Publié par :Published by :Publicación de la :

Faculté des sciences de l’ administrationUniversité LavalQuébec (Québec) Canada G1K 7P4Tél. Ph. Tel. : (418) 656-3644Fax : (418) 656-2624

Édition électronique :Electronic publishing :Edición electrónica :

Céline FrenetteVice-décanat à la recherche et au développementFaculté des sciences de l’ administration

Disponible sur Internet :Available on InternetDisponible por Internet :

http ://www.fsa.ulaval.ca/rdrd@fsa.ulaval.ca

DOCUMENT DE TRAVAIL 1998-003

DIMENSIONS OF THE CONSUMERS ORIENTATION TOWARDS

SPORTING EVENTS : PROPOSITION AND VALIDATION OF THE

OSE SCALE

Frank Pons, Simon Nyeck

Version originale :Original manuscript :Version original :

ISBN - 2-89524-034-5ISBN -ISBN -

Série électronique mise à jour :One-line publication updated :Seria electrónica, puesta al dia

01-1998

1

Dimensions of the consumers Orientation towards Sporting Events : Proposition andvalidation of the OSE scale.

Frank PonsSimon Nyeck1

Université Laval, Québec

1 Université Laval, FSA, G1K7P4, Québec, Qc, Canada. Tel :418 656 2688. Fax : 418 656 2624.E-mail : simon.nyeck@mrk.ulaval.ca. Support for this research from the Fonds pour la Formation de Chercheurset l'Aide à la Recherche (FCAR grant) to the 2nd author is gratefully acknowledged.

2

Dimensions of Consumers Orientation towards Sporting Events (OSE) : Proposition and

Validation of a Scale.

Abstract :

Beside the growing importance of leisure activities in western society, consumers researchers have

paid little attention to sporting events. This research underlines the subjective dimensions of of

consumers orientation towards sporting events. Based on two studies conducted in Quebec, a

confirmatory factor analysis is used to assess the validity of the OSE scale. Lastly, the managerial

implications of the findings are discussed.

Orientation des consommateurs par rapport aux évènements sportifs (OES) : Proposition et

validation d’un outil de mesure

Résumé :

Malgré l'importance grandissante des activités de loisirs dans la société occidentale, très peu de

recherches en comportement du consommateur se sont intéressées aux événements sportifs. Cette

recherche met en évidence les dimensions déterminant l’orientation des consommateurs par

rapport à l'événement sportif. Sur la base de 2 études réalisées au Québec, l’analyse factorielle

confirmatoire permet d’évaluer la validité de l’outil de mesure proposé (OES). Enfin, les

implications pratiques sont discutées.

3

1. Introduction

Under globalization influences such as satellite television, travel and movies, the sporting eventis now a worldwide activity recognized as one of the main and best communication mediabetween individuals of all kind. The social and economic impact of sport in our societies hasbeen emphasized by several authors (Gilbert, 1995; Kindall, 1983; Lever, 1983). In fact, sport isthe key to important, financial and economical, stakes. For example, it generated 60 milliondollars for the american soccer federation during the last world cup (Gilbert, 1995), the televisioncoverage fees were approximatively about 907 million dollars for the Atlanta 1996 OlympicGames (Fortin, Dubois, 1995), in 1994 the National Basket-Ball Association covered 141countries and reached over 400 million people. Sport is also an important part of the culture andthe values of many countries. More than just a communication medium, it is also compared to themodern equivalent of the traditional religious ritual because of its ability to transmit and todevelop values and traditions in a social group (Lever, 1983; Gilbert, 1995). Surprisingly, despitethe fact that sporting events are mass leisure activities and are active parts of consumers' lives,sources of high levels of amusement, sensations and emotions (Holbrook, Hirschman,1982;Unger, Kernan, 1983), they have received very little attention in marketing research exceptrecently with researchers like Holt (1995) for example.

In fact, most of the studies made in the entertainment consumption field generally focused on"high culture" topics as Opera, theater or classical music. Therefore, studies about the hedonistconsumption (Hirschman, Holbrook, 1982; Holbrook, 1986, 1981, 1982) or about subjectivedimensions of leisure activities (Iso-Ahola, 1979; Unger, Kernan, 1983) were the first steps to abetter understanding of the phenomenon. The particular nature of the "sporting event" productsustains the fact it has to be studied in a specific way to get a better vision of all its aspects. Thesport and the sporting event more specifically are complex phenomenons and can not be seenonly under the restrictive definition of leisure. Indeed, the consumption experience of a sportingevent is an elaborate process which involves many values of the consumer (Holt, 1995).

The objective of this research is to present the dimensions which determine the attitudes and thebehaviors of sporting events consumers. These dimensions will be used as the basis to define theorientation of individuals towards sporting events. In concrete terms, the main objective lies indeveloping a valid scale presenting the dimensions of the orientation of individuals towardssporting events. This article briefly analyzes the concept of the sporting event and the way itinteracts with consumers as well as the orientation concept. A conceptual model is presented.Second, the propositions of the study are presented. Third, the methodology of the detailedresearch is presented in detail and finally, the results are discussed.

1. The conceptual framework

A better understanding of the phenomenon assumes the analysis is done based on a conceptualframework involving the orientation and leisure concepts. The "leisure" side of the sport will bepresented with the study of the potential difference between élite sport and mass sport. Thesporting event will be studied through its social and relational sides. Finally, the sensations andknowledge seeking made by the consumers will be included.

4

The consumers orientation :

In consumer behavior, the orientation is defined as the specific inclination of an individual toadopt a foreseeable behavior during an act of consumption. The orientation finds its roots in theneeds and the values of an individual. Parsons and Shils (1967) underlined the importance ofmotivation and values concepts as characteristics of the orientation in general. The "motivation"aspect of the orientation refers to the ability of the activity or the product to satisfy the particularneeds of the individual. Then, we would say that he is oriented to this product or activity. The"value" aspect of the orientation refers to the values of the invidual that, in a particular context,orient his behavior. In the sporting event case, the "motivation" refers to the ability of the eventto fullfill the specific needs of the individual such as attending unforgettable moments of thesport history or such as being able to show his happiness in particular occasions. The "value"usually appears when the consumer has to make a choice between several leisure activities andwhen he ends up choosing a particular sporting event. The orientation of an individual for asporting event can rely on many dimensions or modalities which form the content of theorientation, its qualitative aspect (Parsons, Shils, 1967; Hirschman, 1984, Laaksonen, 1987).

Therefore, a spectator may have an orientation for the sporting event because of the excitment orthe pleasure it brings and an other may have the same orientation but for the relational andeducational aspects gained from the experience.Nevertheless, the dimensions of the orientation should include all these modalities to allow theconsumer to make a choice in harmony with his needs and values. Once these dimensions areestablished, it is important to know the strength of the inclination for the sporting event for anindividual. The implication or the level of significance of particular aspects of the sporting eventsuch as excitment, pleasure... are good indicators of the total intensity of the orientation for thesporting event for a consumer. After these reports about the orientation, it is now important todefine the sporting leisures according their kind.

For a better vision of the many aspects of a phenomenon which represents 60 % of expendituresfor companies specialised in marketing events, the sport will be first studied as a communicationmedium between individuals, allowing to develop values in a social group. Then, theconsumption of sporting events will be considered as an emotional and sensorial experience.Finally, the sport will be analysed as a means to develop the knowledge of the consumers.

The sporting event : an elite leisure or a mass leisure.

In our society, the leisure consists in consuming free time, in making it profitable in order tofulfill our needs such as escape, rest or relaxation. Nevertheless, even if the leisure is effectivelya means to spend time, it also has a distinctive value for the consumer (Veblen, 1978). Indeed,the leisure, still unevenly distributed among the consumers, stands as a selection and a culturaldistinction factor (Baudrillard, 1970). By choosing his own leisure activity, which in fact is another way to spend his free time, the individual asserts himself and produces a social value. Thissocial value results in a categorization of the individual. He, then, faces a structure ofdifferentiation from a group or he joins an already existing group. Then, sport and the sportingevent more specifically can be considered as leisure-witnesses of the social changes. The world

5

of sport appears to any newcomer as a wide range of choices between tradition levels, rules,values and symbols which gain a social meaning from the system they constitute. The consumerclassifies himself in a group represented by his fellow individuals (Settle, Aldeck, Belch, 1979).Those groups with hermetic frontiers are the combined result of the economic capital availableand, more important, of the group entry "fees" such as family tradition or early learning(Bourdieu, 1979). Therefore, it is important to present the possible distinctions existing betweenthe different kind of sports. These differences can come from the variations of the perceivedprofits that a particular sport is supposed to bring. They also can be the result of the variations interms of perceived "costs" (cultural, economic...) that a specific sport might lead to. In fact, thepractice of a specific sport or the fact of attending a specific sporting event depend on the profitsor the extrinsic or intrinsic costs for the individual, in the limits fixed by economic, free time andcultural capitals available (Bourdieu, 1979). The cultural capital (tradition and early learning)predispose the consumer to express all his differences. For the elite sports, the activity (practiceor attendance) offers a forum for high social exchanges, whereas, mass sports have usually a"vulgar" connotation. To escape from "vulgar" entertainment, privileged persons have to preserveand affirm their distinction by avoiding the popular crowd and then show their disagreements.

In this study about the sport area, it is essential to clarify this distinction because the results willbe affected by the sporting events selected and the way it is perceived by the fans (mass or elite).The values,then, attributed to different sporting events by the consumers are different accordingto the social group to which the consumer belongs (Tinsley, Barrett, Kass, 1977).

The sporting event : a show.

Nowadays, the sport reaches the status of a complete show because it offers many new featuresfor the consumer such as escape, dream, feeling of belonging. At this point, marketing has acrucial role.It orients the consumer to the satisfaction of those new attributes by creating an artificial andparallel world where the image becomes the substance, the essence of everything (Baudrillard,1987; Postman, 1985). Thus, the sporting event and its new attributes appear as the new realityfor the consumer. This reality is defined as a hyperreality which underlines the will of consumersto build their own "reality", to simulate it (Baudrillard, 1983, 1987; Eco, 1986; Firat, Venkatesh,1993). Marketing may be perceived as a support to the fragmentation of the sporting event byextracting it from its original context to clarify its different functions in a specific context (Eco,1986; Kellner, 1989; Sherry, 1984). The sporting event might be perceived as the gathering offragmented moments, as a range of products (violence, beautiful plays...), as a multitude ofevents bringing various and often contradictory feelings to the consumer. Therefore, it isnecessary to understand the sporting event as a show in which the consumption becomes asymbolic act. Indeed, the objective differences between the consumer and the object vanish, thesubject becomes also a product and he has to conform to the rules of the "market". In fact, theobject and the subject are not longer separate (Hassan, 1987; Jameson, 1983). By choosing asporting event, the consumer can feel himself as a product which attributes, chosen by the fan,can please to his fellow fans or not. He is a part of the product, of the consumption object. Then,the sporting event is a show because of the significance transmitted by the objects, by theemphasis on the instrumental functions of the product (Mc Cracken, 1988), giving birth to anaesthetisation of the consumption (Cova, Svanfeldt, 1993, 1995).

6

The sporting event : an hedonic activity.

The search of aesthetisation, of feelings, of emotions through sporting events present it as anhedonic activity, underlining the play and the experiential aspects of the consumption act(Hirschman, Holbrook, 1982). Usually, in the case of cultural products and services or leisureactivities, the experience of consumption is the final goal. The main benefits expected during theconsumption are the pleasure, the feelings and the emotions. The consumption of those goodsand services assume an amount of time available and refer to a lifestyle. The alternativesconsidered by the consumers are other time-consuming activities (Cooper-Martin, 1991, 1992).Then, it is possible to assume that sporting events are high implication activities for theconsumers. Laurent and Kapferrer (1985) define the importance of implication profiles for anindividual. The personal interest for the category of product, the sign value attributed by theconsumer to the product and the hedonic value of the product are the aspects which cancontribute to create a durable implication for the consumer. This ability of the product to givepleasure is one of the main determinants of the hedonic value of a sporting event.This aspect allows the individual to escape from the reality during the consumption (Berlyne1969, Stephenson, 1967). This hedonic approach is essential to understand the consumption ofsporting events.The sporting event : an activity allowing the knowledge development.

During the consumption of a sporting event, the spectator can also try to satisfy some needs ofknowledge and understanding of the sporting events world. The "cognitive" individuals arepeople who appreciate to discover how the things are working and who wants to growintellectually (Venkatraman, MacInnis, 1984; Venkatraman, Price, 1990; Nyeck, 1994). During aleisure activity as well as the attendance at a sporting event, the cognitive mastery appears as oneof the most important aspects for a consumer to fully appreciate his occupation (Hawes, 1978).The evaluation of a sporting event by a consumer involves comparisons between what he seesand a full set of rules, histories, memories, records. Furthermore, the consumer has to improvehis knowledge about the event in order to socialise, exchange with the others and finally to beable to appreciate the event using the same reference framework as the other spectators (Holt,1995). This cognitive search can be underlined by the need of the spectator to identify himselfwith the actor. This identification usually allows the spectator to consider himself as an expertand then to affirm his mastery (Mullin, Hardy, Sutton, 1993).The sporting event is an experience of consumption which has the ability to satisfy specific needsof the consumer when he is in a particular context. The affective, cognitive and social integration(socialisation) dimensions constitute the different modalities the sporting event can offer.Therefore, it is logical to propose the following conceptual framework to define the orientationof an individual for a sporting event.

Figure 1. Conceptual framework : Dimensions of the Orientation for the sporting event

7

The sportingevent

consumer...

... sensations,emotions through his

experience.

...searches for... ...searches for....

...searches for.........a better understandingand knowledge of his

expérience.

...a way to fullfil hissocialisation needsby his expérience.

2. Propositions of the research and methodology

The conceptual framework leads to the idea that leisures tend generally to satisfy the search forsensations and emotions, the need for knowledge and the search for social integration (Nyeck andal, 1996; Bergadaa, Nyeck, 1995; Holt, 1995; Kantanen, 1993; Lacher, Mizerski, 1993;Hirschman, 1984). These cognitive, sensorial and socialisation needs are essential dimensions forthe understanding of the sporting event consumer. The main objective of this research can bedivided in two parts. The first one consists in developing a new scale to measure the consumerorientation for a sporting event (OSE) while keeping in mind the second one : verify thereliability and the validity of this instrument. The dimensionality of the measurement scale willbe tested. The validity (convergent, discriminant and predictive) will be tested by including inour measures some existing scales or part of existing scales which are supposed to measurecocepts close or opposite to those of the OSE scale (Synder, Fromkin, 1977).

P1 : The existence of a three dimensions model acting as the determinant of the orientation ofconsumers for sporting events is postulated. These three dimensions constitute the construct.P2 : The measurement scale of the consumer orientation for sporting events (OSE) presentssatisfying psychometric proprieties.P3 : The research will evaluate the validity of the OSE scale. To establish the construct validityof the OSE, the convergent and discriminant validities are evaluated (Campbell, 1960; Peter,1981; Fornell, Larcker, 1981).P4 : The OSE scale presents a nomological validity (Nunnaly, 1978; Peter, Churchill, 1986;Steenkamp, Van Trijp, 1991). We can establish the network of relationships of the consumerorientation for sporting events with other constructs. Then, the measures of constructs such as theneed for cognition (NFC, Petty, Caccioppo, 1982), the inclination to search new sensations andchanges (CSI, Steenkamp, Baumgartner, 1992, 1995), the emotional search (PAD, Mehrabian,

8

Russell, 1974) and the external orientation of social preference (I-O, Kassarjian, 1962) should besignificantly related to the OSE measures.P4 (a) : The search for sensations dimension of the OSE scale is positively related to theemotional dimension measured by the PAD scale (Mehrabian, Russell, 1974) and is not stronglyor not at all related with the cognitive need measured by the NFC scale (Petty, Caccioppo, 1982).P4 (b) : The cognitive dimension of the OSE scale is positively related to the cognitive needmeasured by the NFC scale (Petty, Caccioppo, 1982) and is not strongly or not at all related withthe emotional dimension measured by the PAD scale (Mehrabian, Russell, 1974).P4 (c) : The search for sensations dimension is positively related to the consumer tendancy tosearch for new sensations and changes as measured by the CSI scale (Steenkamp, Baumgartner,1992, 1995).P4 (d) : The socialisation dimension of the OSE scale is positively related to the measure of thethe external orientation of social preference measured by the I-O scale (Kassarjian, 1962).P5 : In the case of the predictive validity, the consumer orientation for sporting events issignificantly and positively correlated with the fact of buying goods related with the sportingevents as well as with the amount of money dedicated to the sporting events.

Kantanen (1983) showed the intensity (frequency) of the attendance of a spectator to a show hasan important influence on his global orientation. The frequent user becomes more demanding onspecific aspects of the show. It is logical to assume that we can find this relationship in thecontext of the sporting event.

Hirschman (1984) states that consumers with emotional tendancy present a stronger inclination toconsume other hedonic products, in fact products to experiment with pleasure. On the contrary,the cognitive consumers present a tendancy to have interest for more utilitarian products. Thispoint is expressed by Holt (1995) who explins that the baseball consumer enjoys products inwhich he can find sensations, experiences. However, he adds that if the baseball consumer alsobuys products with his team signs such as hats or tee-shirts, then he buys an image more than anutilitarian aspect.

In this study, the confirmatory process used (Bollen, 1989; Gerbing, Anderson, 1988) ispresented in the figure 2.

Figure 2 Methodological process

9

Study

1

Specifydomain of

theconstruct

Generatesample of

items

Purifymeasure

AssessDimensionality

AssessValidity

Study

2

PrincipalComponents

Analysis

(PCA)

ConfirmatoryFactor

Analysis

(CFA)

Study 1 : Items generation and pretest

45 items have been generated from semi-directive interviews and discussions conducted amongpeople involved in the sporting events field (fans, promoters of sporting events, marketing andsales people) and from the existing litterature (Bearden, Netemeyer, Mobley, 1993; Unger,Kernan, 1983; Holbrook, 1982, 1986, 1990; Hawes, 1978, Holt, 1995...). The items presentedunder a likert 5-points scale were given to a six-person judgmental panel of experts. They had toindicate what dimension was, according to them, measured by the items they had just received.An item was retained, following an a-priori rule, if the six persons identified it as measuring thesame dimension. The six dimensions were: the learning process, the escape, the pleasure, theexpertise, the arousal and the socialisation. Each dimension had between 6 and 8 pre-supposeditems.

The pretest has been conducted in a sample of graduate students in a large central canadianuniversity (n=88). The craze for university sporting events is high and the students have easyaccess (prices and events) to sporting events on the campus. The questionnaires wereadministered at the sports activities building by the researchers to voluntary students.

Purification of the instrument

A principal components analysis was used to determine the structure of the instrument. Indeed,the exploratory analysis can be used to confirm whether the number of dimensionsconceptualized can be verified empirically (Churchill, 1979). This optimal factorial structure wasdone by iterations : the elimination of "garbage" items was mainly made by examining the resultsof the exploratory factor analysis and the reliability analysis of the instrument. The determinationof the number of factors underlying the construct was made by using the eigenvalue criterionwith a limit of 1. The "Scree test" procedure was used to confirm the number of factors (Cattell,1966). An oblic rotation was performed because the dimensions of the consumer orientation forsporting events are conceptually not independent (Cattell, 1978), the content of the factors wasevaluated by checking the items loadings on those factors. Scale reduction was also performed byimposing stringent requirements on the loadings. This requirement on the relationship betweenitems and the underlying factor that they are intended to measure was established at 0.5.Therefore, items should have high loadings on the factor they are supposed to belong to and low

10

ones on the other facors (Pedhazur, Schmelkin, 1991). Finally, the reliability of the instrumentwas evaluated by alpha of Cronbach coefficient (Nunnally, 1978).

Study 2 : Scale validation

The use of an exploratory factor analysis in the early stage of the research led to a purified 21items OSE scale. This scale will be used for the factor analysis in the confirmatory stage of theresearch (Gerbing, Anderson, 1988). The questionnaire includes questions on behaviors andpractices regarding the sporting event activities, on demographics categories of the respondentsas well as on individual characteristics scales (usually short form) to estimate the validity of theCOSE scale. We used 11 items of the need for cognition scale (NFC, Petty, Caccioppo, 1982;Petty, Caccioppo, Kao. 1984), 7 items (short form) of the inclination to search new sensationsand changes CSI scale (Garlington, Shimota, 1964; Steenkamp, Baumgartner, 1992, 1995), theemotional response (PAD, Mehrabian, Russell, 1974, Haevlena, Holbrook, 1986) and 10 items ofthe social preference I-O scale (the external dimension)(Kassarjian, 1962). For the pupose of ourstudy, all the scales used have been adapted to the 5-point likert model to fit with the COSE scaleone. The items were translated and their meanings were checked with French and Englishspeaking translators using a backtranslation procedure.

The sample

The questionnaires were administered to a sample of undergraduate students in the businessschool of the university. 240 questionnaires were completed and 234 were finally used for thestudy. The samples size is larger than 200 (Boomsma, 1987) and the 5 to 10 ratio of respondentsto items of the larger construct is considered adequate (Peterson, 1982). Our sample respondentstended to be male (55,6%). The average age of sample respondents is 24,2 years with a standarddeviation of 5,4 years. An exploratory factor analysis was performed on this new sample to checkthe adequacy of the results with those of the pretest. Iterations were made again to purify theinstrument. Scale reduction conducted to a final instrument of 15 items used to perform theconfirmatory analysis.

The data analysis

The confirmatory factor analysis was performed with the EQS software (Bentler, 1989). Theestimation method used was maximum of likelihood (ML). Nevertheless, in case of non-normality of the data, EQS allows the use of ROBUST statistics to estimate parameters or fitindicators (Wothke, 1993) This option gives the estimation of Satorra and Bentler (1988) χì

statistic and the ROBUST statistics errors which has been recalculated and corrected to assumethe non normality of large samples. Therefore, the quality of the model is estimated by analysingthese indicators.

3. Results

Study 1 : The pretest

11

Table 1 shows a summary of the first study results. The successive purifications of the principalcomponents analysis conducted to the presence of 4 factors explaining 88,5% of the phenomenonvariance. This number was determined using the Scree test and the eigenvalue criterion of 1(Kaiser). The alpha coefficients are all above the 0,8 level and some of them are above the 0,9point.

Table 1. Exploratory factor analysis

Itemn°

12 23 25 34 06 18 16 22 10 5 29 8 24 14CumulateVariance

explained %

Meaning alpha

Factor1

.91 .91 .90 .89 .84 .80 49.9 Cognitivesearch

.96

Factor2

.92 .91 .90 .89 66.1 Socialorientation

.96

Factor3

.92 .92 81.2 relaxationsearch

.82

Factor4

.92 .90 88.5 Sensationsearch

.92

The most evident dimension is the cognitive one (factor 1 with 49,9% of the explained variance).This dimension includes items from both the learning and expertise dimensions presented in theconceptual framework. A focus on the other items regroupings shows the presence of twodimensions : the first one can be defined as the search for relaxation and the other one asmeasuring the search for sensations or emotions. The social orientation dimension is also reallystrong and its presence is undeniable. The results are not exactly what was expected after theconceptual presentation. They underline the complex aspect of sensorial or affective dimensionand show the possibility of more than a single dimension to measure those aspects. This option isevaluated with the next and last pre-confimatory analysis. The instrument includes 14 items keptafter purification and 7 new ones. Indeed, for the next stage of the study, new items were addedto preserve the balance between factors regarding the number of items in each dimension.

Study 2 : Scale Validity

Pre-confirmatory factor analysis

The scale was tested on the final sample to validate the factorial structure determined by thepretest and to check the quality of items added. This analysis should lead to a better choice ofitems to retain in the model to be subsequently evaluated during the confirmatory analysis. Theresults of this analysis in principal components after an oblimin rotation are presented in Table 2.Items were renamed according the dimension that they are supposed to measure. Loadings under|.4| were not entered. Three factors explain 75 % of the variance. This number of dimensions isconfirmed by the Scree-test and the Kaiser criterion. This first test tends to confirm the presenceof three dimensions in the measure of a consumer orientation for a sporting event : a socialdimension, a cognitive dimension and a search for sensation dimension. The alpha coefficienteasily meets the requirements (Nunnally, 1978). However, before launching the confirmatory

12

analysis process, it is important to keep only the "cleanest" items in the scale. Therefore, becauseof their relatively high loadings on more than one factor, some of the items are withdrawn. Thefinal instrument tested in confirmatory analysis has 15 items (5 in each dimension)2.

Table 2. Pre-confirmatory factor analysis

Itemn°

SE1

SE2

SE3

SE4

SE5

SO1

SO2

SO3

SO4

SO5

CO1

CO2

CO3

CO4

CO5

CumulateVarianceexplained

%

Meaning alpha

Factor1

.87 .81 .80 .76 .72 45.6 Sensationssearch

.93

Factor2

.90 .88 .86 .86 .84 64.5 Socialorientation

.95

Factor3

.87 .84 .84 .84 .79 75.0 Cognitivesearch

.94

The confirmatory factor analysis

In univariate analysis of the data, no Kurtosis coefficient is above |1.5| (Byrne, 1994).Consequently, there is no evidence of univariate non normality of the data. Nevertheless, someitems (SO1, SO3, CO4, CO5, SE5) show significantly high kurtosis values to correspond to aperfect normality. However, it is possible to infer that the data present only a slight multivariatenon-normality which will not bear a strong influence on Maximum Likelihood estimations (ML).Indeed, Bentler (1992) suggests that a Mardia's estimate above 10 testifies to light multivariatenon-normality of the data. Finally, the ROBUST statistics, which assume non normality of data, donot bring significant improvement to estimation of the model.Dimensionality of the instrument

The three factor model was tested and the main indicators of the fit qualities of the model arepresented in table 4. The estimates are given after each modification at the original model henceproviding the opportunity to better judge the overall fit of the eventual model. Each modificationis also justified. Next, the identification of the extreme cases which contribute most heavily to theMardia's estimate (EQS Bentler, 1989) and their withdrawal allowed refinement of the multivariatenormality of the data. Two other cases were eliminated because of their non contribution to theestimates homogeneity (models 1A and 1B).The quality of the overall fit of the model was evaluated according to a set of criterions. The χì

value, the χì/df (degree of freedom) value, the NFI and CFI3 indicators demonstrate fairly good fit 2 For example, Items were :For Socialisation : SO2 Generally, I share my opinions about sporting events with other persons.

SO5 I love talking about sports with people around me.For Search for sensations SE2 Attending sporting events is a good way to relax.

SE3 I am always excited about the fact of attending sporting events.For Cognitive search CO3 I consider myself as an experts of the sports I am interested in.

CO4 Regarding the sporting events field, we can say that I am ignorant.3 NFI = (χì0-χìk) / χì0CFI = | ((χì0 -df0) - (χìk -dfk)) / (χì0 -df0) |

13

for the initial model. The χì to degrees of freedom ratios are each under 5, the first sign of anadequate model fit (Wheaton et al 1977). The 1B model presents a ratio close to 3, whichconstitutes a good value to judge the fit quality of a confirmatory model (Carmines et Mc Iver1981). The values of the NFI and CFI indicators (Bentler et Bonnett 1980; Bentler 1990) wereabove 0.90. Therefore, given the size of the sample, they indicate a fairly good fit (Bentler, 1992).Finally, the number of iterations of the EQS program to converge to the solution is a good sign ofthe quality of fit of the model because it maintains a value under 10 (Byrne, 1994). The residualanalysis also leads to the conclusion of good fit of the model. Indeed, the 0,13 average value forthe off diagonal standardised residuals indicator is good even if not ideal and only 60% of theresiduals are distributed between -1 and +1. This study of the highest normalised residualssuggests a problem related to the construction of the SE2 item.

Table 4. Results of the three factors model.

Tests Model 1 M 1A Final model 1B Comp.Model C

Normalized MultivariateKurtosis estimate (Mardia's)

14.85 13.44 12.64 12.64

Average off-diagonal value ofstandardised residuals

0.131 0.130 0.115 0.115

standardised residualsdistribution

Fairly Good Fairly Good Fairly Good Fairly Good

χìì 377.5 320.7 285.4 χì SB 258.6Degrees of Freedom (Df) 88 86 85 85

χì/df 4.29 3.72 3.35 3.04NFI 0.89 0.91 0.92 0.92CFI 0.917 0.933 0.943 CFI* = .95

Number of iterations 7 7 6 6

Respecification/precedentmodel

Originalmodel

1 case withdrawn, free 2covariances between errors (SO1and SO2 and also SE1 and SE2).

1 case withdrawn,free item SE2 on

factors SO and SE.

ROBUSTestimationmethod.

After respecification4, the final 1B model presents a better quality of fit. Also, the use ofROBUST statistics (comparison model C) allows a slight improvement of the model comparedto the 1B model. Therefore, the light violation of the multivariate normality is not prejudicial tothe final fit of the 1B model. Consequently, the three dimensions structure of the OSE isestablished and proposition 1 is validated. The final model is represented in figure 3.

Figure 3. Final 3 factors model.

4 During resprcifications of 1A and 1B models, 2 covariances between error terms and SE2 item on socialisationfactor have been freed. Acccording to Lagrange Multiplier (LM procedure), it is possible to identify parameterswhich could most contribute to a χì reduction (Byrne, 1994). Conceptually, the error terms measure the sameconcepts and a redundance phenomenon can appear, intoducing a strong correlation between these error terms(Byrne, 1994). Also, SE2 item can be related to the socialisation factor, it can be interpretated as a measure of thesearch for sensations that we want to share, that we want to use to socialise.

14

Socialisation CognitiveSearch

SensationsSearch

SO1 SO2 SO3 SO4 SO5 CO1 CO2 CO3 CO4 CO5 SE1 SE2 SE3 SE4 SE5

E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14 E15

Psychometric properties of the OSE scale

The OSE scale presents good psychometric properties. The alpha coefficients for the "search forsensations", "cognitive search" and "socialisation" dimensions are respectively 0.89, 0.94 et 0.96,and are each above the 0,70 criterion (Nunnally 1978). The factor-item correlation for a samedimension are all above 0,64. Finally, the t-values associated with the loadings estimates are allabove 1,96, hence supporting the internal consistency of the scale (Gerbing et Anderson, 1988).Consequently, proposition 2 is validated.

Construct Validity of the scale.

Table 5 shows the evaluation of the construct validity for the OSE scale. Indeed, the averagevariances shared between the constructs and their measures (AVE) are each above 50%, therebysupporting the convergent validity of the scale (Fornell et Larcker 1981; Valette-Florence, 1988;Netemeyer, Durvasula, Lichtenstein, 1991). Also, the variance shared between the constructs isalways lower than the AVE. The discriminant validity is therefore supported (Fornell et Lacker1981; Valette-Florence 1988). The construct validity can be assumed and proposition 3 validated.

15

Table 5. Construct Validity of the OSE.

Nomological Validity

The nomological validity of the Orientation for Sporting Events is estimated by studying thecorrelation of its measures with measures from personality traits and behaviors as discussed in thepropositions section. The correlations are presented in table 6. By considering the shape and thestrength of the relationships between two criteria or predictors, the results lead to the validation ofan existing network of relationships. The nomological validity of the Orientation for SportingEvents scale is then established (Cronbach et Meehl, 1955). From a compound index for eachdimension of the OSE scale, strong and significant positive relationships are established betweenthe search for sensations dimension of the OSE and the emotional dimension from the PAD scale,between the cognitive search of the OSE and the Need for Cognition of Petty and Cacciopposcale, and between the socialisation dimension of the OSE and the external dimension of the I-OKassrjian's scale. Finally, the search for sensations dimension is positively and significantly relatedto the Change Seeking Index which is not the case for the two other dimensions. Indeed, thesocialisation dimension has no theoritical relationship with the CSI index. However, some weakpositive relationships can be established between the search for sensations dimension of the OSEscale and the Need for Cognition, and between the cognitive search of the OSE and the emotionaldimension from the PAD scale. These relationships can be explained by the fact that they sharesome variance associated to the measurement method5. The correlations are significant and theexpected direction, therefore the propositions 4 dealing with the Criterion Validity. The threedimensions are also related to the purchase of sports products, the predictive validity of the OSEcan then also be assumed (proposition 5)

5 All the instruments have been adapted to the Likert scales format.

Convergent Validity Average Loading AVESocialisation 0.80 0.64

Recherche Cognitive 0.87 0.75Recherche de sensations 0.77 0.59

Moyenne 0.81 0.66

Discriminant Validity Variance shared betweenconstructs

Socialisation- cognitivesearch

0.36

search for sensations-Cognitive search

0

Search for sensations-socialisation

0.10

16

Table 6. Nomological Validity of the OSE scale.

* p<0.05, ** p<0.01.

4. Discussion and Conclusion

Sport in general and sport products have been ignored or have received very little consideration inthe past. This research leads to the development of a valid scale to measure the Orientation forSporting Events. In this respect, the subjective dimensions which determine the attitudes of thesporting events consumers are established (Cognitive search, search for sensations andsocialisation). The results give support to qualitative research in the area (Holt, 1995) and extendresults found in the field of leisure and in the consumption of hedonic products (Kantanen, 1993;Lacher, Mizerski, 1993; Unger , Kernan, 1983). Indeed, Holt (1995) presents four types of sportconsumption experiences. Some of these types are confirmed in our study. Also, Unger andKernan (1983) proposed an instrument oriented to the dimensions of leisure. The six dimensionsfound in the leisure area are not all found in the sporting event field which could be explained bythe specific nature of the sporting event compared to general leisure. Nevertheless, the Unger andKernan (1983) instrument served as a good basis once adapted to the sporting event field.

However, these results can not be considered as conclusive. This research should be extended tosamples other than students. The student samples are not considered representative of thepopulation Netemeyer et al, 1991). However, in the sporting events field, students represent animportant segment of the consumers. To improve the validation of the OSE instrument,forthcoming research should focus on specific sports. Also, it may be advantageous to performcross-cultural studies.

This research highlights the dimensions which determine the consumers orientation regarding the"Sporting Event" product. This product stands for more than 60% of expenditures of promotionand events marketing firms. Consequently, this study may identify segments or "tribes" of potentialconsumers differently than the classical demographic variables such as age or Salary. Thissegmentation of the potential sporting events potential customers may allow better allocation ofcompanies resources allocated to promotion campaigns or choices of support. Thusly, theirchoices would be more aptly adapted to their target (Pope et Vogues 1997; Lardinoit 1997).

Alpha(Cronbach)

Socialisation CognitiveSearch

SensationsSearch

Criterion ValidityEmotional Dimension of PAD

scale0.86 0.41 ** 0.21 ** 0.78 **

Need for Cognition (NFC) 0.93 0.13 * 0.65 ** 0.19 **Dimension OUT of the I-O scale 0.90 0.77 ** 0.29 ** ns

Change Seeking Index (CSI) 0.85 ns ns 0.30 **Predictive Validity

Purchase of sports products 0.77 0.69 ** 0.62 ** 0.27 **Budget for the Sporting Events 0.52 ** 0.39 ** ns

17

References

Baudrillard J. (1987a), “Forget Foucault”, New York, NY : semiotext.Baudrillard J. (1987b), “America”, translated by C.Turner, London : Verso.Baudrillard J. (1983), “The ecstasy of communication”, The anti-aesthetic : Essays onpostmodern culture, editor H. Foster, Port Townsend, WA : Bay Press, pp 126-134.Baudrillard J. (1970), “La société de consommation”, éditions Denoël, pp 242-252.Bearden W, Netemeyer R et Mobley M (1993), “Handbook of Marketing scales,” Newbury Park,CA : Sage.Bentler P. (1992), “EQS : Structural equations program manual”, Los Angeles : BMDPStatistical Software.Bentler P (1990), “Comparative fit indexes in structural models”, Psychological Bulletin, 107, pp238-246.Bentler P (1989), “EQS : Structural equations program manual”, Los Angeles : BMDP StatisticalSoftware.Bentler P(1984), “Structural equation models in longitudinal research”,Handbook of longitudinalresearch, New York, Praeger, pp 88-105.Bentler P, Bonett D (1980), “Significance test and goodness of fit in the analysis of covariancestructures”, Psychological Bulletin, 88, pp 588-606.Berlyne D. F. (1969), “Laughter, Humor and Play”, The Handbook of Social Psychology, Vol.3, editor Gardner Lindzey and Elliot Aronson, Reading, MAA : Addison-Wesley, pp 795-852.Bollen K. (1989), “Structural equations with Latent variables”, New York, John Wiley.Boomsma A. (1987), “The robustness of maximum likelihood estimation in structural equationmodels”, Structural equation modelling by example : Applications in educational, sociologicaland behavioral research, New York, Cambridge University Press, pp 160-188.Bourdieu P. (1979), “La distinction : critique sociale du jugement”, éditions de minuit, Paris.Byrne B. (1994), “Structural Equation Modeling with EQS and EQS/Windows.”, ThousandOaks, CA, Sage Publications.Caccioppo J., Petty R. (1982), “The need for cognition”, Journal of Personality and SocialPsychology, 42 (1) pp 116-131.Carmines E., Mc Iver J., (1981), “Analyzing models with unobserved variables : analysis ofcovariance structures”, Social measurement : Current issues. Newbury Park, CA, Sage, pp 65-115.Cattell R (1978),“The scientific use of factor analysis in behavioral and life sciences,” NewYork, Plenum.Cattell R (1966),“The Scree test for the number of factors,” Sociological methods and research,1, pp 245-276.Churchill G (1979), “A paradigm for constructing better measures of marketing concept,”Journal of Marketing Research, 16(1), pp 64-73.Cooper-Martin, E. (1992), Consumers and Movies: Information Sources for ExperiencialProducts, Advances in Consumer Research, John F. Sherry and Brian Sternthal, eds., vol 19,Provo UT: Association for Consumer Research. 756-761.Cooper-Martin, E. (1991), Consumers and Movies: Some Findings on Experiential Products,Advances in Consumer Research, Rebecca Holman and Michael R. Solomon, eds., vol 18,Provo UT: Association for Consumer Research. 372-378.

18

Cova B, Svanfeldt C. (1995), “Le marketing à l’épreuve du postmoderne”, Actes de l’AFM, pp414-442.Cova B, Svanfeldt C. (1993), “Societal innovations and the postmodern aestheticization ofeveryday life”, International Journal of Research in Marketing, vol 10, pp297-310.Cronbach L, Meehl P (1955), “Construct validity in psychological tests”, Psychological Bulletin,52 (4), pp 281-302.Eco U. (1986), “Travels in Hyperreality”, translated by W Weaver, San Diego, CA : HarcourtBrace Jocanovitch.Firat A. F., Venkatesh A. (1993), “Postmodernity : the age of marketing”, International Journalof Research in Marketing, Vol 10, North Holland, pp 227-249.Fornell C, Larcker D. (1981), “Evaluating Structural equation models with unobservablevariables and measurement error,” Journal of Marketing Research, 18(1), pp 39-50.Gerbing, D. W, Anderson J. C. (1988), An Updated Paradigm for Scale DevelopmentIncorporating Unidimensionality and Its Assessment, Journal of Marketing Research, vol 25(may), 186-192.Gilbert N. (1995), “Kickoff time for Soccer”, Financial World, 14th february, pp 79-86.Haevlena W, Holbrook M B,(1986), “The varieties of consumption Experience : Comparing twotypologies of Emotion in Consumer behavior”, Journal of Consumer Research, 13(12), pp 394-404.Hassan I (1987), “The postmodern turn : Essays in postmodern Theory and Culture”.Hawes Douglas K (1978), “Satisfactions Derived From Leisure Time Pursuits : An ExploratoryNationwide Survey”, Journal of Leisure Research, 10(4), pp 247-264.Hirschman, E. C., (1986), Humanistic Inquiry in Marketing Research, Philosophy, Method andCriteria, Journal of Marketing Research, 23, August, pp. 237-249.Hirschman E. C. (1984), “Experience seeking : a subjectivist perspective of consumption”,Journal of Business Research, 12:1, pp 115-136.Hirschman, E. C., M. B. Holbrook (1982), Hedonic Consumption: Emerging Concepts,Methods and Propositions, Journal of Marketing, 46 (summer), 92-101.Holbrook M B., (1986), “Emotion in the Consumption Experience : Toward a New Model ofthe Human Consumer”, The Role of Affect in Consumer Behaviour : Emerging Theories andApplications, editors Peterson R et al., Lexington, MA: Heath, pp 17-52.Holbrook M B.,(1982), “Hedonic Consumption : Emerging Concepts, Methods andPropositions”, Journal of Marketing, Vol. 46, pp 92-100.Holbrook M B, Hirschman E. C. (1982), “The Experientrial Aspects of Consumption:Consumer Fantasies, Feelings, Fun”, Journal of Consumer Research, 9(2), pp 132-140.Holbrook M B.,(1981), “Symbolic Consumer Behaviour”, Ann Arbor, MI : Association forConsumer Research.Holt D. (1995), “How consumers consume : a typology of consumption practices”, Journal ofConsumer Research, 22(2), pp 1-21.Jameson F (1983), “postmodernism and consumer society”, The anti-aesthetic : Essays onpostmodern culture, editor H Foster, Port Townsend, WA : Bay Press, pp 111-125.Kantanen T (1993), “Analysis of the contents of theatre visiting orientation”, Thesis inMarketing, Faculty of business administration, University of Vaasa.Kassarjian W (1962), “A study of Riesman’s theory of social character”, Sociometry, 26 (9),pp 213-230.Kellner D. (1989), “Postmodernism, Jameson, critique”, Washington, DC : Maisonneuve Press.

19

Kindall J. (1983), “Sports Illustrated Baseball”, Library of Congress cataloging in publishingdata, p 14.Iso-Ahola (1979), “Basic Dimensions of Definitons of Leisure”, Journal of Leisure Research,11(1), 28-39.Laaksonen, M. (1981), “Shopping orientation. An attitude structure approach.,” Proceedings ofthe university of Vaasa, Research papers 77, Business administration 21.Lacher, K, Mizerski Richard (1994), “An Exploratory Study of the Responses andRelationships Involved in the Evaluation and in the Intention to Purchase New Rock Music,”Journal of Consumer Research, 21(9), pp 366-378.Lardinoit T. (1997), "Interaction des parrainages terrain et T.V : mythe ou réalité? Le cas de laréaction attitudinale", in Actes du Congrès annuel de la section Marketing de l’Association dessciences administratives du Canada- ASAC, Vol 18-3, St. John’s NF, juin 1997, 134-149.Laurent G, Kapferrer J-N. (1985), “Measuring Consumer involvement Profiles”, Journal ofMarketing Research, 22 (February), pp 41-53.Lever J. (1983), “Soccer Madness”, Chicago : University of Chicago Press. p 64.McCraken, G. (1988), Culture and Consumption: New Approaches to the Symbolic Character ofConsumer Goods and Activities, Bloomington: Indiana University Press.Mehrabian A., Russel J. (1974), “An approach to Environemental psychology”, Cambridge,MA, MIT Press.Mullin Bernard, Hardy Stephen, Sutton William A (1993), “Sports Marketing”, Champaign,IL: Human Kinetics Publishers Illinois.Netemeyer R, Durvasala S, Lichtenstein D (1991), “A cross national assessment of thereliability and validity of the CET scale,” Journal of Marketing Research, 28(3), pp 320-327.Nunnally, Jum C., (1978), Psychometric Theory, 2nd ed, New York:McGraw-Hill.Nyeck, Simon, Sylvie Paradis, J-M. Xuereb et J-C Chebat (1996) " Standardisation ouAdaptation des échelles de mesure à travers différents contextes nationaux: l’exemple d’uneéchelle de mesure de l'innovativité". Recherche et Applications en Marketing. Vol.11, 3.57-74.Nyeck, Simon, Bergadaà Michelle (1995), "Quel marketing pour les activités artistiques: uneanalyse qualitative comparée des motivations des consommateurs et producteurs de théâtre,"Recherche et Applications en Marketing, Vol 10-4, pp 27-46.Nyeck, Simon (1994),"Structure de l'Espace Cognitif: Ouverture d'esprit et Comportement duConsommateur," Thèse de doctorat non publiée de l’Université Paris Dauphine et duProgramme doctoral ESSEC (Paris, Juillet 1994).Parsons T., E. A. Shills (1967), “Toward a general theory of action”, Harper and Row,Cambridge.Pedhazur E, Pedhazur-Schmelkin L (1991), “Measurement, Design and analysis : an integratedapproach ”, Hillsdale, NJ : Lawrence Erlbaum Associate Publishers.Peter P (1981), “Construct Validity : A review of basic issues and marketing practices,”Journal of Marketing Research, 18(5), pp 133-145.Peter P., Churchill G. (1986), “Relationships among research design choices and psychometricproperties of rating scales : a Meta analysis,” Journal of Marketing Research, 18(5), pp 1-10.Pope, N. K., K. E. Vogues (1997), “An Exploration of Sponsorship Awareness by ProductCategory and Message Location in televised Sporting Events,” Cyber-Journal of SportMarketing, vol 1(1), pp 16-27.Postman N (1985), “Amusing ouselves to death : Public discourses in the age of showbusiness”, New York, NY : Penguin books.

20

Tinsley H., Barrett T., Kass R. (1977), “Leisure activities and need satisfaction”, journal ofleisure research, vol 9, pp 110-120.Satorra A, Bentler P (1988), “Scaling corrections for chi square statistics in covariancestructure analysis”, American Statistical Association 1988, Proceedings of the business andeconomic sections, Alexandria, VA, American Statistical association, pp 308-313.Settle R., Aldeck P., Belch M. (1979), “Social class determinants of leisure activity”,Advances in consumer research, editor Wilkie W, Vol 6, Ann Arbor, MI : Association forconsumer research, pp 139-145.Sherry J. F (1984), “some implications of consumer oral tradition for reactive marketing”,Advances in consumer research, editor Kinnear T, Vol 11, Provo, UT : Association forconsumer research, pp 741-747.Steenkamp, E. M. Jan-Benedict, and Hans Baumgartner (1992), The Role of OptimumStimulation Level in Exploratory Consumer Behavior, Journal of Consumer Research, 19,december, 434-448.Steenkamp J-B, Baumgartner H. (1995), “ Development and cross-cultural validation of a shortform of CSI as a measure of optimum stimulation level,” International Journal of Research inMarketing, 8, pp 283-299.Stephenson W (1967), “The Play Theory of Mass Communication”, Chicago: University ofChicago Press.Synder C, Fromkin H (1977), “Abnormality as a Positive Characteristic : The development andValidation of Scale measuring Need for Uniqueness,” Journal of Abnormal Psychology, 86, pp518-527.Unger L, Kernan J. (1983), “On the meaning of leisure : an investigation of some determinantsof the subjective experience”, Journal of consumer research, 9th of march, pp 457-472.Valette-Florence P (1988), "Analyse structurelle comparative des composantes des systèmes devaleurs selon Kahle et Rokeach". Recherche et Applications en Marketing, vol III, n 1.Veblen T. (1978), “Théorie de la classe de loisir”, translated by L. Evrard, édition Gallimard.Venkatraman M, Price L (1990), “Differentiating between Cognitive and Sensoryinnovativeness”, Journal of Business Research, 20, pp 293-315.Venkatraman M, MacInnis D (1985), “Epistemic and sensory exploratory behaviors ofHedonic and Cognitive consumers”, Advances in Consumer research, Elizabeth Hirschman etMorris Holbrook editors, 12, Provo : Ut, Association for consumer research, pp 102-107.Wheaton B, Muthèn B, Alwin D, Summers G (1977), “Assessing reliability and stability inpanel models”, Sociological Methodology, San Fransisco, Jossey-Bass, pp 84-136.Wothke W (1993), “Nonpositive definite matrices in structural modelling”, Testing Structuralequation models, Newbury Park, CA, Sage, pp 256-293.