involves communicating one’s feelings about the brand (So et al., 2014). Brand communities constitute an important forum for brand enthusiasts to demonstrate their connection with the brand (Merrilees, 2016). However, time and distance pose critical challenges to this platform (Ukpabi & Karjaluoto, 2017). By contrast, social media has liberalised CE; through many platforms, engaged customers can post, write reviews, blog and share content (images, videos) of their favourite brands and experiences. While there is a scarcity of scholarly evidence that recognises interaction as a distinct factor in measuring CE, it has been recognised as a valid measurement of CE in hospitality and tourism brands (Harrigan et al., 2017; So et al., 2014). Adapting and Validating Scale of Customer Engagement in Online Travel Communities 6 Continuance participation The introduction of a piece of information system is usually accompanied with studies on users’ attitudes and intentions to using it (Boateng, Adam, Okoe & Anning-Dorson (2016; Israel, Tscheulin & Zerres, 2019). Within the information system literature, the technology acceptance model (TAM) has been prominently used to examine users’ attitude at the pre-adoption stage (Ooi & Tan, 2016). However, exposure to the system can influence can influence their attitude to continue or discontinue its usage. Thus, continuance usage intention, which is underpinned by users’ satisfaction, is critical to the survival of a piece of an information system (Bhattacherjee, 2001). As many online communities are facing critical challenges of retaining customers (Zhou et al. 2012), thus, understanding the interrelationships between customer engagement and continuance participation would be important for theory and practice. Extant studies have linked continuance intention to customer loyalty as they related to post adoption behaviours (Zhou et al. 2012) and the consequence of customer satisfaction (Anastassova, 2011; Cao et al., 2013; Moise, Gil-Saura & Ruiz-Molina, 2018). Methodology Survey design and data collection As mentioned previously, this study builds on So et al., (2014) customer engagement in tourism brands by adapting the scale in online travel communities. To test, validate, and adapt the scale that measures CE in the context of online travel communities, a questionnaire was designed and was distributed, using online panel company, to panels in eight countries, including Australia, Canada, Hong Kong (Chinese territory), New Zealand, South Africa, Singapore, the United Kingdom and the United States of America. The English-speaking countries were chosen due to limited resources to translate the questionnaire into other non-English languages. The designed questionnaire targeted only members of OTCs who reside in the eight countries. An individual was considered a member of an OTC if they had either one or multiple user accounts with an online travel website, such as TripAdvisor or LonelyPlanet, or had liked a social media page related to tourism and travel and are regularly following posts on those pages, websites or blogs. Measures Measurement scales for CE were adapted from the initial list of 28 items in So et al. (2014). The scale was composed of five dimensions: (1) identification, (2) enthusiasm, (3) attention, (4) absorption, and (5) interaction (Table 3). To check for content validity this list of items was emailed to 17 researchers who are professors or have deep expertise in tourism marketing, tourism management, and quantitative research methods especially, development of psychometric measures. In addition to these researchers, five tourism marketing managers were also consulted. These researchers and marketing managers were asked to assess the content validity of each dimension and the overall domain validity of the scale. They were provided with operational definition of customer engagement as behavioural manifestations that have a brand or firm focus, beyond purchase, resulting from motivational drivers (Verhoef, Reinartz, & Krafft, 2010); and online travel community as a group of individuals with shared travel, tourism or hospitality interests brought together by a travel, tourism or hospitality related brand using an online platform. Of the 17 researchers that were emailed, 14 responded with mixed responses. Some commented on the list of items that it is too long and, in many contexts, it is impractical to ask all those questions or there will be a serious concern with common method bias (Baumgartner & Weijters, 2012; Podsakoff, Mackenzie, & Podsakoff, 2012). Some researchers commented on the wording of the items so as to properly adapt the scale and make it suitable for customer engagement in online travel communities. A few other researchers expressed their doubts that five dimensions are too much, there are a lot of redundant items and that some need to be dropped out. All researchers seemed to agree that customer engagement in online travel communities Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 7 is a sub-scale/sub-component of the general customer engagement in tourism brands thus, in part, supporting the conceptualization and justification put forward in earlier sections of this study. All marketing managers responded and their comments, as among researchers above, were a little mixed. Critical comments were on the number of items being too long; the wording of items needs to be general but focusing on specific dimension and remain relevant on online travel communities, and item statements need to be shorter to reduce ambiguity among non-native English speakers. The authors in this study decided to retain all 28 items, made sure that the wording of items sufficiently and fully capture the nomological radii of the dimensions and also use of plain language to reduce ambiguity. Table 2 shows a list of all items and dimensions. Continuance participation (CPA) was used as a criterion variable for examining criterion validity and the scale for CPA was adopted from Hur, Terry, Karatepe, & Lee, (2017). The scale contains seven items (Table 3). Pre-test, a priori power analysis, and data collection A pre-test of the questionnaire was conducted in which a total of 50 respondents filled out the questionnaire. The distribution to panels was done under the help of an online panel company. Preliminary analyses (descriptive and correlation statistics) were conducted to gain a picture of the study, how variables are related so as to inform improvements if there were any in the final questionnaire. Power analysis was of most importance. A two-stage a priori power analysis was conducted; stage i) to determine a minimum sample size to detect model misfit (Kline, 2016) and stage ii) to estimate minimum sample size for sufficient power to detect an effect when testing for construct predictive validity and thus avoiding type II error (Farrokhyar, Reddy, Poolman, & Bhandari, 2013). Power analysis for stage (i) was conducted using semPower() package in R (Moshagen, 2018; Moshagen & Erdfelder, 2016) and the code argument contained the following minimum specifications for fit indices at multiple phases; power = 0.80, AGFI = 0.95, CFI = 0.95, GFI = 0.95, RMSEA = 0.05, alpha = 0.05, df = 994, and p = 47. Power analysis at this stage suggested a minimum sample size to detect model misfit is 110. Stage (ii) of power analysis was conducted using G*Power (Faul, Erdfelder, Buchner, & Lang, 2019), an online tool useful for a priori power analysis. In conducting power analysis using G*Power, a minimum correlation (r) between variables of .2 (based on pre-test results) and the highest correlation of .8 were used. The result suggested a sample size bigger than 53 will be needed. Since this project had other objectives beyond those mentioned in this study it was decided to target a minimum of 450 participants so as to allow for advanced and sophisticated analyses at subgroup levels. Minimum quotas for each country were abruptly suggested based on the country’s population as follows: Australia (50), Canada (50), Hong Kong (30), New Zealand (30), South Africa (30), Singapore (30), the UK (50) and the USA (80). Filters were set in the questionnaire to exclude untargeted participants and those who did not qualify for this study. Questionnaire validation was set ‘force response’ for questions related to measurement scales and ‘request response’ for general demographic questions. Algorithms were programmed to discard all incomplete responses. Since pre-test results showed that the median time to complete the questionnaire was 7.5 minutes; an extra filter was set to discard all questionnaires that were completed below two-thirds of the median time (i.e. below five minutes). In the final version of the questionnaire, items were measured on a seven-point Likert-type scale strongly disagree (1) to strongly agree (7). The questionnaire was distributed to online panels using Qualtrics algorithms. The analysis of data went through numerous rigorous stages namely; exploratory factor analysis (EFA), a Adapting and Validating Scale of Customer Engagement in Online Travel Communities 8 two-phase differential item functioning (DIF) testing, item response theory (IRT) testing, confirmatory factor analysis (CFA) for assessing measurement models, test for invariance (configural and factorial) and finally CFA for testing criterion validity of the scale in the structural model. The analysis procedure is presented in the following sections. Table 2. List of items and dimensions for CE adapted from So et al., (2014) Dimension Code Item statement Identification (CEID) ID1 When someone criticizes my online travel community, it feels like a personal insult. ID2 I am very interested in what others think about my online travel community. ID3 When I talk about my online travel community, I usually say WE rather than THEY. ID4 This online travel community’s successes are my successes. ID5 When someone praises this online travel community, it feels like a personal compliment. Enthusiasm (CEEN) EN1 I spend a lot of my discretionary time thinking about this online travel community. EN2 I am heavily into this online travel community. EN3 I am passionate about this online travel community. EN4 My days would not be the same without this online travel community. EN5 I am enthusiastic about this online travel community. EN6 I feel excited about this online travel community. Attention (CEAT) AT1 I would like to learn more about this online travel community. AT2 I pay a lot of attention to anything about this online travel community. AT3 Anything related to this online travel community grabs my attention. AT4 I concentrate a lot on this online travel community. AT5 I spend a lot of time thinking about this online travel community. AT6 I focus a great deal of attention on this online travel community. Absorption (CEAB) AB1 When I am interacting with this online travel community, I forget everything else around me. AB2 Time flies when I am interacting with this online travel community. AB3 I get carried away when I am interacting with this online travel community. AB4 It is difficult to detach myself when I am interacting with this online travel community. AB5 I become immersed when I am interacting with this online travel community. AB6 I feel happy when I am interacting intensely with this online travel community. Interaction IT1 In general, I like to get involved with this online travel Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 15 AB4 3.39 1.70 0.89 SRMR = 0.014, CFI = 0.99, RMSEA = 0.094 90%CI = [.000, .185] AB5 3.79 1.62 0.89 Interaction (CEIT) IT1 4.33 1.49 0.91 0.79 0.95 0.96 0.85 0.89 S-Bχ2 = 15.48, df = 5, p < 0.01, SRMR = 0.02, CFI = 0.98, RMSEA 0.079, 90%CI = [.044, .153] IT2 4.18 1.52 0.92 IT3 4.50 1.51 0.87 IT4 4.49 1.47 0.86 IT5 4.09 1.60 0.87 Higher order GOF: S-Bχ2 = 109.9, df = 62, p < 0.001, SRMR = 0.037, CFI = 0.98, RMSEA = 0.059, 90%CI = [.040, .076] n = 224 Continuance participation (CPA) (N = 449) Construct Code M SD Loading (λ) AVE α CR (ρ) 2nd Order (λ) SQRTAVE Fit Indices Continuance participation (CPA) CPA1 5.01 1.21 0.83 0.66 0.94 0.94 NA 0.81 S-Bχ2 = 17.6, df = 14, p = 0.09, CFI = 0.99, SRMR = 0.02, RMSEA = 0.04, 90%CI = [.000, .061] CPA2 4.97 1.28 0.87 CPA3 4.97 1.32 0.89 CPA4 4.91 1.31 0.87 CPA5 5.36 1.16 0.72 CPA6 4.83 1.45 0.74 CPA7 4.84 1.47 0.76 Table 9. Factor correlation matrices Calibration sample Validation sample S/N Dimension 1 2 3 Dimension 1 2 3 1 CEEN (0.85) CEEN (0.86) 2 CEAB 0.84 (0.90) CEAB 0.85 (0.88) 3 CEIT 0.82 0.76 (0.90) CEIT 0.81 0.78 (0.89) Diagonal values in brackets are the square roots of AVEs Testing for invariance In addition to DIF, testing for invariance assesses whether the scale measures the same construct across different samples drawn from the same population (Meredith, 1964; Putnick, Diane & Bornstein, Mark, 2016). While DIF assessed how individual items functions across demographical groups in the same population, invariance in this study assessed both items and dimensions across different samples from the same population. Testing for invariance is an essential step to make sure that differences, if there are any, between samples are not due to scale’s psychometric properties (Lee, 2018; Meredith, 1993). Two assessments were conducted to test for invariance between calibration and validation samples; configural and metric invariances (Cheung & Rensvold, 2002; Lee, 2018; Timmons, 2010). On the one hand, testing for configural invariance assesses if the general construct structure is consistent across different samples; calibration and validation in this study. On the other hand, metric Adapting and Validating Scale of Customer Engagement in Online Travel Communities 16 invariance examines if pattern coefficients (loadings) do not vary significantly across calibration and validation samples (Byrne, 2012; Kline, 2016). In both assessment procedures, EQS version 6.4 software was used. The results for configural invariance suggested that the proposed scale as examined across two samples, calibration (S-Bχ2 = 104.79, df = 62, p < 0.001, SRMR = 0.039, CFI = 0.98, RMSEA = 0.056) and validation (S-Bχ2 = 109.9, df = 62, p < 0.001, SRMR = .0037, CFI = 0.98, RMSEA = 0.059) is configural invariant (S-Bχ2 = 214.52, df = 124, Δχ2 = 0.17, SRMR = 0.038, CFI = 0.98, RMSEA = 0.057, 90%CI = [.044, .070]) meaning that the construct structure is consistent across different samples from the same population. In examining metric invariance, constraints were placed in each pair of loading path in calibration and validation samples. The results for metric invariance also suggested that item loadings are invariant (S-Bχ2 = 228.76, df = 137, Δχ2 = 14.24, Δdf = 13, p = 0.36, SRMR = .054, CFI = .98, RMSEA = .055, 90%CI = [.042, .067]) across the two samples (Byrne, 2006). In addition to the fit indices of overall metric invariance model, all applied loading constraints across two samples were found to be statistically not significant suggesting that the loadings do not vary significantly across samples of online travel communities. After having confirmed the proposed scale is invariant, the two samples; calibration and validation were re-combined into one and used as a single sample in the subsequent analyses and examinations. Testing for criterion validity Criterion validity could simply be defined as the extent to which there is a significant relationship between a given construct (which the scale is being developed for) and performance on another construct of particular relevance, commonly referred to as criterion variable (Cronbach & Meehl, 1955; DeVellis, 2017; Raykov & Marcoulides, 2011). Criterion validity exists in two main forms concurrent validity and predictive validity (Cronbach & Meehl, 1955; Dima, 2018). Concurrent criterion validity is the degree to which the construct (which a scale is developed for) has a stronger relationship with a theoretically-supported relevant criterion variable made at the time of data collection or shortly afterward (Cronbach & Meehl, 1955; DeVellis, 2017). Predictive criterion validity is the extent to which a construct (which the scale is developed to measure) predicts the response to another criterion variable which it is expected to relate with (Boateng et al., 2018; Chan, 2014) and therefore the scale should be able to predict expected behavioural intention or actual behaviour in the future. In this study, continuance participation (CPA) is used as a criterion construct. An increase in customer engagement with the online travel community should sustain participation of that particular customer on the OTC platform, thus CPA is a suitable criterion construct to examine predictive criterion validity. The scale for CPA is adopted from Hur, Terry, Karatepe, & Lee, (2017). The fact that data for both proposed construct and criterion construct were collected at the same time, makes it (CPA) also a relevant criterion construct for evaluating concurrent criterion validity. Prior to examining criterion validity, a Harman’s single factor test was conducted to examine for common method bias in which was found that a single factor accounts for 39% of the total variance. Even though a single factor accounted for less than 50% of the total variance (Harman, 1976) a common latent factor (CLF) was added into the structural model to control for common method bias by accounting for common variance (Craighead, Ketchen, Dunn, & Hult, 2011; Eichhorn, 2014; Podsakoff et al., 2012; Podsakoff & Organ, 1986). EQS version 6.4 software was used. The structural model was adequately specified (S-Bχ2 = 284.86, df = 130, p <.001SRMR = 0.060, CFI = 0.98, RMSEA = 0.052, 90%CI = [.043, .060]) and as expected customer engagement (CE) in online travel communities (OTC) significantly predicts continuance participation (CPA) (Figure 1, Table 8) a standard deviation increase in customer engagement in OTC increases continuance participation by 0.832 standard Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 17 deviation. In a plain language, a unit increase in CE on a 7-point Likert scale increases CPA by 0.832 units on a 7-point Likert scale. Customer engagement in OTC explains 69% of the variance in continuance participation (Figure 1). The results supported that the scale is criterion valid. Figure 1. Examining criterion validity Table 10. Examining criterion validity Criterion variable Predictive variable B Beta S.E. t-statistic p - value R2 Result CPA CE 0.239 0.832 0.064 3.734 0.00011 0.69 Supported Key: B = Unstandardized path coefficient; Beta = Standardized path coefficient; CE = Customer engagement; S.E. = Standard Error; R2 = Coefficient of determination; CPA = Continuance participation. Discussion and conclusion While So et al. (2014) suggested five factors for general customer engagement with tourism brands, this study suggests three dimensions to be used when measuring customer engagement in online travel communities. Findings of this study partly support those of Harrigan et al., (2017) (who used only Facebook fans) but the factor structures of the construct differ. The items in three extracted factor were further examined through DIF and IRT to make sure that each item is consistent in measuring what it is supposed to measure regardless of demographic differences among participants in samples drawn from the same populations (Baker & Kim, 2017; Hanson, 1998). Additionally, the scale invariance was examined for configural and metric and was found to be invariant. Finally, the scale was found to be criterion valid. A total of 15 items with poor psychometric properties were dropped out. The key objective of this study was to adapt a general customer engagement scale with tourism brands into online travel communities (OTCs). Contrary to So et al. (2014), this study suggests that customer engagement in online travel communities should be measured using three dimensions of affection, absorption and interaction Furthermore, these findings suggest that i) the construct domain of CE in OTC is smaller than the general customer engagement in tourism brands; ii) the nomological radius of CE construct in OTC is sufficiently covered by the dimensions of affection, absorption, and interaction, while attention and identification are redundant; they bring nothing unique to the higher-order construct of CE in OTC. Theoretically, this study contributes to the existing literature by adapting and validating the CE scale to OTCs. The findings of this study suggest that CE in OTC could be operationalized as a threedimensional construct. The three dimensions in their over-identification forms, thus providing an opportunity to assess model fit properly from the dimensional level. In covariance-based SEM, the proper assessment of model fitness, especially at the level of the measurement model, is vital before Adapting and Validating Scale of Customer Engagement in Online Travel Communities 18 one can look at how constructs relate structurally (Tabachnick & Fidell, 2012; Hair, Celsi, Money, Samouel, & Page, 2015; Kline, 2015). This is because covariance-based SEM models are founded on how the estimated model can best reproduce the covariance and variance matrices of the observed data; the only way of knowing that is by assessing both the measurement and the structural models’ fitness using established fit indices (absolute, incremental and parsimony) (MacCallum, 1986; Schermelleh- Engel, Moosbrugger, & Müller, 2003; Kline, 2015). While this scale could be broadly used in different customer engagement contexts, it is most appropriate in measuring customer engagement in online travel communities on a 7-point Likert or semantic scale. Managerially, this study presents two main implications: Managers can confidently measure CE using the three dimensions of affection, absorption, and interaction. Correctly measuring CE in OTCs is important because OTCs have become competitive marketing channels through which tourism and travel-related firms can attract, interact with, convert and retain customers. As such, management of the activities needed to engender CE is critical for continuous patronage. Managers of tourism and travel-related businesses should understand that enhancing individual dimensions of CE is important for driving overall CE in their business’ OTCs. In addition to educating members and promoting different offerings and prices, managers should consciously create opportunities to enhance CE among existing and potential customers. Members can be encouraged to interact with one another by sharing travel-related experiences, reviewing destinations/attractions and recommending destinations to visit, where possible. CE in OTCs could also be fostered when members of the OTC are considered first for offers, incentives, rebates, and discounts before they are made available to the general public via commercial media. Limitations and future research While the sample of this study was composed of individuals who are residents of eight countries from five continents, it included only English-speaking economies. This might limit the generalization of the findings within English speaking economies. The debate regarding CE in OTCs is yet to mature. This study presents with confident the findings especially the adaption and validation of CE in OTCs. Due to limited space, this publication is unable to examine nomological validity of the scale; instead it is presented in a separate publication. Other authors are encouraged to validate this study in other non-English speaking destinations and markets and to explore how the construct of customer engagement in OTCs fits in the wider nomological network (Kock, Josiassen, & Assaf, 2018) by including other variables like the credibility of OTCs platforms and other forms of behavioural intentions. Acknowledgement This study benefitted from the grant to the second author by the Jenny and Antti Wihuri Foundation, Finland (Grant Number – 00180408) References Algesheimer, R., Dholakia, U. M., & Herrmann, A. (2005). The Social Influence of Brand Community: Evidence from European Car Clubs. Journal of Marketing, 69(3), 19–34. Anastassova, L. (2011). Tourist loyalty and destination brand image perception: the case of Sunny beach resort, Bulgaria. European Journal of Tourism Research, 4(2), 191-204. Baker, F. B., & Kim, S.-H. (2017). The Basics of Item Response Theory Using R. Springer. https://doi.org/10.1007/978-3-319-54205-8 Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 19 Baumgartner, H., & Weijters, B. (2012). Commentary on “Common Method Bias in Marketing: Causes, Mechanisms, and Procedural Remedies.” Journal of Retailing, 88(4), 563–566. https://doi.org/10.1016/j.jretai.2012.10.003 Bhattacharya, C. B., & Sen, S. (2003). Consumer-Company Identification: A Framework for Understanding Consumers’ Relationships with Companies. Journal of Marketing, 67(2), 76–88. Boateng, G. O., Neilands, T. B., Frongillo, E. A., Melgar-Quiñonez, H. R., & Young, S. L. (2018). Best Practices for Developing and Validating Scales for Health, Social, and Behavioural Research: A Primer. Frontiers in Public Health, 6(June), 1–18. https://doi.org/10.3389/fpubh.2018.00149 Bowden, J. L.-H. (2009). The Process of Customer Engagement: A Conceptual Framework. Journal of Marketing Theory and Practice, 17(1), 63–74. https://doi.org/10.2753/MTP1069-6679170105 Brochado, A., Rita, P., & Gameiro, C. (2015). Exploring backpackers’ perceptions of the hostel service quality. International Journal of Contemporary Hospitality Management, 27(8), 1839–1855. https://doi.org/10.1108/IJCHM-03-2014-0145 Brodie, R. J., Hollebeek, L. D., Juric, B., & Ilic, A. (2011). Customer Engagement: Conceptual Domain, Fundamental Propositions, and Implications for Research. Journal of Service Research, 14(3), 252– 271. https://doi.org/10.1177/1094670511411703 Brodie, R. J., Ilic, A., Juric, B., & Hollebeek, L. (2013). Consumer engagement in a virtual brand community: An exploratory analysis. Journal of Business Research, 66(2013), 105–114. https://doi.org/10.1016/j.jbusres.2011.07.029 Byrne, Barbara M. (2012). Structural Equation Modeling with Mplus: Basic Concepts, Applications, and Programming. New York, USA: Routledge. Byrne, Byrne M. (2006). Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming (2nd Editio). Mahwah, NJ: Routledge. Casaló, L. V., Flavián, C., & Guinalíu, M. (2010). Antecedents and Consequences of Consumer Participation in On-Line Communities: The Case of the Travel Sector. International Journal of Electronic Commerce, 15(2), 137–167. https://doi.org/10.2753/JEC1086-4415150205 Chalmers, R. P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1–29. https://doi.org/10.18637/jss.v048.i06 Chan, E. K. H. (2014). Standards and Guidelines for Validation Practices: Development and Evaluation of Measurement Instruments. In B. D. Zumbo & E. K. H. Chan (Eds.), Validity and Validation in Social, Behavioural, and Health Sciences (pp. 9–24). Springer International Publishing. Cheung, C. M. K., Lee, M. K. O., & Jin, X.-L. (2011). Customer Engagement in an Online Social Platform: A Conceptual Model and Scale Development. In Thirty Second International Conference on Information Systems (pp. 1–8). Shanghai, China. Cheung, G. W., & Rensvold, R. B. (2002). Evaluating Goodness-of-Fit Indexes for Testing Measurement Invariance. Structural Equation Modeling, 9(2), 233–255. Conway, J. M., & Huffcutt, A. I. (2003). A Review and Evaluation of Exploratory Factor Analysis Practices in Organizational Research. Organizational Research Methods, 6(2), 147–168. https://doi.org/10.1177/1094428103251541 Craighead, C. W., Ketchen, D. J., Dunn, K. S., & Hult, G. T. M. (2011). Addressing common method variance: Guidelines for Survey Research on Information Technology, Operations, and Supply Chain Management. IEEE Transactions on Engineering Management, 58(3), 578–588. https://doi.org/10.1109/TEM.2011.2136437 Crocker, L., & Algina, J. (2006). Introduction to Classical and Modern Test Theory. Wadsworth Pub. Co. Cronbach, L. J., & Meehl, P. E. (1955). Construct Validity in Psychological Tests. Psychological Bulletin, 52(4), 281–302. Adapting and Validating Scale of Customer Engagement in Online Travel Communities 20 Dessart, L., Veloutsou, C., & Morgan-thomas, A. (2015). Consumer engagement in online brand communities: a social media perspective. Journal of Product & Brand Management, 24(1), 28–42. https://doi.org/10.1108/JPBM-06-2014-0635 Dessart, L., Veloutsou, C., & Morgan-Thomas, A. (2016). Capturing consumer engagement: duality, dimensionality and measurement. Journal of Marketing Management, 32(5–6), 399–426. https://doi.org/10.1080/0267257X.2015.1130738 DeVellis, R. F. (2017). Scale Development: Theory and Applications (4th Editio). Thousand Oaks, USA: SAGE Publications, Inc. Dima, A. L. (2018). Scale Validation in Applied Health Research: Tutorial for a 6-step R-based Psychometrics Protocol. Health Psychology and Behavioural Medicine, 6(1), 136–161. https://doi.org/10.1080/21642850.2018.1472602 Dwivedi, A. (2015). A higher-order model of consumer brand engagement and its impact on loyalty intentions. Journal of Retailing and Consumer Services, 24(2015), 100–109. https://doi.org/10.1016/j.jretconser.2015.02.007 Dziuban, C. D., & Shirkey, E. C. (1974). When is a Correlation Matrix Appropriate for Factor Analysis? Psychological Bulletin, 81(6), 358–361. Eichhorn, B. R. (2014). Common Method Variance Techniques (No. AA11-2014). Cleveland. Enginkaya, E., & Esen, E. (2014). Dimensions of Online Customer Engagement. Journal of Business, Economics and Finance, 3(1), 106–114. Farrokhyar, F., Reddy, D., Poolman, R. W., & Bhandari, M. (2013). Why Perform a priori sample size calculation? Canadian Journal of Surgical, 56(3), 207–213. https://doi.org/10.1503/cjs.018012 Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2019). G*Power 3.1.9.4 Manual. Germany. Faul, Franz, Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioural, and biomedical sciences. Behaviour Research Methods, 39(2), 175–191. Field, A., Miles, J., & Field, Z. (2012). Discovering Statistics Using R. SAGE Publications, Inc. Finch, W. H., Immekus, J. C., & French, B. F. (2016). Applied Psychometrics Using SPSS and AMOS. Charlotte, NC - USA: Information Age Publishing Inc. Floyd, F. J., & Widaman, K. F. (1995). Factor Analysis in the Development and Refinement of Clinical Assessment Instruments. Psychological Assessment, 7(3), 286–299. https://doi.org/10.1037/1040- 3590.7.3.286 Fornell, C., & Larcker, D. (1981). Evaluating Structural Equation Models with Unobservable Variables and Measurement Error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.2307/3151312 Gallaugher, J., & Ransbotham, S. (2010). Social Media and Customer Dialog Management at Starbucks. MIS Qarterly Executive, 9(4), 197–212. Gulliksen, H. (1950). Theory of Mental Tests. New York: Wiley. Hair, J. F., Celsi, M., Money, A., Samouel, P., & Page, M. (2015). The Essentials of Business Research Methods (3rd Editio). New York: Routledge. Hanson, B. A. (1998). Uniform DIF and DIF Defined by Differences in Item Response Functions. Journal of Educational and Behavioural Statistics, 23(3), 244–253. Retrieved from https://www.jstor.org/stable/1165247 Harman, H. H. (1976). Modern Factor Analysis (3rd Revise). Chicago, Illinois, USA: University of Chicago Press. Harrigan, P., Evers, U., Miles, M., & Daly, T. (2017). Customer engagement with tourism social media brands. Tourism Management, 59(2017), 597–609. https://doi.org/10.1016/j.tourman.2016.09.015 Hollebeek, L. (2011). Exploring customer brand engagement: definition and themes. Journal of Strategic Marketing, 19(7), 555–573. https://doi.org/10.1080/0965254X.2011.599493 Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 21 Hollebeek, L. D. (2011). Demystifying customer brand engagement: Exploring the loyalty nexus. Journal of Marketing Management, 27(7–8), 785–807. https://doi.org/10.1080/0267257X. 2010.500132 Hollebeek, L. D. (2013). The customer engagement/value interface: An exploratory investigation. Australasian Marketing Journal, 21(2013), 17–24. https://doi.org/10.1016/j.ausmj.2012.08.006 Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer Brand Engagement in Social Media: Conceptualization, Scale Development and Validation. Journal of Interactive Marketing, 28(2014), 149–165. https://doi.org/10.1016/j.intmar.2013.12.002 Hur, K., Terry, T. K., Karatepe, O. M., & Lee, G. (2017). An exploration of the factors influencing social media continuance usage and information sharing intentions among Korean travellers. Tourism Management, 63(2017), 170–178. https://doi.org/10.1016/j.tourman.2017.06.013 Janssens, W., Wijnen, K., Pelsmacker, P. De, & Kenhove, P. Van. (2008). Marketing Research with SPSS (1st Edition). Pearson Education. Kaiser, H. F. (1974). Little Jiffy, Mark IV. Educational and Psychological Measurement, 34, 111–117. Kline, R. B. (2015). Principles and Practice of Structural Equation Modeling (Methodology in the Social Sciences) (Fourth Edi). New York: The Guilford Press. Kline, R. B. (2016). Principles and Practice of Structural Equation Modeling (Fourth). New York, USA: The Guilford Press. Kock, F., Josiassen, A., & Assaf, A. G. (2018). Scale Development in Tourism Research: Advocating for a New Paradigm. Journal of Travel Research, 58(7), 1227–1229. https://doi.org/10.1177/ 0047287518800391 Ku, E. C. S. (2014). Distributed Fascinating Knowledge Over an Online Travel Community. International Journal of Tourism Research, 16, 33–43. https://doi.org/10.1002/jtr Kumar, V., Aksoy, L., Donkers, B., Venkatesan, R., Wiesel, T., & Tillmanns, S. (2010). Undervalued or Overvalued Customers: Capturing Total Customer Engagement Value. Journal of Service Research, 13(3), 297–310. https://doi.org/10.1177/1094670510375602 Lee, S. T. H. (2018). Testing for Measurement Invariance: Does Your Measure Mean the Same Thing for Different Participants? Retrieved from https://www.psychologicalscience.org/observer/ testing-for-measurement-invariance Linacre, J. M. (1994). Sample Size and Item Calibration or Person Measure Stability. Rasch Measurement Transactions, 7(4), 328. Retrieved from https://www.rasch.org/rmt/rmt74m.htm Linden, W.J. (Ed.). (2016). Handbook of Item Response Theory: Applications (Volume 3:). Boca Raton, FLUSA: A Chapman & Hall/CRC. Linden, Wim J. (Ed.). (2016a). Handbook of Item Response Theory: Models (Volume 1:). Boca Raton, FLUSA: Chapman & Hall/CRC. Linden, Wim J. (Ed.). (2016b). Handbook of Item Response Theory: Statistical Tools (Volume 2:). Boca Raton, FLUSA: A Chapman & Hall/CRC. Little, T. D., Lindenberger, U., & Nesselroade, J. R. (1999). On Selecting Indicators for Multivariate Measurement and Modeling with Latent Variables: When “Good” Indicators Are Bad and “Bad” Indicators Are Good. Psychological Methods, 4(2), 192–211. MacCallum, R. C. (1986). Specification Searches in Covariance Structure Modeling. Psychological Bulletin, 100(1), 107–120. https://doi.org/10.1037/0033-2909.100.1.107 Meredith, W. (1964). Notes on Factorial Invariance. Psychometrika, 29(2), 177–185. Meredith, W. (1993). Measurement Invariance, Factor Analysis and Factorial Invariance. Psychometrika, 58(4), 525–543. Merrilees, B. (2016). Interactive brand experience pathways to customer-brand engagement and value co-creation. Journal of Product & Brand Management, 25(5), 402–408. https://doi.org/10.1108/JPBM-04-2016-1151 Adapting and Validating Scale of Customer Engagement in Online Travel Communities 22 Moreau, C. P. (2011). Inviting the Amateurs into the Studio: Understanding How Consumer Engagement in Product Design Creates Value. Journal of Product Innovation Management, 28, 409–410. Moreno, H. (2017). Three Rules of Customer Engagement. Retrieved October 5, 2018, from https://www.forbes.com/sites/forbesinsights/2017/11/03/3-rules-of-customer- engagement/#392e48e85609 Moshagen, M. (2018). Package ‘semPower’ 1.0.0. Germany. https://doi.org/10.1080/10705511. 2014.950896>.License Moshagen, M., & Erdfelder, E. (2016). A New Strategy for Testing Structural Equation Models. Structural Equation Modeling: A Multidisciplinary Journal, 23(1), 54–60. https://doi.org/10.1080/10705511.2014.950896 Nakamura, J., & Csikszentmihalyi, M. (2014). The concept of flow. In M. Csikszentmihalyi (Ed.), Flow and the Foundations of Positive Psychology (pp. 239–263). Springer, Nertherlands. https://doi.org/10.1007/978-94-017-9088-8 Nering, M. L., & Ostini, R. (Eds.). (2010). Handbook of Polytomous Item Response Theory Models. Routledge. Novick, M. R. (1966). The Axioms and Principal Results of Classical Test Theory. Journal of Mathematical Psychology, 3(1), 1–18. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory (3rd ed.). McGraw-Hill, Inc. Podsakoff, P. M., Mackenzie, S. B., & Podsakoff, N. P. (2012). Sources of Method Bias in Social Science Research and Recommendations on How to Control It. Annual Review of Psychology, 63, 539– 569. https://doi.org/10.1146/annurev-psych-120710-100452 Podsakoff, P. M., & Organ, D. W. (1986). Self-Reports in Organizational Research: Problems and Prospects. Journal of Management, 12(4), 531–544. https://doi.org/10.1177/014920638601200408 Putnick, Diane, L., & Bornstein, Mark, H. (2016). Measurement Invariance Conventions and Reporting: The State of the Art and Future Directions for Psychological Research. Developmental Review, 41, 71–90. https://doi.org/10.1016/j.dr.2016.06.004.Measurement Raykov, T., & Marcoulides, G. A. (2011). Introduction to Psychometric Theory. Routledge, New York, NY 10017, USA. Rizopoulos, D. (2006). ltm: An R Package for Latent Variable Modelling and Item Response Theory Analyses. Journal of Statistical Software, 17(5), 1–25. https://doi.org/http://www.jstatsoft.org/v17/ i05/ Sarkar, A., & Sreejesh, S. (2014). Examination of the roles played by brand love and jealousy in shaping customer engagement. Journal of Product & Brand Management, 23(1), 24–32. https://doi.org/10.1108/JPBM-05-2013-0315 Sashi, C. M. (2012). Customer engagement, buyer-seller relationships, and social media. Management Decision, 50(2), 253–272. https://doi.org/10.1108/00251741211203551 Schermelleh-Engel, K., Moosbrugger, H., & Müller, H. (2003). Evaluating the Fit of Structural Equation Models: Tests of Significance and Descriptive Goodness-of-Fit Measures. Methods of Psychological Research Online, 8(2), 23–74. https://doi.org/10.1002/0470010940 Slocum, S. L., Gelin, M. N., & Zumbo, B. D. (2004). Statistical and Graphical Modeling to Investigate Differential Item Functioning for Rating Scale and Likert Item Formats. Vancouer: Edgeworth Laboratory, University of British Columbia. So, K. K. F., King, C., & Sparks, B. (2014). Customer Engagement with Tourism Brands: Scale Development and Validation. Journal of Hospitality and Tourism Research, 38(3), 304–329. https://doi.org/10.1177/1096348012451456 Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 23 Stokburger-Sauer, N., Ratneshwar, S., & Sen, S. (2012). Drivers of consumer – brand identification. International Journal of Research in Marketing, 29(2012), 406–418. https://doi.org/10.1016/j.ijresmar.2012.06.001 Tabachnick, B. G., & Fidell, L. S. (2012). Using Multivariate Statistics (6th Editio). United States of America: Pearson Education Inc. Tajfel, H. (1982). Social Psychology of Intergroup Relations. Annual Review of Psychology, 33, 1–39. Thakur, R. (2016). Understanding Customer Engagement and Loyalty: A Case of Mobile Devices for Shopping. Journal of Retailing and Consumer Services, 2016, 151–163. https://doi.org/10.1016/j.jretconser.2016.06.004 Thissen, D., Pommerich, M., Billeaud, K., & Williams, V. S. L. (1995). Item Response Theory for Scores on Tests Including Polytomous Items with Ordered Responses. Applied Psychological Measurement, 19, 39–49. Timmons, A. C. (2010). Establishing Factorial Invariance for Multiple-group Confirmatory Factor Analysis. Timmons, A. C. Manhatan, Kansas. Retrieved from http://crmda.dept.ku.edu/resources/kuantguides/22.Factorial_Invariance_Guide.pdf. Ukpabi, D., & Karjaluoto, H. (2017). Influence of Social Media on Corporate Heritage Tourism Brand. In R. Schegg & B. Stangl (Eds.), Information and Communication Technologies in Tourism 2017 (pp. 697–709). Cham: Springer. https://doi.org/https://doi.org/10.1007/978-3-319-51168-9_50 van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C. (2010). Customer Engagement Behaviour: Theoretical Foundations and Research Directions. Journal of Service Research, 13(3), 253–266. https://doi.org/10.1177/1094670510375599 van Tonder, E., & Petzer, D. J. (2018). The Interrelationships Between Relationship Marketing Constructs and Customer Engagement Dimensions. Service Industries Journal, 38(13–14), 948– 973. https://doi.org/10.1080/02642069.2018.1425398 Verhoef, P. C., Reinartz, W. J., & Krafft, M. (2010). Customer Engagement as a New Perspective in Customer Management. Journal of Service Research, 13(3), 247–252. https://doi.org/10.1177/1094670510375461 Vivek, S. D., Beatty, S. E., Dalela, V., & Morgan, R. M. (2014). A Generalized Multidimensional Scale for Measuring Customer Engagement. Journal of Marketing Theory and Practice, 22(4), 401–420. https://doi.org/10.2753/MTP1069-6679220404 Vivek, S. D., Beatty, S. E., & Morgan, R. M. (2012). Customer Engagement: Exploring Customer Relationships Beyond Purchase. Journal of Marketing Theory and Practice, 20(2), 122–146. https://doi.org/10.2753/MTP1069-6679200201 Woodbury, M. A. (1963). The Stochastic Model of Mental Testing and an Application. Psychometrika, 28, 391–394. Woods, C. M., & Grimm, K. J. (2011). Testing for Nonuniform Differential Item Functioning with Multiple Indicator Multiple Cause Models. Applied Psychological Measurement, 35(5), 339–361. https://doi.org/10.1177/0146621611405984 Xiang, Z., & Gretzel, U. (2010). Role of Social Media in online travel information search. Tourism Management, 31(2010), 179–188. https://doi.org/10.1016/j.tourman.2009.02.016 Yong, A. G., & Pearce, S. (2013). A Beginner’s Guide to Factor Analysis: Focusing on Exploratory Factor Analysis. Tutorials in Quantitative Methods for Psychology, 9(2), 79–94. https://doi.org/10.20982/tqmp.09.2.p079 Received: 22/07/2019 Accepted: 20/10/2019 Coordinating editor: Stanislav Ivanov Adapting and Validating Scale of Customer Engagement in Online Travel Communities 24 Appendix 1: Test for differential item functioning (DIF) Item Grouping variable Significance (p-value) Uniform DIF Non uniform DIF EN1 Gender 0.095 0.236 Country of residence 0.645 0.229 ID1 Gender 0.389 0.290 Country of residence 0.460 0.061 ID4 Gender 0.002* 0.031 Country of residence 0.061 0.501 ID5 Gender 0.004* 0.001* Country of residence 0.037 0.781 EN4 Gender 0.552 0.598 Country of residence 0.604 0.981 AT5 Gender 0.074 0.981 Country of residence 0.005* 0.103 AT6 Gender 0.003* 0.080 Country of residence 0.019 0.081 ID3 Gender 0.049 0.103 Country of residence 0.004* 0.027 EN2 Gender 0.895 0.941 Country of residence 0.921 0.016 EN3 Gender 0.002* 0.045 Country of residence 0.613 0.301 AT4 Gender 0.006* 0.001* Country of residence 0.041 0.482 IT4 Gender 0.763 0.282 Country of residence 0.615 0.766 IT3 Gender 0.106 0.944 Country of residence 0.797 0.101 IT2 Gender 0.049 0.823 Country of residence 0.391 0.823 IT1 Gender 0.243 0.985 Country of residence 0.568 0.329 IT5 Gender 0.068 0.836 Country of residence 0.811 0.328 AB4 Gender 0.091 0.799 Country of residence 0.430 0.846 AB5 Gender 0.795 0.829 Country of residence 0.462 0.473 AB3 Gender 0.072 0.592 Country of residence 0.039 0.482 AB2 Gender 0.051 0.377 Country of residence 0.536 0.921 AB1 Gender 0.006* 0.218 Country of residence 0.341 0.801 Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 31 Appendix 6: Examining IRT - Item endorsement on a 7-point Likert type scale continue Adapting and Validating Scale of Customer Engagement in Online Travel Communities 32 Appendix 6: Examining IRT - Item endorsement on a 7-point Likert type scale continue Mkumbo et al. (2020) / European Journal of Tourism Research 25, 2501 33 Appendix 6: Examining IRT - Item endorsement on a 7-point Likert type scale continue