A Multilevel Approach to Scale Development in Sport Tourist Nostalgia

Abstract

Nostalgia has been identified as an essential factor to understand sport tourists’ behavioral intentions. However, a measurement model to examine nostalgia has not been developed in the field of sport tourism. The purpose of this study was to develop a valid and reliable Nostalgia Scale for Sport Tourism (NSST) to measure sport tourists’ nostalgia. A multilevel analysis was used in order to avoid biases caused by common characteristics within a travel group. The scale conceptualized sport nostalgia as a five-dimensional construct reflecting sport tourists’ nostalgia of sport team, environment, socialization, personal identity, and group identity and showed adequate psychometric properties in assessing sport-specific nostalgia. The NSST scale developed here can be a useful tool for future empirical studies aiming to better understand sport spectator nostalgia and identify the role of nostalgia in sport tourism.

Keywords

nostalgia sport tourist scale development multilevel analysis

Introduction

Strong emotional memories from the past evoke nostalgic longing for the good old days. According to Chase and Shaw (1989), people living in today’s society reminisce about the past—desiring to return to a simpler, less corrupt time—as an escape from their busy lives. In sports, sport tourists maintain strong emotional attachments with their favorite teams based on previous positive experiences. Nostalgic recollections of unforgettable moments in sports become a strong motive for sport tourists to attend the home and, even distant, away games (Fairley 2003).

Spectator sport itself has unique characteristics such as sport-specific atmosphere, group rituals, norms, irrational passion of fans, vicarious identification, competitive balance, and limited availability (Edwards 1973; Fairley 2003; Hinch and Higham 2001; Kelly 1982; Uhrich and Benkenstein 2010). According to Cho, Ramshaw, and Norman (2014), these unique characteristics of sport affect the generation of nostalgia. They further specified the diverse experiences in sport settings that can evoke and are connected to nostalgia, such as identifying with a subculture through a group experience, socializing with others, cheering for a team or player, and childhood memories (e.g., learning rules, participating in sport). In addition, they asserted that nostalgia plays an essential role in understanding the behavior of consumers and tourists.

Whereas having regular visitors from the area is a unique aspect of sports (Wann and Branscombe 1993), whether it is a large-scale event or a small-scale event, sporting events attract various home fans and fans coming from out of town (Gibson, Willming, and Holdnak 2003; Kaplanidou and Vogt 2007). Nonetheless, sport tourists have to overcome higher levels of spatial and temporal constraints compared to the local fans (Gibson 1998). Cho, Ramshaw, and Norman (2014) highlighted nostalgia as a motivator bringing such sport tourists to the events and emphasized the need for development of a nostalgia scale specific to the context of sport tourism. That is, a scale that takes into account the uniqueness of sports to better understand the dynamics of sport-specific nostalgia. It is the aim of this study to develop a nostalgia scale germane to the needs of the sports travel and tourism industry.

A large body of research in consumer behavior identified the significant role and effects of nostalgia (Havlena and Holak 1991; Holbrook 1993; Schindler and Holbrook 2003). In particular, Havlena and Holak (1996) suggested categories of nostalgia, and Holak, Havlena, and Matveev (2006) further developed a scale of nostalgia from the perspective of the consumer. This scale, however, focuses on general consumer behavior (e.g., consuming tangible commodity) and as such has limited applicability to the spectator sports setting, which is an experiential consumption. For instance, the items ask about movies, daydreaming, toys, and society in general while these seem not to be the points of attachment to sports. Moreover, their scale is not conceptually consistent with the extant literature nor is it methodologically sound. Foremost, the researchers’ operationalization and the dimensionality of measures were ambiguous. The initial questionnaire contained 31 items, including 21 items measuring personal nostalgia, 2 measuring interpersonal nostalgia, 4 measuring cultural nostalgia, and 4 measuring virtual nostalgia. However, Holak, Havlena, and Matveev did not take a deductive approach in naming each factor, and the items of each factor of nostalgia were not loaded properly onto the reflective factors.

Secondly, Holak, Havlena, and Matveev’s (2006) nostalgia scale exhibited low reliability values, and they did not conduct any validity tests. This is against the psychometric theories that a measurement scale should have adequate psychometric properties, evidenced by reliability and validity tests (Kline 2011). In other words, the previous scale does not hold up under statistical scrutiny. Furthermore, Holak, Havlena, and Matveev simultaneously measured direct and indirect experiences of nostalgia in a unidimensional manner. Based on the conceptualization, these two experiences should be separately examined to identify the types of experience (Cho, Ramshaw, and Norman 2014). The reason for this separate consideration is that a direct experience of nostalgia is derived from an individual’s personal experience (Davis 1979), whereas an indirect experience of nostalgia is a nostalgic feeling due to word of mouth and culture, which is not through actual (direct) experience (Havlena and Holak 1991; Stern 1992). That is, unlike for the direct experience of nostalgia, respondents’ ages, memory status, and the time of the event should be considered in measuring the indirect experience of nostalgia.

Previous research in sport tourism has emphasized the imperative role of nostalgia (e.g., Fairley 2003; Fairley and Gammon 2005; Ramshaw 2005). Although tourists coming from out of town are inevitable consumers of sporting events bringing in economic impact, there has not been a rigorous scale developed to examine sport nostalgia as a multidimensional construct. Recently, Cho, Ramshaw, and Norman’s (2014) conceptual research suggested a conceptual model of nostalgia indicating the relationship among memory, types of experience, and nostalgia and developed a classification of nostalgia in the context of sport tourism based on the unique features of sports. Further, they called for future research to develop a measurement model to assess the multidimensional nature of sport-specific nostalgia. However, to date, such a precise scale of nostalgia has not yet been developed in the field of sport tourism. Furthermore, in the context of tourism, there has not been a scale assessment considering the biases caused by the common characteristics within a travel group (e.g., fans coming from out of town).

To address this need, this study developed and assessed a Nostalgia Scale for Sport Tourism (NSST) focusing on direct experience of sport tourism. Specifically, we conducted a multilevel analysis, which is an effective method to analyze hierarchically structured data. This was to avoid biased results caused by common characteristics within a group (Raudenbush and Bryk 2002) such as those found in the group-based nature of sport tourism (Fairley 2003). The NSST developed in this study contributes to an increased understanding of sport nostalgic behavior, expanding the domain of nostalgia research into experiential aspects of sport tourism. Specifically, we explicate the multiple dimensions of sport nostalgia and provide a rigorously assessed scale for use in the sport travel and tourism industry. By elucidating nostalgia as a strong motive for sport fans, we argue that managers in the travel and tourism industry need to better understand and utilize a rigorous nostalgia scale.

Literature Review

The Concept of Nostalgia

Nostalgia is different from reminiscing (Castelnuovo-Tedesco 1980), sentimentality (Wilson 2005), and autobiographical memory (Sedikides, Wildschut, and Baden 2004), all of which are considered cognitive mental processes (Brown and Schopflocher 1998; Skowronski, Walker, and Betz 2003), whereas nostalgia is considered an emotional mental process (Castelnuovo-Tedesco 1980; Cavanaugh 1989). Furthermore, Sedikides, Wildschut, and Baden (2004) explained that reminiscence and autobiographical memory occurred when the individual remembers certain events, ones that “do not have to be, and typically are not, important or affect-laden” (p. 205). However, Havlena and Holak (1991) and Holak and Havlena (1998) pointed out that nostalgia is caused by chance by external stimuli such as friends, family members, farewell parties, music, scent, mood, and objects. Further differentiating nostalgia, Castelnuovo-Tedesco (1980, cited by Sedikides, Wildschut, and Baden 2004) explained that reminiscence and autobiographical memory are “cold” processing while conversely, nostalgia is considered “hot” processing. More specifically, Davis (1979) defined the characteristic of nostalgia:

The nostalgic feeling is infused with imputations of past beauty, pleasure, joy, satisfaction, goodness, happiness, love, and the like, in sum, any or several of the positive effects of being. Nostalgic feeling is almost never infused with those sentiments we commonly think of as negative—for example, unhappiness, frustration, despair, hate, shame, abuse. (p. 14)

Kaplan (1987) held a view similar to Davis, defining nostalgia as “a universal affect that results in a heightened mental state, an enhancing, uplifting mood related to particular memories of past” (p. 465), a definition also similar to Castelnuovo-Tedesco (1980, p. 122), who considered nostalgia to be an emotional aspect, which “is sweet because the original object or event gave pleasure. . . . It is bitter not only because it cannot be made to come back but also because, even in its original setting, it contained conflict and disappointment.” Reminiscing and remembering are broader concepts than nostalgia, the primary difference being that reminiscence and nostalgia are a desire to return to the past, involving positive emotions, while reminiscence does not include either. According to Batcho (2007), “one can remember without being nostalgic, but one cannot be nostalgic without remembering” (p. 362). Distinguishing nostalgia from other ways of remembering is important to researchers who are interested in nostalgia. Though the concepts that are related to remembering are similar, their distinguishing features are important.

Nostalgia is based on positive memories that have already been experienced directly or indirectly by individuals (Havlena and Holak 1991; Stern 1992). People can experience nostalgic feelings even if they have negative memories of the past if their positive memories outweigh their negative emotions. If an individual experiences nostalgia, it is believed that he or she has more positive memories than negative ones or has overcome their negative memories. As this analysis suggests, the concept of nostalgia primarily focuses on positive emotions triggered by positive memories of the past.

Classification of Nostalgia in the Context of Sport Tourism

Nostalgia is more generally considered an important motive in travel and tourism (see Sellick 2004; Phau et al. 2016). In sport tourism, nostalgia remains one of the least studied topics in the field (Gordon 2013), although this may be in part because of a conceptual split between nostalgia sport tourism and heritage sport tourism (see Ramshaw and Gammon 2005). More recently, Cho, Ramshaw, and Norman (2014) developed a classification of nostalgia in the context of sport tourism derived from past literature related to nostalgia (Aden 1995; Wilson 2005), nostalgia sport tourism (Gibson 1998; Fairley 2003; Fairley and Gammon 2005), identity theory (Stets and Burke 2000), and social identity theory (Tajfel 1981). Their classification system involves two dimensions: (a) the purpose of nostalgia (experience-based nostalgia and identity-based nostalgia) and (b) the structure of nostalgia (object-based nostalgia and interpersonal relationship–based nostalgia). Their structure of nostalgia was developed based on the research conducted by Fairley and Gammon (2005), who contended that nostalgia is generated by both objects (people, places, or things) and social experiences. Thus, they suggested two categories for the structure of nostalgia: (a) object-based nostalgia and (b) interpersonal relationship–based nostalgia. Second, the purpose of nostalgia could be explained by one’s willingness to pursue and place a value based on past experiences. Aden (1995) and Wilson (2005) both found that people have nostalgic feelings because of their own past experiences and the continuity of their identity. In other words, a person can value the pursuit of the nostalgic experience by itself and as a means of verifying his/her identity. Based on this, Cho, Ramshaw, and Norman classified the purpose of nostalgia as (a) experience-based nostalgia and (b) identity-based nostalgia. Based on these two dimensions, they suggested a four-way classification of nostalgia: (a) nostalgia as experience, (b) nostalgia as socialization, (c) nostalgia as personal identity, and (d) nostalgia as group identity as seen in Table 1.

Table 1.

Classification of Nostalgia in the Context of Sport Tourism.

		Purpose of Nostalgia
		Experience-Based Nostalgia	Identity-Based Nostalgia
Structure of nostalgia	Object-based nostalgia	Nostalgia as experience	Nostalgia as personal identity
Structure of nostalgia	Interpersonal relationship–based nostalgia	Nostalgia as socialization	Nostalgia as group identity

Nostalgia as Experience

The classification of nostalgia as experience focuses on the relationship between one’s nostalgic feelings caused by past experiences and objects. According to Fairley (2003), a person feels nostalgia through objects, including places, people, and experiences. In the field of sport tourism, objects can be considered as one’s favorite sport players, teams, stadia, among others. That is, there can be various points of attachment in spectator sports. Robinson and Trail (2005) defined this as “being attached to a specific team, or perhaps even rather than being attached to a specific team, an individual might be attached to, for example, the coach or a specific player, among other things related to the experience” (p. 61). Viewing from a point of attachment perspective, nostalgia itself is an important point of attachment. Testing the many attachment points, Funk and James (2006) evidenced nostalgia as a significant factor affecting fans’ attachment. The positive past memories related to sport objects evoke nostalgic recollection regarding sport objects, such as sport players, teams, coaches, sport venues, food, music, and atmosphere (Cho, Ramshaw, and Norman 2014). Kulczycki and Hyatt (2005) found that even old sport facilities evoke nostalgia as one’s positive memories are related to the facilities. Nostalgia as an experience focuses on feelings as evoked by positive memories associated with diverse sport objects. Further, it is implied that nostalgia can stimulate the attitude and behavioral intentions for a tourist or a traveler living away from the sporting scene to be willing to reconnect with the venue.

Nostalgia as Socialization

The second category considers nostalgia as socialization, important because as Fairley (2003) and Fairley and Gammon (2005) pointed out, social interaction plays a significant role in nostalgia. This research examined group travelers who have annually followed one team in the Australian Football League (AFL) annually, emphasizing how the small group experience is important in understanding nostalgic behavior:

Therefore, it is reasonable to expect that sport nostalgia can be derived from group (or social) experiences which themselves become the basis for tourism. Thus, the focus of some sport tourism may be on the travel group itself and therefore on reliving a sport-based group (social) experience, rather than on visiting a particular site or destination. (p. 285)

Nostalgia as socialization emphasizes the relationship between nostalgic feelings and social experience. According to Cho, Ramshaw, and Norman (2014), this aspect of nostalgia is evoked by past social interaction experiences, such as sharing useful information and current news, building friendships, gaining diverse benefits (e.g., emotional, health, financial, and educational) by meeting group members, making new friends, and socializing with others. Moreover, an individual’s positive childhood memory of socializing with family and friends evokes nostalgic recollections, and in the process, sport rules learned during childhood act as a vehicle to promote the feelings of nostalgia.

Nostalgia as Fan (Personal) Identity

The third aspect, nostalgia as personal identity, is based on identity theory, with Hogg, Terry, and White (1995) indicating that its “central characteristics…are that 1) it represents a social psychological model of self in that social factors are seen to define self; 2) the social nature of self is conceived as derived from the role positions that people occupy in the social world; 3) in an enduring sense, these role identities are proposed to vary in regard to their salience; and 4) although identity theorists acknowledge that reciprocal links exist between self and society, they have been most interested in individualistic outcomes of identity-related process” (p. 259). As such, this concept of nostalgia is closely related to identity (Davis 1979), further supported by Barclay and DeCooke (1988), who pointed out that a person can realize himself or herself through nostalgic memories. In the sport tourism field, an individual’s behavior and emotions are influenced by the level of identification with one’s favorite players or teams (Schurr et al. 1988; Wann and Branscombe 1993). In addition, sport consumers use specific sport players or teams to build their identity (Smith and Stewart 2007), and in the process, can recognize their identity, which, in turn, causes nostalgia (Cho, Ramshaw, and Norman 2014). As Cho, Ramshaw, and Norman maintain, one has nostalgic feelings because of memories related to a sense of accomplishment, a desire to confirm one’s identity as a fan, self-respect, value, and pride of being a fan. In other words, an individual’s participation in sporting events effects the establishment of his or her identity, and this established identity can evoke nostalgic sentiments. Each person desires a role as a fan in sport settings or sport stadia, identifying his or her “true self” by taking part in sporting events. These experiences strengthen one’s personal identity and facilitate nostalgic feelings.

Nostalgia as Group Identity

The last aspect is nostalgia as group identity. Fairley and Gammon (2005) explored the relationships among nostalgia, personal identity, and group identity, concluding that the “memories that an individual holds include both self and collective memories that reflect an individual’s identification with, and belongingness to, a particular social group” (p. 183). Group identity is well explained by social identity theory, with Hogg, Terry, and White (1995) suggesting that this theory “is intended to be a social psychological theory of intergroup relations, group processes and the social self” (p. 259). Social identity is considered as one’s emotional attachment to a certain group membership with a sense of belonging (Tajfel 1981). Further, Jenkins (1996) indicated that social identity “refers to the ways in which individuals and collectives are distinguished in their social relations with other individuals or collectives” (p. 4). That is, a group has its own culture, rituals, and norms, wanting to be seen as unique from other groups. The experience of being acknowledged as a group member, thus, increases group identity (Cho, Ramshaw, and Norman 2014).

Positive memories and group experiences associated with group identity play a vital role in evoking nostalgia. In the field of sport tourism Fairley and Gammon (2005) noted that individual identity is affected by group identity, saying that “as both sport and tourism represent salient personal and collective identities for many, it is not surprising that memories of sport and tourism form the basis of nostalgic recollections” (p. 184). In other words, positive memories associated with the group behavior of sport spectators evoke nostalgia. Based on this classification of nostalgia developed by Cho, Ramshaw, and Norman (2014), this research employed these four dimensions to develop the NSST, which to enhance the understanding of the specific nostalgic behavior of attendees at sporting events.

Methods

The Development of the Scale

This study referenced Menor and Roth’s (2007) scale development procedure for developing its valid and reliable nostalgia scale for sport tourism (NSST). Menor and Roth suggested a two-stage approach to develop measurement items and scales, specifically involving seven steps: (1) specify the theoretical domain and operational definitions of constructs, (2) generate items, (3) purify and pretest items, (4) develop the questionnaire, (5) collect the survey data, (6) conduct a confirmatory analysis, and (7) refine the items and the scale. The first stage includes the first three steps and the second stage consists of the remaining four. By emphasizing the item assessments in the first stage, Menor and Roth’s procedure can result in improving overall measurement quality.

The first and second steps of this process are accomplished by semistructured interviews, literature review, Q-sort, and expert review to enhance clarity (Babbie 2010). A mixed purposeful sample was selected for an intensive interview. A homogeneous and typical group of 20 undergraduate student sport fans in the Department of Parks, Recreation and Tourism Management participated in the semistructured interviews. We note that this purposeful sampling does not ensure randomness. They were asked to complete an open-ended questionnaire to establish the list of nostalgia factors (e.g., What is your most memorable sporting event? [Reasons]. What made you feel nostalgia at the most memorable sporting event? [Reasons]). Then we analyzed the semistructured interviews identifying 75 keywords. The keywords were sorted to match the previously known four domains of nostalgia. The findings from this process supported Cho, Ramshaw, and Norman’s (2014) classification (Table 1). Results obtained from the literature and the information from the semistructured interviews were integrated and a pool of items was developed sampling each of four domains: experience, socialization, fan (personal) identity, and group identity. Initially, 69 items were developed from the previous nostalgia scales (Batcho 1995; Holbrook 1994; Pascal, Sprott, and Muehling 2002; Routledge et al. 2008; Rodrigues 2012), prior qualitative research (Fairley 2003; Fairley and Gammon 2005), and different types of nostalgia experiences (Stern 1992).

After generating the item pool, Q-sort and expert reviews were conducted to provide evidence of face validity and content validity. According to Zait and Bertea (2011), “the Q-sort procedure aims to separate items in a multi-dimensional construct according to their specific domain” (p. 218), meaning that this procedure helps the researchers identify inappropriate and ambiguous items. For this research, 16 graduate students and three professors in the field of parks, recreation, and tourism conducted the Q-sort. Gould et al. (2008) stated that this process is an important step for identifying heterogeneous and homogeneous types of items, and researchers can generate more valid items by revising and rewording them based on the results of a Q-sort (Little, Lindenberge, and Nesselroade 1999). After completing the Q-sort data collection, the researchers conducted a frequency analysis to confirm the matching percentages between the items and the constructs. Of the 69 items in the initial item pool, 49 items were retained with high (80%–100%) consensus percentages.

A second expert panel of scholars was then asked to review and to analyze each item and construct of the second item pool to further contribute to content validity. This panel of experts consisted of seven professors, two from tourism, three from leisure and recreation, one from sport marketing, and one from psychology. After their review, 16 items were reworded or modified, and 6 items were deleted since the meanings of the questions overlapped and/or contained vague noun phrases. The resulting pool containing 43 items was retained and used for the pilot test in which all nostalgia scale items were measured on 7-point Likert-type scale, ranging from 1 = strongly disagree to 7 = strongly agree. Next, a pilot test was conducted because as McMillan and Schumacher (1989) maintain, it is an effective method for testing an initial proposed model. The targeted samples for the pilot study were nonstudent sport tourists from out of town at a college football game. One hundred sixty-two responses were collected for the pilot test, and seven responses that were not more than 50% completed were excluded. The systematic sampling technique was used for gathering the data. Specifically, to collect the data, first, a research team member approached an individual in a parking lot and inquired about his or her willingness to participate in a survey. If they agreed to fill out the surveys, the research team member briefly explained the content of the study. The reason the research team member asked individuals in each group was to measure group effects on each individual through multilevel analysis. After the first group finished the survey, a research assis tant moved in a clockwise direction and asked individuals in a group to conduct a survey who parked at every third parking space. The model is tested for reliability and dimensionality, with the pilot test being used to improve instrument validity and provide more important information about the model (Gay 1996). The Statistical Package for the Social Science (SPSS 18.0) was used to analyze the data, including descriptive statistics, exploratory factor analysis (EFA), and reliability tests. After conducing EFA on the nostalgia scale, the researchers conducted confirmatory factor analysis (CFA) using EQS 6.2. In this study, EFA was necessary because, to date, a nostalgia scale has not yet been developed in the field of sport tourism. The findings of EFA indicated that of the 43 items, 19 were included in the nostalgia as experience factor; however, the meanings of the items were different and diverse (e.g., sport players, teams, coaches, stadium, and atmosphere). Thus, in order to determine the number of factors, this study conducted a parallel analysis and a scree test. Parallel analysis is a comparison of the observed component variance (eigenvalues) with a random analysis of 1,000 data sets with characteristics similar to the sample. Based on the results of the scree plot and by comparing initial eigenvalues for factors with random data eigenvalues, the evidence supported five components for the nostalgia scale. Once the number of factors was determined, the principal axis factoring (PAF) procedure with oblique (Promax) rotation was used to conduct the EFA.

The results indicated that the initial 19 items reflecting nostalgia as experience were divided into two factors. Based on previous research and the content of these items, the researchers identified these two separate factors from the “nostalgia as experience” factor as “nostalgia as sport team” and “nostalgia as environment.” In addition, the results of the EFA indicated that one item cross-loaded, and four items exhibited low factor loadings. Thus, these five items were deleted, and a reliability test and CFAs were conducted to test the nostalgia construct.

Based on the CFA, the initial model fit the data poorly: χ²(df) = 1,787.79(655), root mean square error of approximation (RMSEA) = .079, non-normed fit indices (NNFI) = .774, comparative fit index (CFI) = .842. Eight items that showed low reliability were, thus, deleted. The revised model exhibited an improved fit: χ²(df) = 786.75 (390), RMSEA = .060, NNFI = .902, CFI = .931. Finally, 30 items remained, all of which for the NSST presented appropriate reliability and validity values.

Before performing the main data analysis, data screening processes were conducted to eliminate extreme outliers based on Mahalanobis distance, and missing values were imputed using the expectation maximization (EM) algorithm. In addition, univariate normality was also evaluated by examining skewness and kurtosis using SPSS 18.0, and multivariate normality using Mardia’s (1985) multivariate kurtosis coefficients. If the normality of distribution is violated, Satorra-Bentler scaled statistic (S-B χ²) (Satorra and Bentler 1994) and robust standard errors (Bentler and Dijkstra 1985) are used to provide robust estimates (Bentler 2005).

A Multilevel Approach

A multilevel CFA with robust maximum likelihood estimation was performed to explain group effects as a single-level CFA cannot assess them. For example, most people attend home football games with their family, friends, or someone close to them. In this case, individuals in the same group are apt to have common characteristics or experiences, and the diverse characteristics of groups affect the responses of individuals within groups differently. As Allua (2007) noted, dependencies of data should be considered to avoid incorrect results drawn from the results of an inflated model chi-square statistic, standard errors, and parameters biases. Moreover, multilevel data structures should be modeled using hierarchical covariance modeling to prevent incorrect interpretations from biased results (Raudenbush and Bryk 2002). To avoid these biases, this study performed multilevel CFA by considering both variation among groups and variation among individuals within groups. In other words, there are two observed variables, travel group and individual observations. Below is a representation of a multilevel CFA.

η_{g i} = α + η_{B g} + η_{W i}

where $η_{g i}$ represents the latent factor, $α$ the vector of intercepts or overall expectation, $η_{B g}$ a random factor that models group effects, and $η_{W i}$ a random factor that varies over individuals within groups (Muthén 1994).

Before conducting multilevel CFA, an intraclass correlation coefficient (ICC) was examined to identify nesting or if multilevel CFA was necessary. The ICC, the ratio of the between-group variance to total variance, is presented in the equation format below (Muthén 1989, 1991). According to Muthén (1997), a multilevel analysis is required if the ICC values are larger than 0.1. In addition, Preacher, Zhang, and Zyphur (2011) suggested that an ICC value of .05 is small, of .10, medium, and of .20, large.

I C C = \frac{σ_{B}^{2}}{σ_{B}^{2} + σ_{W}^{2}}

where:

$σ_{B}^{2}$ = between-group variance

$σ_{W}^{2}$ = within-group variance

Absolute fit and comparative fit indices were used to evaluate goodness of fit. First, for the absolute fit, the chi-square (χ²) statistic was assessed to investigate overall fit. However, the chi-square statistic is sensitive and influenced by sample size (Kenny 2014). Therefore, this study used the RMSEA and standardized root mean squared residual (SRMR). A good fit for the RMSEA value is less than 0.06 (Hu and Bentler 1999), while Browne and Cudeck (1992) suggest that the RMSEA values of less than 0.08 can be considered as a reasonable fit. In addition, the SRMR values of less than 0.08 indicate a good fit (Hu and Bentler). Second, for the comparative fit indices, NNFIs and CFI were used, both of which should be greater than 0.9, the value that indicates acceptable fit (Marsh and Hau 1996).

This study performed validity and reliability tests for multilevel CFAs. More specifically, the validity test, construct (convergent and discriminant) validity, and criterion validity were evaluated. This study used AVE values and each indicator’s coefficient on each construct to test convergent validity. Next, the correlations among each factor and the square root of the AVEs of each construct were employed to test discriminant validity. Third, the relationship between the five scales of nostalgia and the three external variables (i.e., cognitive attitude, behavioral attitude, intention to visit football-related places) were assessed using Pearson’s r statistic to identify criterion validity.

Results

Identical to the questionnaire pilot study’s sampling procedure, the targeted sample for the main study was nonstudent sport tourists from out of town at college football games. Data were obtained using a systematic sampling technique. Face-to-face surveys were conducted near parking/tailgating areas on the university campus. According to the content validity of the intended measures, the collected samples were also filtered based on two criteria: definition of nostalgia and sport tourism.

First, this study defined nostalgia as longing for the past with positive memories. Thus, target samples should have positive memories for the past to qualify under the criteria of nostalgia. Participants without nostalgic memories were excluded from the study.

Second, Gibson (1998) defined sport tourism as “leisure-based travel that takes individuals temporarily outside of their home communities to participate in physical activities, to watch physical activities, or to venerate attractions associated with physical activities” (p. 49). This study asked respondents to identify their current residency. Answers to this question were classified into three groups: students at a university, residents of near the host city (Pickens, Oconee, or Anderson County), and neither of them. Based on the respondents’ answers, this study only used the data from nonresidents of the host city. The benefit of this filtering process was to identify the impact of college football games on the host community by measuring the behavioral intentions of sport tourists. According to Daniels (2002), a county is an effective territorial division to set the boundary of an economic impact study measuring tourists’ impact to a host community. The host city is located in Pickens County, but there is a high probability that individuals who reside in Oconee or Anderson County also consider the host city as their home community. Following Gibson’s suggestions, we excluded the three counties located within 50 miles of the stadium.

In sum, this study collected 985 responses from five college football games for a total response rate of 84.7%. Out of the 985 collected questionnaires, 71 with less than 50% complete and 2 with extreme multivariate outliers based on the Mahalanobis distance were excluded from this study. Further, 16 respondents who answered that they did not have any positive memories regarding college football games and respondents who answered that they were students of the university and/or residents of Pickens, Oconee, or Anderson County were eliminated for this research. As a result, 619 responses were used for the main study.

Population

Of the 619 respondents, 55.3% were males and 44.2% females. Age was reported as categorical variable, completed as an open-ended question. Based on the results of frequency analysis, the age variable was recoded into the categorical variable of 18 to 22 (5.5%), 23 to 29 (26.2%), 30 to 39 (18.7%), 40 to 49 (16.0%), 50 to 59 (20.0%), 60 to 69 (9.4%), and 70 and older (1.9%). The average age of the entire sample was 40.31. As for marital status, married (62.5%) was the highest followed by single, never married (27.9%), separated/divorced (6.6%), and widowed (1.9%). Most attendees attended football events as a group, such as family (30.7%), friends (19.5%), and family and friends (48.5%). Only 1.3% of respondents attended alone (see Table 2).

Table 2.

Demographic Factors.

Variable	Category	N (%)	Total
Gender	Male	342 (55.3)	619
	Female	274 (44.2)
	No response	3 (0.5)
Age	18–22	34 (5.5)	619
	23–29	162 (26.2)
	30–39	116 (18.7)
	40–49	99 (16.0)
	50–59	124 (20.0)
	60–69	58 (9.4)
	≥70	12 (1.9)
	No response	14 (2.3)
Marital status	Single, never married	173 (27.9)	619
	Married	387 (62.5)
	Separated/divorced	41 (6.6)
	Widowed	12 (1.9)
	No response	6 (1.0)
Type of group	Family	190 (30.7)	619
	Friends	121 (19.5)
	Alone	8 (1.3)
	Family and friends together	300 (48.5)

Multilevel Analysis

Before conducting multilevel CFA, model-based interclass correlations coefficients (ICCs) were estimated to identify nesting at the group level. If the ICC values are nonzero (<.05 or .10), then multilevel CFA should be performed to estimate individual and group effects. Table 3 displays the results of the model-based ICCs. According to the results, the ICC values of the most variables are greater than .1 except for NFI 5. Thus, more than 10% of the variance in the responses are due to group membership. As these results indicate, multilevel CFA is the appropriate approach (Muthén 1997).

Table 3.

Interclass Correlation Values of Variables.

Model-Based Interclass Correlation Coefficients
NST1	NST2	NST3	NST4	NST5	NE1	NE2	NE3	NE4	NS1
.274	.262	.205	.151	.153	.185	.113	.152	.173	.133
NS2	NS3	NS4	NPI1	NPI2	NPI3	NPI4	NPI5	NPI6	NPI7
.196	.128	.226	.156	.207	.167	.223	.096	.102	.146
NGI1	NGI2	NGI3	NGI4	NGI5	NGI6	NGI7	NGI8	NGI9
.110	.123	.153	.142	.362	.306	.308	.353	.270

Note: NST = Nostalgia as Sport Team; NE = Nostalgia as Environment; NS = Nostalgia as Socialization; NPI = Nostalgia as Personal Identity; NGI = Nostalgia as Group Identity.

An initial multilevel CFA was conducted to verify model fit indices. The multilevel CFA model showed acceptable fit for the data: χ²(df) = 1920.887 (734), RMSEA = .072, SRMR = .035, NNFI = .909, CFI = .918. Next, we assessed convergent validity, discriminant validity, and internal consistency of both level 1 and level 2. In the level 1 and level 2 models, the Rho α coefficients of the total measurement model were .969 and .992, respectively. The factor loadings, the α coefficients, the Rho, and the AVEs of both the level 1 and level 2 models are displayed in Table 4.

Table 4.

Factor Loadings, Reliability Coefficients, and AVEs of a Multilevel Model.

Factors and itemsMy pleasant feelings for the past are evoked by ___________.	λ		α		Rho		AVE
	Level 1	Level 2	Level 1	Level 2	Level 1	Level 2	Level 1	Level 2
Nostalgia as Sport Team			.888	.966	.889	.967	.616	.855
(NST 01) remembering my favorite athletes	.773	.968
(NST 02) my favorite team from the past	.707	.942
(NST 03) a particular player with a good reputation	.873	.941
(NST 04) a particular coach with a good reputation	.806	.931
(NST 05) a particular team with a good reputation	.755	.835
Nostalgia as Environment			.872	.929	.880	.934	.655	.781
(NE 01) the music played at the sport venue	.528	.679
(NE 02) the appearance of the sport venue	.828	.889
(NE 03) the size of the sport venue	.929	.968
(NE 04) the design of the sport venue	.890	.968
Nostalgia as Socialization			.919	.926	.920	.927	.616	.762
(NS 01) positive memories shared with others at previous sports event	.776	.764
(NS 02) memories of building friendships with others at previous sports events	.847	.915
(NS 03) memories of socializing with others at previous sports events	.929	.906
(NS 04) memories of making friends through sport events	.890	.899
Nostalgia as Personal identity			.926	.968	.927	.968	.648	.815
(NPI 01) identifying myself as a sport fan	.715	.854
(NPI 02) how important I am to the team or players as a sport fan	.653	.916
(NPI 03) feelings of being a sport fan	.829	.927
(NPI 04) a sense of accomplishment as a sport fan	.827	.915
(NPI 05) pride of being a sport fan	.903	.903
(NPI 06) a feeling of satisfaction as a sport fan	.893	.933
(NPI 07) positive feelings about myself as a sport fan	.785	.867
Nostalgia as Group identity			.930	.979	.935	.980	.619	.846
(NGI 01) experiences of group bonding at past sporting event	.824	.849
(NGI 02) the fact that a particular sport team or player has enhanced a community’s reputation	.727	.768
(NGI 03) unique characteristics of my sport event social group in comparison to sport event social groups from the past	.817	.895
(NGI 04) the traditions of my group while attending a sporting event	.861	.935
(NGI 05) memories of being accepted as a member of my group at past sporting event	.708	.970
(NGI 06) pride of being a part of my group at past sporting event	.830	.968
(NGI 07) group rituals at past sporting event	.784	.968
(NGI 08) behavior my group expected at past sporting event	.659	.958
(NGI 09) shared memories which affect my group identity	.732	.946

At level 1, the AVEs of all factors are larger than .5, indicating acceptable convergent validity. Rho and α coefficients range from .880 for environment to .935 for group identity, indicating good internal consistency. The square root of AVE scores of each factor were greater than the correlations between all pairs of factors, providing evidence for discriminant validity (see Table 5).

Table 5.

Correlations among All Constructs in the Level 1 Model.

	1	2	3	4	5
1. Nostalgia as sport team	.785^a
2. Nostalgia as environment	.627	.809^a
3. Nostalgia as socialization	.509	.432	.785^a
4. Nostalgia as personal identity	.654	.613	.498	.805^a
5. Nostalgia as group identity	.575	.487	.757	.734	.787^a

Square root of average variance extracted.

For the level 2 model, AVE scores ranged from .762 for socialization to .855 for sport team, and all Rho and α coefficients were greater than .7, indicating good convergent validity and internal consistency as seen in Table 6. In addition, the square root of AVE scores of each factor were greater than the correlations between all pairs of factors, indicating acceptable discriminant validity. As these results from the multilevel analysis show, all AVEs and reliability coefficients indicate strong evidence of convergent validity and internal consistency in both level 1 and level 2 models. In addition, the correlations between factors were less than the square root AVEs of each relevant factor in both level 1 and level 2 models, indicating good discriminant validity.

Table 6.

Correlations among All Constructs in the Level 2 Model.

	1	2	3	4	5
1. Nostalgia as sport team	.925^a
2. Nostalgia as environment	.764	.884^a
3. Nostalgia as socialization	.507	.696	.873^a
4. Nostalgia as personal identity	.888	.824	.690	.903^a
5. Nostalgia as group identity	.711	.771	.845	.854	.920^a

Square root of average variance extracted.

Criterion validity was evaluated by comparing the five factors of nostalgia to the three external variables at both level 1 and level 2: Cognitive attitude (a five-item scale with a Rho coefficient in the level 1 model = .837, a Rho coefficient in the level 2 model = .941), behavioral attitude (a seven-item scale with a Rho coefficient in the level 1 model = .916, a Rho coefficient in the level 2 model = .959), and intention to visit football-related places (a four-item scale with a Rho coefficient in the level 1 model = .854, a Rho coefficient in the level 2 model = .977). At level 1, all subscales of nostalgia showed a significant correlation with cognitive attitude (range: r = .350 for nostalgia as environment to .551 for nostalgia as personal identity), behavioral attitude (range: r = .417 for nostalgia as environment to .631 for nostalgia as personal identity), and intention to visit football-related places (range: r = .253 for nostalgia as socialization to .408 for nostalgia as sport team) (see Table 7). At level 2, all subscales of nostalgia were significantly correlated with cognitive attitude (range: r = .502 for nostalgia as sport team to .673 for nostalgia as group identity), behavioral attitude (range: r = .468 for nostalgia as socialization to .611 for nostalgia as personal identity), and intention to visit football-related places (range: r = .490 for nostalgia as socialization to .678 for nostalgia as sport team) (see Table 8). The significant correlations between the nostalgia factors and the criterion variables are indicative of criterion validity for the nostalgia scale.

Table 7.

Assessment of Criterion Validity in the Level 1 Model.

Pearson Correlation	CA	BA	FRP
1. Nostalgia as Sport Team	.471**	.579**	.408**
2. Nostalgia as Environment	.350**	.417**	.264**
3. Nostalgia as Socialization	.376**	.556**	.253**
4. Nostalgia as Personal Identity	.551**	.631**	.390**
5. Nostalgia as Group Identity	.526**	.627**	.356**

Note: CA = Cognitive Attitude; BA = Behavioral Attitude; FRP = Football Related Places to visit.

Correlation is significant at the 0.01 level (two-tailed).

Table 8.

Assessment of Criterion Validity in the Level 2 Model.

Pearson Correlation	CA	BA	FRP
1. Nostalgia as Sport Team	.502**	.557**	.650**
2. Nostalgia as Environment	.568**	.514**	.678**
3. Nostalgia as Socialization	.517**	.468**	.490**
4. Nostalgia as Personal Identity	.621**	.611**	.602**
5. Nostalgia as Group Identity	.673**	.561**	.499**

Note: CA = Cognitive Attitude; BA = Behavioral Attitude; FRP = Football Related Places to visit

Correlation is significant at the 0.01 level (two-tailed).

Discussion

The main purpose of this study was to develop a theoretically based measure of the NSST. This study initially developed nostalgia items based on a literature review and then conducted EFAs and CFAs in a pilot study. As Teijlingen and Hundley (2001) explained, conducting a pilot study is important since it provides “advance warning about where the main research project could fail, where research protocols may not be followed, or whether proposed methods or instruments are inappropriate or too complicated” (p. 1). After conducting a pilot test, this study performed multilevel CFA in the main study.

As Floyd and Widaman (1995) asserted, identifying dimensions of a domain and data reduction are the main primary functions of EFA. Based on the results of the EFA from the pilot study, the main study found that there were five subfactors in the NSST rather than only the four found in the classification of nostalgia in the sport tourism context. In the classification of nostalgia, the first category, nostalgia as experience, was divided here into two, “nostalgia as sport team” and “nostalgia as environment.” Even though the number of subfactors from the results of the EFA was different from the number of factors indicated in the classification of nostalgia, the concepts of nostalgia as sport team and nostalgia as environment matched with the concept of nostalgia as experience, which is one factor in the classification of nostalgia. Nostalgia as experience is defined as being based on one’s past experience and on an individual’s desire to pursue experience for its own sake, creating nostalgic feelings regarding sport objects including athletes, teams, and sport venues. In other words, nostalgia as sport team factor focuses on athletes and teams, whereas nostalgia as environment emphasizes sport venue and atmosphere. Finally, the study found five subfactors of nostalgia: nostalgia as sport team, nostalgia as environment, nostalgia as socialization, nostalgia as personal identity, and nostalgia as group identity. After conducting an EFA, this study performed CFA to specify the models. In total, 13 items with poor psychometric properties were deleted based on the results of the EFA and CFA.

In the main study, multilevel CFA was conducted to clarify group effects since most respondents were nested in a group. Multilevel analysis is used to analyze hierarchically structured data, meaning that by using it, this research could avoid biased results triggered by the shared common characteristics within groups. In the multilevel CFA, the interclass correlation coefficients should be examined to identify the variation in responses due to group membership. Initially, from the results of multilevel CFA, this study found acceptable model fit and that all scales had good reliability, convergent validity, discriminant validity, and criterion validity. The results of the multilevel CFA from the main study indicated that the final scale has adequate psychometric properties. First, the scale established content validity by a review of the literature, Q-sort, and expert review. Second, reliability values for all constructs were higher than the accepted cutoff value (α > .7). Third, AVE values for all constructs were greater than the cutoff criterion, indicating good convergent validity. Fourth, discriminant validity was established by comparing the square root of AVE values for all factors to the correlations between each factor. Fifth, criterion validity was evidenced by the significant correlations between the predictors and the external variables. Therefore, the final scale demonstrated good psychometric properties and significant structural relations with the three external constructs.

Holak, Havlena, and Matveev’s (2006) scale had three major limitations to be applied to sport travelers: it ignored direct experiences and was not able to distinguish direct and indirect experiences; there were cross-loadings of items across four unnamed factors that made the operationalization of the factors ambiguous; and scales showed low reliability. In this study, we operationalized experiences of sport travelers through a mixed method and identified five specific reflective factors. Results of the multilevel approach showed adequate convergent and discriminant validity with high reliability. Furthermore, the NSST had significant associations with criterion variables of cognitive attitude, behavioral attitude, and intention to visit football-related places. Consequently, the NSST captured the unique features of nostalgia related to sports (e.g., Stewart and Smith 1999) and this specific categorization of nostalgia was rigorously assessed by a multilevel analysis. Development of this quantitative scale can expand the boundaries of nostalgia studies in sport tourism that were mostly investigated with qualitative methods (Fairley 2003; Fairley and Gammon 2005).

Limitations and Suggestions for Future Research

The current study is not without limitations. Foremost, the NSST was developed only using football travelers. Thus, further assessment of external validity is necessary for generalizability of the scale. For instance, the scale should be tested in other football settings and across other disciplines of sport. Moreover, while the investigation focuses on travel research, cross-validation between sport tourists and locals need to be conducted. Within the scope of the current study, characteristics of fans who are not tourists cannot be explained. This needs to be addressed in future studies. Researchers should expand and compare nostalgia experience in other tourism contexts. Further, there can be more dynamics in travelers’ motivation. In some cases, for example, travelers might hold nostalgic feelings but not be willing to travel. The constructs were sorted from responses derived from semi-structured interviews by researchers in the field of parks, recreation and tourism. There could be a missing motive that is unexplored and bias in the sorting implemented by the researchers within a single field. Future studies should investigate such dynamics of tourist behaviors that may be missing in this study. In addition, while the NSST factors were significantly associated with criterion variables of cognitive attitude, behavioral attitude, and intention to visit football related places, results of this cross-sectional study do not prove causation and the antecedents and consequences of sport nostalgia need further investigation. Also, our sample showed sport tourists, responding to the NSST, visited the football venue in groups. Investigation about the social network and how a vestige of nostalgia can be succeed to others—perhaps compared across other levels of competition or other sport disciplines—could provide further practical implications.

Conclusion

Attention given to studying sport nostalgia has increased in the field of sport tourism. This study developed the NSST using the best practices in scale development. It demonstrated convergent validity, discriminant validity, criterion validity, and internal consistency, indicating that it is an appropriate tool for measuring nostalgia in sport settings. The scale developed here theoretically and practically contributes to the increasing knowledge about nostalgia and to identifying the role of nostalgia in sport tourism. From a theoretical standpoint, this research provides a measurement scale, which is important because one had not yet been developed. Thus, future research can apply the NSST scale in diverse empirical research of sport nostalgia to understand and measure sport tourists’ behavior. Furthermore, the effect of direct and indirect experience on the degree of nostalgia needs to be examined as two concepts of nostalgia that are conceptually distinct. From the practical standpoint, after further assessments of external validity, sport event managers and administrators can employ the NSST scale to investigate the economic impact of the sport consumer caused by nostalgic behavior. As both of these perspectives suggest, this study adds to the growing body of research being conducted on nostalgia in the field of sport tourism.

Footnotes

Authors’ Note

Heetae Cho is now working at Nanyang Technological University (starting from 2017 spring).

Declaration of Conflicting Interests

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

Funding

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

Author Biographies

Heetae Cho is an assistant professor of Sport Management in the Department of Physical Education and Sports Science at Nanyang Technological University. His major area is sport tourism.

Hyun-Woo Lee is an assistant professor of Sport Management in the School of Health and Kinesiology at Georgia Southern University. His area of interest is human emotion and fan behavior.

DeWayne Moore is a professor in the Department of Psychology at Clemson University. His research interests is in quantitative methods.

William C. Norman is a professor in Clemson University’s Department of Parks, Recreation and Tourism Management and the former director of the Recreation, Travel and Tourism Institute. Dr. Norman works closely with graduate students from around the world conducting theoretically-based and applied recreation and tourism research that assists South Carolina’s tourism industry and leisure service providers as well as other industry sectors (e.g., aquaculture).

Gregory Ramshaw is an associate professor in the Department of Parks, Recreation and Tourism Management. His research explores the social construction and cultural production of heritage, with a particular interest in sport-based heritage.

References

Aden

R. C.

1995. “Nostalgic Communication as Temporal Escape: When It was a Game’s Reconstruction of a Baseball Work Community.” Western Journal of Communication 59 (1): 20–38.

Allua

S. S.

2007. “Evaluation of Single- and Multilevel Factor Mixture Model Estimation.” PhD diss., the University of Texas, Austin.

Babbie

E. R.

2010. Introduction to Social Research. Belmont. CA: Thompson Wadsworth.

Barclay

C. R.

DeCooke

P. A.

1988. “Ordinary Everyday Memories: Some of the Things of Which Selves are Made.” In Remembering Reconsidered: Ecological and Traditional Approaches to the Study of Memory, edited by Neisser

Winograd

, 91–125. New York: Cambridge University Press.

Batcho

K. I.

1995. “Nostalgia: A Psychological Perspective.” Perceptual and Motor Skills 80 (1): 131–43.

Batcho

K. I.

2007. “Nostalgia and the Emotional Tone and Content of Song Lyrics.” American Journal of Psychology 120 (3): 361–81.

Bentler

P. M.

2005. EQS 6 Structural Equations Program Manual. Encino, CA: Multivariate Software.

Bentler

P. M.

Dijkstra

1985. “Efficient Estimation via Linearization in Structural Models.” In Multivariate Analysis VI, edited by Krishnaiah

P. R.

, 9–42. Amsterdam, the Netherlands: North-Holland.

Browne

M. W.

Cudeck

1992. “Alternatives Ways of Assessing Model Fit.” Sociological Methods and Research 21:230–58.

10.

Brown

N. R.

Schopflocher

1998. “Event Clusters: An Organization of Personal Events in Autobiographical Memory.” Psychological Science 9 (6): 470–75.

11.

Castelnuovo-Tedesco

1980. “Reminiscence and Nostalgia: The Pleasure and Pain of Remembering.” In The Course of Life: Psychoanalytic Contributions toward Understanding Personality Development, edited by Greenspan

S. I.

Pollack

G. H.

, 104–18. Washington, DC: U.S. Government Printing Office.

12.

Cavanaugh

J. C.

1989. “I Have this Feeling about Everyday Memory Aging.” Educational Gerontology 15 (6): 597–605.

13.

Chase

Shaw

1989. “The Dimensions of Nostalgia.” In The Imagined Past: History and Nostalgia, edited by Shaw

Chase

, 1–17. Manchester: Manchester University Press.

14.

Cho

Ramshaw

Norman

2014. “A Conceptual Model for Nostalgia in the Context of Sport Tourism: Re-classifying the Sporting Event.” Journal of Sport & Tourism 19:145–67.

15.

Daniels

M. J.

2002. “A Comparison of Social Accounting Matrix and Occupation-Based Modeling Techniques for Estimating the Personal Income Distribution Effects of Sport Tourism Events across Different Household Income Levels.” PhD diss., Clemson University, Clemson, SC.

16.

Davis

1979. Yearning for Yesterday: A Sociology of Nostalgia. New York: Free Press.

17.

Edwards

1973. Sociology of Sport. Homewood, IL: Dorsey.

18.

Fairley

2003. “In Search of Relived Social experience: Group-Based Nostalgia Sport Tourism.” Journal of Sport Management 17 (3): 284–304.

19.

Fairley

Gammon

2005. “Something Lived, Something Learned: Nostalgia’s Expanding Role in Sport Tourism.” Sport in Society 8 (2): 182–97.

20.

Floyd

F. J.

Widaman

K. F.

1995. “Factor Analysis in the Development and Refinement of Clinical Assessment Instruments.” Psychological Assessment 7 (3): 286–99.

21.

Funk

D. C.

James

J. D.

2006. “Consumer Loyalty: The Meaning of Attachment in the Development of Team Allegiance.” Journal of Sport Management 20 (2): 189–217.

22.

Gay

L. R.

1996. Educational Research: Competencies for Analysis and Application. Englewood Cliffs, NJ: Prentice-Hall.

23.

Gibson

H. J.

1998. “Sport Tourism: A Critical Analysis of Research.” Sport Management Review 1 (1): 45–76.

24.

Gibson

H. J.

Willming

Cynthia

Holdnak

2003. “Small-scale Event Sport Tourism: Fans as Tourists.” Tourism Management 24 (2): 181–90.

25.

Gordon

K. O.

2013. “Emotion and Memory in Nostalgia Sport Tourism: Examining the Attraction to Postmodern Ballparks through an Interdisciplinary Lens.” Journal of Sport & Tourism 18 (3): 217–39.

26.

Gould

J. M.

Moore

McGuire

Stebbins

2008. “Development of the Serious Leisure Inventory and Measure.” Journal of Leisure Research 40 (1): 47–68.

27.

Havlena

W. J.

Holak

S. L.

1991. “The Good Old Days: Observations on Nostalgia and Its Role in Consumer Behavior.” Advances in Consumer Research 18 (1): 323–29.

28.

Havlena

W. J.

Holak

S. L.

1996. “Exploring Nostalgia Imagery through the Use of Consumer Collages.” In Advances in Consumer Research, edited by Corfman

K. P.

Lynch

J. G.

Jr. 35–42. Prove, UT: Association for Consumer Research.

29.

Hinch

Higham

2001. “Sport Tourism: A Framework for Research.” International Journal of Tourism Research 3 (1): 45–58.

30.

Hogg

M. A.

Terry

D. J.

White

K. M.

1995. “A Tale of Two Theories: A Critical Comparison of Identity Theory with Social Identity Theory.” Social Psychology Quarterly 58:255–69.

31.

Holak

S. L.

Havlena

W. J.

1998. “Feelings, Fantasies, and Memories: An Examination of the Emotional Components of Nostalgia.” Journal of Business Research 42 (3): 217–26.

32.

Holak

S. L.

Havlena

W. J.

Matveev

A. V.

2006. “Exploring Nostalgia in Russia: Testing the Index of Nostalgia-Proneness.” European Advances in Consumer Research 7:195–200.

33.

Holbrook

M. B.

1993. “Nostalgia and Consumption Preferences: Some Emerging Patterns of Consumer Tastes.” Journal of Consumer Research 20:245–56.

34.

Holbrook

M. B.

1994. “Nostalgia Proneness and Consumer Tastes.” Buyer Behavior in Marketing Strategy 2:348–64.

35.

Bentler

P. M.

1999. “Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria versus New Alternatives.” Structural Equation Modeling 6:1–55.

36.

Jenkins

1996. Social Identity. New York: Routledge.

37.

Kaplan

H. A.

1987. “The Psychopathology of Nostalgia.” Psychoanalytic Review 74:465–86.

38.

Kaplanidou

Vogt

2007. “The Interrelationship between Sport Event and Destination Image and Sport Tourists’ Behaviours.” Journal of Sport & Tourism 12 (3-4): 183–206.

39.

Kelly

J. R.

1982. Leisure. Englewood Cliffs, NJ: Prentice Hall.

40.

Kenny

D. A.

2014. Measuring model fit. http://davidakenny.net/cm/fit.htm (accessed February 6, 2014).

41.

Kline

R. B.

2011. Principle and Practice of Structural Equation Modelling. New York: Guilford.

42.

Kulczycki

Hyatt

2005. “Expanding the Conceptualization of Nostalgia Sport Tourism: Lessons Learned from Fans Left Behind after Sport Franchise Relocation.” Journal of Sport & Tourism 10:273–93.

43.

Little

T. D.

Lindenberge

Nesselroade

J. R.

1999. “On Selecting Indicators for Multivariate Measurement and Modelling with Latent Variables: When ‘Good’ Indicators Are Bad and ‘Bad’ Indicators Are Good.” Psychological Methods 4 (2): 192–211.

44.

Mardia

K. V.

1985. “Mardia’s Test of Multinormality.” In Encyclopedia of Statistical Sciences, edited by Kotz

Johnson

N. L.

, 217–21. New York: John Wiley.

45.

Marsh

H. W.

Hau

1996. “Assessing Goodness of Fit: Is Parsimony Always Desirable?” Journal of Experimental Education 64:364–90.

46.

McMillan

J. H.

Schumacher

1989. Research in Education: A Conceptual Introduction. Glenview, IL: Scott Foresman.

47.

Menor

Roth

A. V.

2007. “New Service Development Competence in Retail Banking: Construct Development and Measurement Validation.” Journal of Operations Management 25 (4): 825–46.

48.

Muthén

B. O.

1989. “Latent Variable Modelling in Heterogeneous Populations.” Psychometrika 54 (4): 557–85.

49.

Muthén

B. O.

1991. “Multilevel Factor Analysis of Class and Student Achievement Components.” Journal of Educational Measurement 28:338–54.

50.

Muthén

B. O.

1994. “Multilevel Covariance Structure Analysis.” Sociological Methods & Research 22:376–98.

51.

Muthén

B. O.

1997. “Latent Variable Modelling of Longitudinal and Multilevel Data.” In Sociological Methodology 1997, edited by Raftery

A. E.

, 453–80. Washington, DC: American Sociological Association.

52.

Pascal

V. J.

Sprott

D. E.

Muehling

D. D.

2002. “The Influence of Evoked Nostalgia on Consumers’ Responses to Advertising: An Exploratory Study.” Journal of Current Issues & Research in Advertising 24 (1): 39–47.

53.

Phau

Quintal

Marchegiani

Lee

2016. “Looking beyond Pasta and Pizzas: Examining Personal and Historical Nostalgia as Travel Motives.” International Journal of Culture, Tourism and Hospitality Research 10 (3): 296–309.

54.

Preacher

K. J.

Zhang

Zyphur

M. J.

2011. “Alternative Methods for Assessing Mediation in Multilevel Data: The Advantages of Multilevel SEM.” Structural Equation Modeling 18:161–81.

55.

Ramshaw

2005. “Nostalgia, Heritage, and Imaginative Sports Geographies: Sport and Cultural Landscapes.” Paper presented at the Forum UNESCO University and Heritage 10th International Seminar: Cultural Landscapes in the 21st Century, Newcastle upon Tyne.

56.

Ramshaw

Gammon

2005. “More Than Just Nostalgia? Exploring the Heritage/Sport Tourism Nexus.” Journal of Sport Tourism 10 (4): 229–41.

57.

Raudenbush

S. W.

Bryk

A. S.

2002. Hierarchical Linear Models: Applications and Data Analysis Methods. Newbury Park, CA: Sage.

58.

Robinson

M. J.

Trail

G. T.

2005. “Relationships among Spectator Gender, Motives, Points of Attachment, and Sport Preference.” Journal of Sport Management 19:25–80.

59.

Rodrigues

2012. “Constraints Nostalgia and North Americans’ Visits to Rural Europe.” European Journal of Tourism Research 5 (2): 186–90.

60.

Routledge

Arndt

Sedikides

Wildschut

2008. “A Blast from the Past: The Terror Management Function of Nostalgia.” Journal of Experimental Social Psychology 44:132–40.

61.

Satorra

Bentler

P. M.

1994. “Correction to Test Statistics and Standard Errors in Covariance Structure Analysis.” In Latent Variables Analysis: Applications for Developmental Research, edited by Eye

A. V.

Clogg

C. C.

, 399–419. Thousand Oaks, CA: Sage.

62.

Schindler

R. M.

Holbrook

M. B.

2003. “Nostalgia for Early Experience as a Determinant of Consumer Preference.” Psychology & Marketing 20:275–302.

63.

Schurr

K. A.

Wittig

Terry F.

Ruble

V. E.

Ellen

A. S.

1988. “Demographic and Personality Characteristics Associated with Persistent, Occasional, and Non-attendance of University Male Basketball Games by College Students.” Journal of Sport Behavior 11:3–17.

64.

Sedikides

Wildschut

Baden

2004. “Nostalgia: Conceptual Issues and Existential Functions.” In Handbook of Experimental Existential Psychology, edited by Greenberg

Koole

S. L.

Pyszczynski

, 200–14. New York: Guilford.

65.

Sellick

M. C.

2004. “Discovery, Connection, Nostalgia: Key Travel Motives within the Senior Market.” Journal of Travel & Tourism Marketing 17 (1): 55–71.

66.

Skowronski

Walker

Betz

2003. “Ordering Our World: An Examination of Time in Autobiographical Memory.” Memory 11 (3): 247–60.

67.

Smith

A. C. T.

Stewart

2007. “The Travelling Fan: Understanding the Mechanisms of Sport Fan Consumption in a Sport Tourism Setting.” Journal of Sport & Tourism 12 (3/4): 155–81.

68.

Stern

B. B.

1992. “Historical and Personal Nostalgia in Advertising Text: The Fin de Siecle Effect.” Journal of Advertising 21 (4): 11–22.

69.

Stets

J. E.

Burke

P. J.

2000. “Identity Theory and Social Identity Theory.” Social Psychology Quarterly 63:224–37.

70.

Tajfel

1981. Human Groups and Social Categories: Studies in Social Psychology. Cambridge: Cambridge University Press.

71.

Teijlingen

Hundley

2001. “The Importance of Pilot Studies.” Social Research Update 35:1–4.

72.

Uhrich

Benkenstein

2010. “Sport Stadium Atmosphere: Formative and Reflective Indicators for Operationalizing the Construct.” Journal of Sport Management 24:211–237

73.

Wann

D. L.

Branscombe

N. R.

1993. “Sports Fans: Measuring Degree of Identification with Their Team.” International Journal of Sport Psychology 24:1–17.

74.

Wilson

J. L.

2005. Nostalgia: Sanctuary of Meaning. Cranbury, NJ: Bucknell University Press.

75.

Zait

Bertea

P. E.

2011. “Methods for Testing Discriminant Validity.” Management & Marketing 9 (2):217–24.