Leisure Sciences 22297 ndash313 2000Copyright Cdeg 2000 Taylor amp Francis0149-040000 $1200 + 00

Recreation Con ict among Skiers and Snowboarders

JERRY J VASKEPAM CAROTHERSMAUREEN P DONNELLYBIFF BAIRD

Department of Natural Resource Recreation and TourismHuman Dimensions in Natural Resources UnitColorado State UniversityFort Collins Colorado USA

Jacob and Schreyerrsquos (1980) four determinants of recreation con ict along with a mea-sure of perceived safety were used to examine both out-group and in-group normativebeliefs about unacceptable behaviors reported by skiers and snowboarders Data wereobtained from surveys distributed at ve ski resorts in Colorado Bivariate analysesindicated that skiers reported more unacceptable behaviors by snowboarders than byfellow skiers and snowboarders also identi ed more out-group than in-group con ictThe skiers and snowboarders in this sample varied in terms of activity style resourcespeci city mode of experience lifestyle tolerance and concerns with safety Two struc-tural equation models were used to identify which determinants in uenced beliefs aboutunacceptable behavior Only activity style signi cantly in uenced both out-group andin-group con ict in both the skier and the snowboarder path models The relative in u-ence of the other predictors varied according to activity (skiing or snowboarding ) andtype of con ict (out-group or in-group)

Keywords activity style resource speci city mode of experience lifestyle tolerancesafety in-group con ict out-group con ict normative behavior

When visitors with differing views on how to use a recreation resource interact with eachother con ict may occur (Adelman Heberlein amp Bonnicksen 1982 Jackson amp Wong1982 Jacob amp Schreyer 1980 Knopp amp Tyger 1973) Recreation con ict is often asym-metrical where the physical presence or actions of one group interfere with the goals (mo-tivations) of another group but the reverse does not hold true (Gibbons amp Ruddell 1995Lucas 1964 Ramthun 1995 Watson Niccolucci amp Williams 1994 Watson Williams ampDaigle 1991) This phenomenon typically occurs when people engaged in traditional activ-ities (eg skiers) interact with those using newer technologies (eg snowboarders) Otherstudies (Thapa 1996 Thapa amp Graefe 1998 1999) however have shown goal interferencecon ict between individuals engaged in the same activity (ie in-group con ict)

Jacob and Schreyerrsquos (1980) goal interference model identi es four major factors thatcontribute to recreation con ict (a) the meaning individuals attach to the activity (b) thesigni cance of the resource to the individual (c) the extent to which the individual is focused

This is a revised version of a paper presented at the 6th International Symposium on Society and ResourceManagement The Pennsylvania State University University Park May 1996

Pam Carothers is now an assistant professor at Jacksonville State University in Jacksonville AlabamaAddress correspondence to Jerry J Vaske Department of Natural Resource Recreation and Tourism Human

Dimensions in Natural Resources Unit Colorado State University Fort Collins CO 80523 E-mail jerryvcnrcolostateedu

297

298 J Vaske et al

on the environment or activity and (d) the usersrsquo acceptance of different lifestylesAlthoughthis model has provided the framework for most con ict studies (Schneider 2000 Watson1995) other concepts have been proposed For example when multiple groups share thesame physical space safety concerns may in uence con ict (Blahna Smith amp Anderson1995) especially for high-speed activities that attract large numbers of participants inrelatively con ned areas such as ski resorts (Finley 1990 Hughes 1988)

Alpine skiing has traditionally dominated North Americarsquos ski slopes In recent yearshowever ski area managers have expressed concern over declining skier numbers andsought ways to recruit new participants Snowboarding with its youth appeal (Baird 1993Thapa 1996) created a new market segment for these resorts Although snowboarding haswitnessed phenomenal growth thus diversifying the use of ski areas questions have arisenregarding the compatibility of the two activities sharing a resource designed speci callyfor skiing Newspaper accounts (Hughes 1988 Meyers 1991) as well as some empiricalevidence have suggested that skiers have not always willingly embraced this new activity(Thapa amp Graefe 1998 1999 P W Williams Dossa amp Fulton 1994)

The study reported here examined both out-group and in-group recreation con ictamong skiers and snowboarders Bivariate analyses were used to compare individuals en-gaged in these two activities relative to the traditional indicators of con ict (activity styleresource speci city mode of experience lifestyle tolerance) as well other con ict-relatedvariables (eg safety) Multivariate analyses were then used to evaluate the relative impactof these predictors on both out-group and in-group beliefs about unacceptable behaviorsassociated with skier and snowboarder interactions

De ning Con ict

Despite the volume of con ict-related research ldquothere has never been agreement on howrecreation con ict should be measuredrdquo (Watson 1995 p 237) Some studies (Thapa ampGraefe 1999 Watson et al 1994) for example have examined the extent to which visitors nd encounters with others to be desirable or undesirable A more direct measure of goalinterference asks respondents to indicate the extent to which encounters with others interferewith their enjoyment (Thapa amp Graefe 1999 Watson et al 1991) Other researchers (Blahnaet al 1995 Carothers Vaske amp Donnelly in press Ramthun 1995 Vaske DonnellyWittmann amp Laidlaw 1995) have focused on the social acceptability of speci c behaviors(eg feeding wildlife mountain biking out of control discourteous skier behavior) De nedin this manner con ict essentially becomes a normative (Ruddell amp Gramann 1994) asopposed toa motivational (goal) issueNorms are evaluative beliefs (standards)regarding ac-ceptable behavior in a given context (see Vaske Shelby Graefe amp Heberlein 1986 ShelbyVaske amp Donnelly 1996 for reviews) In this articlewe focus on skiersrsquo and snowboardersrsquonormative beliefs about unacceptable behaviors as indicators of recreation con ict

There are at least two sources of unacceptable behavior those resulting from inter-actions with other individuals involved in the same activity (in-group con ict) and thoseassociated with interactions with other individuals involved in different activities (out-groupcon ict) Whereas most research has focused on out-group con ict (Adelman et al 1982Devall amp Harry 1981 Watson et al 1991 1994 P W Williams et al 1994) some investi-gations have explored beliefs about unacceptable behavior occurring as a result of in-groupinteractions Studies by Todd and Graefe (1989) and Thapa and Graefe (1998 1999) forexample found that goal interference was more likely to be attributed to in-group than toout-group con ict In general however the con ict literature has shown that recreationistsare more tolerant of individuals engaged in the same activity as themselves than they arewith those engaged in a different activity (Jackson amp Wong 1982 Gibbons amp Ruddell1995 Knopp amp Tyger 1973 Lucas 1964) We therefore hypothesize

298 J Vaske et al

on the environment or activity and (d) the usersrsquo acceptance of different lifestylesAlthoughthis model has provided the framework for most con ict studies (Schneider 2000 Watson1995) other concepts have been proposed For example when multiple groups share thesame physical space safety concerns may in uence con ict (Blahna Smith amp Anderson1995) especially for high-speed activities that attract large numbers of participants inrelatively con ned areas such as ski resorts (Finley 1990 Hughes 1988)

Alpine skiing has traditionally dominated North Americarsquos ski slopes In recent yearshowever ski area managers have expressed concern over declining skier numbers andsought ways to recruit new participants Snowboarding with its youth appeal (Baird 1993Thapa 1996) created a new market segment for these resorts Although snowboarding haswitnessed phenomenal growth thus diversifying the use of ski areas questions have arisenregarding the compatibility of the two activities sharing a resource designed speci callyfor skiing Newspaper accounts (Hughes 1988 Meyers 1991) as well as some empiricalevidence have suggested that skiers have not always willingly embraced this new activity(Thapa amp Graefe 1998 1999 P W Williams Dossa amp Fulton 1994)

The study reported here examined both out-group and in-group recreation con ictamong skiers and snowboarders Bivariate analyses were used to compare individuals en-gaged in these two activities relative to the traditional indicators of con ict (activity styleresource speci city mode of experience lifestyle tolerance) as well other con ict-relatedvariables (eg safety) Multivariate analyses were then used to evaluate the relative impactof these predictors on both out-group and in-group beliefs about unacceptable behaviorsassociated with skier and snowboarder interactions

De ning Con ict

Despite the volume of con ict-related research ldquothere has never been agreement on howrecreation con ict should be measuredrdquo (Watson 1995 p 237) Some studies (Thapa ampGraefe 1999 Watson et al 1994) for example have examined the extent to which visitors nd encounters with others to be desirable or undesirable A more direct measure of goalinterference asks respondents to indicate the extent to which encounters with others interferewith their enjoyment (Thapa amp Graefe 1999 Watson et al 1991) Other researchers (Blahnaet al 1995 Carothers Vaske amp Donnelly in press Ramthun 1995 Vaske DonnellyWittmann amp Laidlaw 1995) have focused on the social acceptability of speci c behaviors(eg feeding wildlife mountain biking out of control discourteous skier behavior) De nedin this manner con ict essentially becomes a normative (Ruddell amp Gramann 1994) asopposed toa motivational (goal) issueNorms are evaluative beliefs (standards)regarding ac-ceptable behavior in a given context (see Vaske Shelby Graefe amp Heberlein 1986 ShelbyVaske amp Donnelly 1996 for reviews) In this articlewe focus on skiersrsquo and snowboardersrsquonormative beliefs about unacceptable behaviors as indicators of recreation con ict

There are at least two sources of unacceptable behavior those resulting from inter-actions with other individuals involved in the same activity (in-group con ict) and thoseassociated with interactions with other individuals involved in different activities (out-groupcon ict) Whereas most research has focused on out-group con ict (Adelman et al 1982Devall amp Harry 1981 Watson et al 1991 1994 P W Williams et al 1994) some investi-gations have explored beliefs about unacceptable behavior occurring as a result of in-groupinteractions Studies by Todd and Graefe (1989) and Thapa and Graefe (1998 1999) forexample found that goal interference was more likely to be attributed to in-group than toout-group con ict In general however the con ict literature has shown that recreationistsare more tolerant of individuals engaged in the same activity as themselves than they arewith those engaged in a different activity (Jackson amp Wong 1982 Gibbons amp Ruddell1995 Knopp amp Tyger 1973 Lucas 1964) We therefore hypothesize

SkierSnowboarder Con ict 299

H1 Skiers and snowboarders will report more out-group than in-group unaccept-able behaviors (con ict)

Sources of Con ict

Jacob and Schreyer (1980) proposed four major classes of determinants (activity styleresource speci city mode of experience lifestyle tolerance) that in uence recreation con- ict Activity style refers to the personal meaning individuals assign to the activity Theseindividual meanings not the activity itself contribute to con ict evaluations The moreintense an individualrsquos activity style the greater the likelihood that contact with less in-tense participants will result in con ict Intensity of participation has been operationalizedrelative to an individualrsquos level of involvement in a sport (eg total years of participa-tion days of participation per year) P W Williams et al (1994) for example comparedskiers and snowboarders at 16 ski resorts in British Columbia Canada Results indicatedthat as a group the skiers had pursued their sport for signi cantly more years and weremore likely to take advantage of the services offered at the resort (eg lessons) thanwere the snowboarders On the other hand the snowboarders reported more overnighttrips per year and more days of participation per year than the skiers Taken togetherthe ndings from the P W Williams et al study showed that skiers were more involvedwith their activity based on years of participation whereas snowboarders were more in-volved on the basis of amount of participation per year Because these ndings do notindicate a clear pattern of activity style differences between skiers and snowboarders wehypothesize

H2 Skiers and snowboarders will not differ in the importance they attach to theactivity

Resource speci city relates to the signi cance recreationists attach to a speci c re-source Those less attached to the resource are seen to disrupt the traditional uses (Jacobamp Schreyer 1980) Hiking for example represents a traditional activity on most trailswhereas mountain biking is a relatively new sport (Chavez 1999 Woodward 1996) Hogerand Chavez (1998) have shown that hikers view mountain biking as intrusive and are con-cerned with the impact mountain biking has on the environment These ndings as wellas those of other researchers (Watson Zaglauer amp Stewart 1996) indicate that individualsengaged in more traditional activities may place greater signi cance on the resource thanthose participating in nontraditional recreation pursuits

Relative to skiers and snowboarders skiing represents the traditional activity on mostNorth American slopes The empirical evidence supporting greater resource speci cityamong skiers however has shown a mixed pattern of results P W Williams et al (1994)for example suggested that skiers saw themselves as more attached to the resource than thesnowboarders The skiers viewed the snowboarders as intruding on the pristine quality of theresort exhibiting little respect for the natural beauty of the environment The snowboarderson the other hand also expressed a closeness to the natural environment but were moreconcerned about the freedom to pursue their activity without restrictions on where theycould snowboard In other words snowboarding was not allowed on all trails Overall theP W Williams et al study revealed few differences in resource speci city between thetwo groups Participants in the two activities attached importance to the ski resort but fordifferent reasons Therefore we hypothesize

H3 Skiers and snowboarders will not differ in the importance they attach to theresource

300 J Vaske et al

Jacob and Schreyer (1980) described mode of experience as a continuum rangingfrom unfocused to focused ldquoAs the mode of experiencing the environment becomes morefocused an individual produces more rigid de nitions of what constitutes acceptable stimuliand is increasingly intolerant of external stimulationrdquo (Jacob amp Schreyer 1980 p 375)Snowboarders in the P W Williams et al (1994) investigation were focused on technicaland competency-related issues Alternatively the skiers were more focused on the naturalfeatures of the environment and often complained about the snowboarders scraping andrutting the trails and ruining the moguls Similar to the above logic both groups appear tobe focused but for different reasons We hypothesize

H4 Skiers and snowboarders will not differ in their mode of experience

Lifestyle tolerance refers to the tendency to accept or reject lifestyles different thanonersquos own (Jacob amp Schreyer 1980) As noted by Ivy Stewart and Lue (1992) toler-ance is typically associated with beliefs about a particular group rather than reactions tospeci c behaviors When recreationists encounter others a cognitive processing of infor-mation occurs This action often results in the categorization of others according to somegroup membership which helps to simplify and order environmental stimuli Differences inlifestyles are often communicated through visual cues such as the equipment used by recre-ationists engaged in different activities (eg guns for hunting vs binoculars for wildlifeviewing Vaske et al 1995) ldquoRecreation in-groups and out-groups represent categoriesan individual establishes on the basis of perceived or imagined lifestyle similarities anddifferencesrdquo (Jacob amp Schreyer 1980 p 376) Though useful for maintaining a view ofthe world it can also lead to unjusti ed generalizations about other groups (Baron Kerramp Miller 1992 Ramthun 1995) Those who demonstrate low tolerance for persons withdiffering lifestyles will be more likely to experience con ict

P W Williamset al (1994) suggested that skiers and snowboarders have differing viewsof each other Skiers felt threatened by the snowboardersrsquo different approach they evaluatedthe language clothes and on-slope behavior of snowboarders as intimidating and had theperception that snowboarders purposely created con ict situations Snowboarders on theother hand perceived skiers as predictable and showed less concern for their presence onthe slopes The British Columbian snowboarders however were more willing to share theresource with skiers than the skiers were with snowboarders (P W Williams et al 1994)These group differences may increase the potential for a culture clash between skiers andsnowboarders (Hughes 1988)

H5 Skiers will be less tolerant of the snowboardersrsquo lifestyle than vice versa

Although not explicitly addressed by Jacob and Schreyer (1980) safety concerns rep-resent a potential indicator of con ict Skiers and snowboarders often share the slopes withlarge numbers of fellow recreationists Participants in each group traverse the slopes at highspeeds Speed when combined with large numbers of recreationists can lead to potentiallydangerous situations (Finley 1990)

In 1985 only 6 of Colorado ski areas allowed snowboarders to ride their lifts (Meyers1991) Safety was the major consideration in these early bans on snowboarding as ski areamanagers questioned whether theycould coexist with skiers (Asher amp Markels 1992 Finley1990) To some extent these concerns were legitimate as early snowboards lacked steeledges retention devices and sidecuts making control dif cult Insurance carriers declinedto place these early boards in the category of ldquodirectional devicesrdquo and refused to issuecoverage to ski areas that allowed the sport (Aitkens 1990)

300 J Vaske et al

Jacob and Schreyer (1980) described mode of experience as a continuum rangingfrom unfocused to focused ldquoAs the mode of experiencing the environment becomes morefocused an individual produces more rigid de nitions of what constitutes acceptable stimuliand is increasingly intolerant of external stimulationrdquo (Jacob amp Schreyer 1980 p 375)Snowboarders in the P W Williams et al (1994) investigation were focused on technicaland competency-related issues Alternatively the skiers were more focused on the naturalfeatures of the environment and often complained about the snowboarders scraping andrutting the trails and ruining the moguls Similar to the above logic both groups appear tobe focused but for different reasons We hypothesize

H4 Skiers and snowboarders will not differ in their mode of experience

Lifestyle tolerance refers to the tendency to accept or reject lifestyles different thanonersquos own (Jacob amp Schreyer 1980) As noted by Ivy Stewart and Lue (1992) toler-ance is typically associated with beliefs about a particular group rather than reactions tospeci c behaviors When recreationists encounter others a cognitive processing of infor-mation occurs This action often results in the categorization of others according to somegroup membership which helps to simplify and order environmental stimuli Differences inlifestyles are often communicated through visual cues such as the equipment used by recre-ationists engaged in different activities (eg guns for hunting vs binoculars for wildlifeviewing Vaske et al 1995) ldquoRecreation in-groups and out-groups represent categoriesan individual establishes on the basis of perceived or imagined lifestyle similarities anddifferencesrdquo (Jacob amp Schreyer 1980 p 376) Though useful for maintaining a view ofthe world it can also lead to unjusti ed generalizations about other groups (Baron Kerramp Miller 1992 Ramthun 1995) Those who demonstrate low tolerance for persons withdiffering lifestyles will be more likely to experience con ict

P W Williamset al (1994) suggested that skiers and snowboarders have differing viewsof each other Skiers felt threatened by the snowboardersrsquo different approach they evaluatedthe language clothes and on-slope behavior of snowboarders as intimidating and had theperception that snowboarders purposely created con ict situations Snowboarders on theother hand perceived skiers as predictable and showed less concern for their presence onthe slopes The British Columbian snowboarders however were more willing to share theresource with skiers than the skiers were with snowboarders (P W Williams et al 1994)These group differences may increase the potential for a culture clash between skiers andsnowboarders (Hughes 1988)

H5 Skiers will be less tolerant of the snowboardersrsquo lifestyle than vice versa

Although not explicitly addressed by Jacob and Schreyer (1980) safety concerns rep-resent a potential indicator of con ict Skiers and snowboarders often share the slopes withlarge numbers of fellow recreationists Participants in each group traverse the slopes at highspeeds Speed when combined with large numbers of recreationists can lead to potentiallydangerous situations (Finley 1990)

In 1985 only 6 of Colorado ski areas allowed snowboarders to ride their lifts (Meyers1991) Safety was the major consideration in these early bans on snowboarding as ski areamanagers questioned whether theycould coexist with skiers (Asher amp Markels 1992 Finley1990) To some extent these concerns were legitimate as early snowboards lacked steeledges retention devices and sidecuts making control dif cult Insurance carriers declinedto place these early boards in the category of ldquodirectional devicesrdquo and refused to issuecoverage to ski areas that allowed the sport (Aitkens 1990)

SkierSnowboarder Con ict 301

Although improvements in snowboard-manufacturing technology (leading to improvedcontrol) have played a role in the current near-unanimous acceptance of snowboarding atColorado ski resorts skiers may still perceive snowboarders as reckless individuals and feelthreatened by their presence on the slopes (Meyers 1991) Such safety concerns may beattributed to beliefs about unacceptable behaviors such as unsafe jumping or riding out ofcontrol (White 1990) Taken together these observations suggest

H6 Skiers will perceive more safety-related problems associated with snowboard-ing than vice versa

Conceptual Model

All hypotheses proposed thus far have suggested bivariate relationships among the variablesTo address the combined in uence of these variables on out-group and in-group beliefs aboutunacceptable behaviors we developed a multivariate conceptual model On the basis of theresearch and popular literature summarized above the model predicts that activity styleresource speci city mode of experience and safety concerns will increase the likelihoodof con ict (both out-group and in-group) Lifestyle tolerance on the other hand should benegatively associated with perceived con ict These relationships are shown in Figure 1and are stated formally as hypotheses below

H7 As the importance attached to the activity increases out-group and in-groupbeliefs about unacceptable behaviors (con ict) will increase

H8 As the importance attached to the resource increases out-group and in-groupbeliefs about unacceptable behaviors (con ict) will increase

H9 As the mode of experience increases (becomes more focused) out-group andin-group beliefs about unacceptable behaviors (con ict) will increase

H10 As tolerance for lifestyle diversity increases out-group and in-group beliefsabout unacceptable behaviors (con ict) will decrease

H11 As perceptions of safety-related problems increase awareness of out-groupand in-group beliefs about unacceptable behaviors (con ict) will increase

FIGURE 1 Expanded con ict model

302 J Vaske et al

Method

Study Locations and Sampling

Date were collected from ve Colorado ski areas (Arapahoe Basin Copper MountainEldora Winter Park and Steamboat Springs)1 between December 1992 and February 1993Mail-back surveys were distributed on randomly selected days at lift lines and ski arearestaurants At the lift lines every 10th individual was selected In the restaurants anindividual was selected at random from every 5th table Of the 1252 surveys distributed on-site595 usable questionnaires were mailedback (response rate = 48) Funding constraintsdid not allow for any additional follow-up to nonrespondents The sample consisted of 383skiers and 212 snowboarders2

Variables Measured

Con ictA multiple-item index was created to measure observed unacceptable behaviors (con-

ict) between skiers and snowboarders Speci c items asked if skierssnowboarders (a)failed to be aware of others around them (b) were not keeping an adequate distance fromothers (c) failed to yield the right of way to the downhill skiersnowboarder (d) behavedin a discourteous manner (e) cut others off and (f) failed to be aware of and yield toless-advanced skierssnowboarders Respondents indicated how often these behaviors wereseen Response categories were never (1) rarely (2) sometimes (3) frequently (4) andalmost always (5)

Predictors of Con ictAn activity style scale was created that re ected investment in the sport Respondents

reported the number of days per year skiedsnowboarded (responses ranged from 1 to5 to more than 50) the number of skissnowboards owned (zero to more than three) theapproximate amount of money invested in equipment clothing and accessories ($0ndash$100 tomore than $3000) number of years skiingsnowboarding (1 to more than 20) and a rating oftheir skiingsnowboarding ability (beginner to expert) For resource speci city respondentsindicated their agreement with the following place attachment statements (D Williams ampRoggenbuck 1989) (a) ldquothis ski area means a lot to merdquo (b) ldquoa lot of my life is organizedaround this ski areardquo (c) ldquothis ski area is the best place for what I like to dordquo and (d) ldquoIidentify strongly with this ski areardquo Responses were coded on 5-point scales ranging fromstrongly disagree (1) to strongly agree (5) A lifestyle tolerance index was created by askingthe respondents to agree or disagree that snowboarders and skiers have similar (a) lifestyles(b) levels of education (c) incomes (d) attitudes toward the environment and (e) feelingsabout the value of this area Five-point Likert scales ranging from strongly disagree (1) tostrongly agree (5) were used to measure responses to these variables

Mode of experience was measured with a single item Individuals indicated the ex-tent to which they agreed or disagreed that they focused most of their attention on theirskiingsnowboarding skills Responses were coded on a 5-point Likert scale ranging fromstrongly disagree (1) to strongly agree (5)

Safety was also measured with a single-item statement (again using the 5-point Likertdisagree-to-agree scale) that it is not safe to have snowboarders and skiers share the sametrails

Analysis

Reliability analyses were used to determine the internal consistency of each of the scaledmeasurement items Con rmatory factor analyses examined the extent to which the four

SkierSnowboarder Con ict 303

Jacob and Schreyer (1980) determinants of con ict and safety provided a good t to thedata We used t tests to analyze bivariate differences between skiers and snowboarders andstructural equation path analyses to address the predictive validity of the models LISREL814 (Joreskog amp Sorbom 1993) was used for this analysis3

Results

Reliability and Con rmatory Factor Analyses

Tables 1 and 2 show the reliabilityand con rmatory factor analyses for the items in the skierindices and Tables 3 and 4 provide the same information for the snowboarders The primarydependent variables in the models out-group and in-group con ict were computed fromsix beliefs about unacceptable behaviors associated with skiing (Table 1) and snowboarding(Table 3) For the skiers the reliability coef cients for the two indices were 93 (out-group)and 88 (in-group) The reliability coef cients for the snowboarders were 87 (out-group)and 83 (in-group) The con rmatory factor analyses demonstrated that the data providedan acceptable t for both the skiers (factor loadings cedil 68 SE middot048) and snowboarders(factor loadings cedil 61 SE middot069)

Cronbachrsquos alphas for the items in the skiersrsquo activity style (79) resource speci city(79) and lifestyle tolerance (80) indices are given in Table 2 The alphas for the snow-boarders (Table 4) were similar activity style 85 resource speci city 76 and lifestyletolerance 75 For both the skiers and snowboarders the standardized factor loadings ( cedil 48in all cases) and standard errors (SE middot074 in all cases) provided additional support forcombining these items into their respective latent constructs

TABLE 1 Skier Indices for In-Group and Out-Group Con ict

Standardized CronbachrsquosStatement M factor loading SE ta a

Unacceptable snowboarder behavior (out-group con ict)b 93Fail to be aware of others 341 81 046 1745Not adequate distance 334 86 045 1887Fail to yield right of way to 324 86 045 1894

downhill userBehaves in a discourteous 299 81 046 1756

mannerCuts others off 313 85 045 1865Fails to yield to the less 311 82 046 1773

advanced userUnacceptable skier behavior (in-group con ict)b 88

Fail to be aware of others 323 74 047 1571Not adequate distance 331 77 046 1678Fail to yield right of way to 311 81 045 1803

downhill userBehaves in a discourteous 267 71 047 1500

mannerCuts others off 305 75 047 1601Fails to yield to the less 288 68 048 1421

advanced user

aAll ts signi cant at p lt 001bVariables coded on a 5-point scale ranging from never (1) to almost always (5)

302 J Vaske et al

Method

Study Locations and Sampling

Date were collected from ve Colorado ski areas (Arapahoe Basin Copper MountainEldora Winter Park and Steamboat Springs)1 between December 1992 and February 1993Mail-back surveys were distributed on randomly selected days at lift lines and ski arearestaurants At the lift lines every 10th individual was selected In the restaurants anindividual was selected at random from every 5th table Of the 1252 surveys distributed on-site595 usable questionnaires were mailedback (response rate = 48) Funding constraintsdid not allow for any additional follow-up to nonrespondents The sample consisted of 383skiers and 212 snowboarders2

Variables Measured

Con ictA multiple-item index was created to measure observed unacceptable behaviors (con-

ict) between skiers and snowboarders Speci c items asked if skierssnowboarders (a)failed to be aware of others around them (b) were not keeping an adequate distance fromothers (c) failed to yield the right of way to the downhill skiersnowboarder (d) behavedin a discourteous manner (e) cut others off and (f) failed to be aware of and yield toless-advanced skierssnowboarders Respondents indicated how often these behaviors wereseen Response categories were never (1) rarely (2) sometimes (3) frequently (4) andalmost always (5)

Predictors of Con ictAn activity style scale was created that re ected investment in the sport Respondents

reported the number of days per year skiedsnowboarded (responses ranged from 1 to5 to more than 50) the number of skissnowboards owned (zero to more than three) theapproximate amount of money invested in equipment clothing and accessories ($0ndash$100 tomore than $3000) number of years skiingsnowboarding (1 to more than 20) and a rating oftheir skiingsnowboarding ability (beginner to expert) For resource speci city respondentsindicated their agreement with the following place attachment statements (D Williams ampRoggenbuck 1989) (a) ldquothis ski area means a lot to merdquo (b) ldquoa lot of my life is organizedaround this ski areardquo (c) ldquothis ski area is the best place for what I like to dordquo and (d) ldquoIidentify strongly with this ski areardquo Responses were coded on 5-point scales ranging fromstrongly disagree (1) to strongly agree (5) A lifestyle tolerance index was created by askingthe respondents to agree or disagree that snowboarders and skiers have similar (a) lifestyles(b) levels of education (c) incomes (d) attitudes toward the environment and (e) feelingsabout the value of this area Five-point Likert scales ranging from strongly disagree (1) tostrongly agree (5) were used to measure responses to these variables

Mode of experience was measured with a single item Individuals indicated the ex-tent to which they agreed or disagreed that they focused most of their attention on theirskiingsnowboarding skills Responses were coded on a 5-point Likert scale ranging fromstrongly disagree (1) to strongly agree (5)

Safety was also measured with a single-item statement (again using the 5-point Likertdisagree-to-agree scale) that it is not safe to have snowboarders and skiers share the sametrails

Analysis

Reliability analyses were used to determine the internal consistency of each of the scaledmeasurement items Con rmatory factor analyses examined the extent to which the four

SkierSnowboarder Con ict 303

Jacob and Schreyer (1980) determinants of con ict and safety provided a good t to thedata We used t tests to analyze bivariate differences between skiers and snowboarders andstructural equation path analyses to address the predictive validity of the models LISREL814 (Joreskog amp Sorbom 1993) was used for this analysis3

Results

Reliability and Con rmatory Factor Analyses

Tables 1 and 2 show the reliabilityand con rmatory factor analyses for the items in the skierindices and Tables 3 and 4 provide the same information for the snowboarders The primarydependent variables in the models out-group and in-group con ict were computed fromsix beliefs about unacceptable behaviors associated with skiing (Table 1) and snowboarding(Table 3) For the skiers the reliability coef cients for the two indices were 93 (out-group)and 88 (in-group) The reliability coef cients for the snowboarders were 87 (out-group)and 83 (in-group) The con rmatory factor analyses demonstrated that the data providedan acceptable t for both the skiers (factor loadings cedil 68 SE middot048) and snowboarders(factor loadings cedil 61 SE middot069)

Cronbachrsquos alphas for the items in the skiersrsquo activity style (79) resource speci city(79) and lifestyle tolerance (80) indices are given in Table 2 The alphas for the snow-boarders (Table 4) were similar activity style 85 resource speci city 76 and lifestyletolerance 75 For both the skiers and snowboarders the standardized factor loadings ( cedil 48in all cases) and standard errors (SE middot074 in all cases) provided additional support forcombining these items into their respective latent constructs

TABLE 1 Skier Indices for In-Group and Out-Group Con ict

Standardized CronbachrsquosStatement M factor loading SE ta a

Unacceptable snowboarder behavior (out-group con ict)b 93Fail to be aware of others 341 81 046 1745Not adequate distance 334 86 045 1887Fail to yield right of way to 324 86 045 1894

downhill userBehaves in a discourteous 299 81 046 1756

mannerCuts others off 313 85 045 1865Fails to yield to the less 311 82 046 1773

advanced userUnacceptable skier behavior (in-group con ict)b 88

Fail to be aware of others 323 74 047 1571Not adequate distance 331 77 046 1678Fail to yield right of way to 311 81 045 1803

downhill userBehaves in a discourteous 267 71 047 1500

mannerCuts others off 305 75 047 1601Fails to yield to the less 288 68 048 1421

advanced user

aAll ts signi cant at p lt 001bVariables coded on a 5-point scale ranging from never (1) to almost always (5)

SkierSnowboarder Con ict 303

Jacob and Schreyer (1980) determinants of con ict and safety provided a good t to thedata We used t tests to analyze bivariate differences between skiers and snowboarders andstructural equation path analyses to address the predictive validity of the models LISREL814 (Joreskog amp Sorbom 1993) was used for this analysis3

Results

Reliability and Con rmatory Factor Analyses

Tables 1 and 2 show the reliabilityand con rmatory factor analyses for the items in the skierindices and Tables 3 and 4 provide the same information for the snowboarders The primarydependent variables in the models out-group and in-group con ict were computed fromsix beliefs about unacceptable behaviors associated with skiing (Table 1) and snowboarding(Table 3) For the skiers the reliability coef cients for the two indices were 93 (out-group)and 88 (in-group) The reliability coef cients for the snowboarders were 87 (out-group)and 83 (in-group) The con rmatory factor analyses demonstrated that the data providedan acceptable t for both the skiers (factor loadings cedil 68 SE middot048) and snowboarders(factor loadings cedil 61 SE middot069)

Cronbachrsquos alphas for the items in the skiersrsquo activity style (79) resource speci city(79) and lifestyle tolerance (80) indices are given in Table 2 The alphas for the snow-boarders (Table 4) were similar activity style 85 resource speci city 76 and lifestyletolerance 75 For both the skiers and snowboarders the standardized factor loadings ( cedil 48in all cases) and standard errors (SE middot074 in all cases) provided additional support forcombining these items into their respective latent constructs

TABLE 1 Skier Indices for In-Group and Out-Group Con ict

Standardized CronbachrsquosStatement M factor loading SE ta a

Unacceptable snowboarder behavior (out-group con ict)b 93Fail to be aware of others 341 81 046 1745Not adequate distance 334 86 045 1887Fail to yield right of way to 324 86 045 1894

downhill userBehaves in a discourteous 299 81 046 1756

mannerCuts others off 313 85 045 1865Fails to yield to the less 311 82 046 1773

advanced userUnacceptable skier behavior (in-group con ict)b 88

Fail to be aware of others 323 74 047 1571Not adequate distance 331 77 046 1678Fail to yield right of way to 311 81 045 1803

downhill userBehaves in a discourteous 267 71 047 1500

mannerCuts others off 305 75 047 1601Fails to yield to the less 288 68 048 1421

advanced user

aAll ts signi cant at p lt 001bVariables coded on a 5-point scale ranging from never (1) to almost always (5)

304 J Vaske et al

TABLE 2 Skier Indices for Activity Style Resource Speci city and Lifestyle Tolerance

Standardized CronbachrsquosStatement M factor loading SE t a a

Activity styleb 79Days per year skiedc 275 65 051 1285Pairs of skis ownedd 241 69 049 1418Money invested in skiinge 300 63 050 1258Number of years skiing f 509 61 052 1184Rating of skiing abilityg 362 78 047 1664

Resource speci cityh 79This area means a lot to me 370 68 050 1357Lots of my life is organized 222 67 051 1321

around this areaThis area is best for what I like to do 306 64 051 1258I identify strongly with this area 281 79 049 1628

Lifestyle toleranceh Skiers and 80snowboarders have similarLifestyles 259 72 049 1467Education 268 72 049 1447Income 292 71 049 1442Attitudes toward the environment 245 48 055 870Feelings about the arearsquos value 333 60 051 1171

aAll ts signi cant at p lt 001bBecause the items in the activity style index used different response scales all variables were

standardized before computing the indexcVariable coded on a scale ranging from 1ndash5 (1) to gt50 (8)dVariable coded on a scale ranging from 0 (1) to gt3 (8)eVariable coded on a scale ranging from 0 to 100 (1) to gt 3000 (8)f Variable coded on a scale ranging from 1 (1) to gt20 (8)gVariable coded on a scale ranging from beginner (1) to expert (8)hVariables coded on a scale ranging from strongly disagree (1) to strongly agree (5)

Bivariate Analyses

Consistent with Hypothesis 1 we found signi cant differences between skiers (M =320) and snowboarders (M = 305) for unacceptable snowboarder behaviors t (592) =258 p = 010 and for unacceptable skier behaviors (M = 304 and 356 respectively)t (591) = 890 p lt 001 These ndings indicate that skiers reported more unacceptablebehaviors for snowboarders than for fellow skiers Similarly snowboarders identi ed moreout-group than in-group con ict

Skiers and snowboarders were predicted to be similar in activity style (Hypothesis 2)resource speci city (Hypothesis 3) and mode of the experience (Hypothesis 4) The Col-orado data failed to support any of these hypotheses signi cant differences (p lt 001)between skiers and snowboarders were observed for all three constructs The skiers in thissample attached more importance to the activity than did the snowboarders Conversely thesnowboarders rated the resource more highly and were more focused on their activity thanwere the skiers

We predicted that skiers and snowboarders would differ in their tolerances foreach otherrsquos lifestyles (Hypothesis 5) and their perceptions of safety-related beliefs

SkierSnowboarder Con ict 305

TABLE 3 Snowboarder Indices for In-Group and Out-Group Con icts

Standardized CronbachrsquosStatement M factor loading SE t a a

Unacceptable snowboarder behavior (in-group con ict)b 83Fail to be aware of others 319 63 069 914Not adequate distance 306 73 067 1094Fail to yield right of way to downhill 313 75 066 1128

userBehaves in a discourteous manner 302 65 068 955Cuts others off 299 66 068 969Fails to yield to the less advanced 294 61 069 881

userUnacceptable skier behavior (out-group con ict)b 87

Fail to be aware of others 372 74 065 1140Not adequate distance 356 71 066 1083Fail to yield right of way to downhill 366 70 066 1052

userBehaves in a discourteous manner 340 74 065 1139Cuts others off 361 82 064 1284Fails to yield to the less advanced 342 66 067 993

user

aAll ts signi cant at p lt 001bVariables coded on a 5-point scale ranging from never (1) to almost always (5)

(Hypothesis 6) Both of these hypotheses were supported by the data (Table 5) The averagescores for the lifestyle tolerance index indicated that snowboarders were more likely thanskiers to perceive the two groups to be similar Compared with the snowboarders howeverthe skiers reported more unacceptable safety-related behaviors

Multivariate Analysis

The overall t of the skier and snowboarder models was assessed using ve indicators chi-square chi-squaredegree of freedom goodness-of- t index (GFI) comparative t index(CFI) and root mean square residual (RMR Table 6) Although both models produced asigni cant chi-square sample size tends to in ate this statistic Consequently Marsh andHocevar (1985) suggested that the chi-square should be evaluated in relation to the modelrsquosdegrees of freedom with a v 2 df ratio of 21 to 51 indicating an acceptable t This ratiofor both the skier ( v 2 df = 275) and snowboarder ( v 2 df = 168) models fell within thisrange Values for the GFI and CFI ranged from 91 to 93 also indicating an acceptable t forthe two models (Bollen 1989) Finally the RMRswhich measure the average discrepanciesbetween the observed and the model-generated covariances were less than or equal to 061for both skiers and snowboarders suggesting a close t of the data (Church amp Burke1994)

Skier Path ModelConsistent with Hypothesis 7 a signi cant and positive relationship between activity

style and out-group ( b = 023 p lt 05) and in-group ( b = 027 p lt 05) con ict wasobserved in the skier model4 (Figure 2) Resource speci city however did not signi cantly

306 J Vaske et al

TABLE 4 Snowboarder Indices for Activity Style Resource Speci city and LifestyleTolerance

Standardized CronbachStatement M factor loading SE ta a

Activity styleb 85Days per year snowboardc 482 72 064 1133Snowboards ownedd 229 76 060 1262Money invested in snowboardinge 286 61 065 941Number of years snowboarding f 226 76 063 1207Rating of snowboarding abilityg 352 84 058 1455

Resource speci cityh 76This area means a lot to me 379 78 067 1170Lots of my life is organized 292 67 069 967

around this areaThis area is best for what I like to do 313 53 072 736I identify strongly with this area 311 68 069 993

Lifestyle toleranceh Skiers and snowboarders have similar 75Lifestyles 293 60 072 839Education 327 52 074 701Income 312 69 070 980Attitudes toward the environment 273 50 074 673Feelings about the arearsquos value 364 68 070 976

aAll ts signi cant at p lt 001bBecause the items in the activity style index used different response scales all variables were

standardized before computing the indexcVariable coded on a scale ranging from 1ndash5 (1) to gt50 (8)d Variable coded on a scale ranging from 0 (1) to gt3 (8)eVariable coded on a scale ranging from 0 to 100 (1) to gt 3000 (8)f Variable coded on a scale ranging from 1 (1) to gt20 (8)gVariable coded on a scale ranging from beginner (1) to expert (8)hVariable coded on a scale ranging from strongly disagree (1) to strongly agree (5)

in uence beliefs about either out-group ( b = 008 ns) or in-group ( b = 006 ns ) unac-ceptable behavior as predicted by Hypothesis 8 Hypothesis 9 which predicted a positiverelationship between mode of experience and the two con ict constructs was only partiallysupported Similarly the predicted in uence of lifestyle tolerance on perceived con ictreceived only partial support (Hypothesis 10) In-group con ict increased for skiers whowere focused on their activity (b = 014 p lt 05) but the relationship between out-groupcon ict and mode of the experience was not signi cant ( b = 000 ns) Although sig-ni cant paths between the tolerance variable and out-group ( b = iexcl 028 p lt 05) andin-group ( b = 016 p lt 05) con ict were observed only the negative relationship waspredicted by theory Finally safety concerns (Hypothesis 11) in uenced out-group con ict( b = 042 p lt 05) but had no effect on beliefs about in-group unacceptable behavior( b = 010 ns )

Taken together three of the ve predictor variables (activity style lifestyle toler-ance safety) accounted for 44 of the variance in skiersrsquo reported unacceptable behaviorwith snowboarders (out-group con ict) Of these safety concerns had the largest relativein uence Three variables also explained skiersrsquo evaluations of con ict with other skiers

SkierSnowboarder Con ict 307

TABLE 5 Bivariate Analyses Comparing Skiers and Snowboarders

Skier SnowboarderVariable (M N = 383) (M N = 212) t p

Unacceptable snowboarder behaviora 320 305 258 010Unacceptable skier behaviora 304 356 890 001Activity styleb 445 411 322 001Resource speci cityc 295 324 423 001Mode of experienced 381 424 636 001Lifestyle tolerancee 279 315 616 001Safetyd 286 183 1047 001

aSix-variable index coded on a 5-point scale ranging from never (1) to almost always (5)bFive-variable standardized indexcFour-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)dSingle-variable coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)eFive-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)

(in-group) but it is important to note the following First although lifestyle tolerancesigni cantly in uenced in-group con ict the positive relationship was opposite that pre-dicted by theory Second the in-group model accounted for only 10 of the variance incon ict

Snowboarder Path ModelFigure 3 diagrams the ndings from the snowboarder path model5 Three variables (ac-

tivity style b = 022 p lt 05 resource speci city b = 023 p lt 05 lifestyle toleranceb = iexcl 033 p lt 05) signi cantly in uenced snowboardersrsquo beliefs about unacceptableskier behavior (out-group con ict) and accounted for 23 of the variance in the criterionconstruct All relationships were in the predicted directions Relative to snowboardersrsquo eval-uations of other snowboarders (in-group con ict) only activity style ( b = 040 p lt 05)had a signi cant in uence accounting for 21 of the variance

Overall these ndings are consistent with Hypothesis 7 but provide only partial supportfor Hypotheses 8 and 10 The mode of experience (Hypothesis 9) and safety (Hypothesis 11)relationships were not supported for either out-group or in-group con ict

TABLE 6 Goodness-of- t Statistics

Measure Skier model Snowboarder model

Chi-square 31866 19535Degrees of freedom 116 116Chi-squaredegrees of freedom 275 168Goodness-of- t index 92 91Comparative t index 91 93Root mean square residual 059 061

308 J Vaske et al

FIGURE 2 Skier con ict model Only signi cant paths (p lt 05) are shown

Discussion

Recent advances in technology such as snowboards and mountain bikes have changed theway people recreate in the outdoors and have introduced new challenges to land managementagencies and researchers (Hendricks 1995) When these new interest groups share the sameresource with traditional recreationists the potential for con ict increases This study hashighlighted the complexities in understanding and predicting both out-group and in-groupcon ict The ndings reinforce some aspects of previous research raise questions aboutother predicted relationships and suggest other constructs worthy of future investigation

FIGURE 3 Snowboarder con ict model Only signi cant paths (p lt 05) are shown

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

SkierSnowboarder Con ict 305

TABLE 3 Snowboarder Indices for In-Group and Out-Group Con icts

Standardized CronbachrsquosStatement M factor loading SE t a a

Unacceptable snowboarder behavior (in-group con ict)b 83Fail to be aware of others 319 63 069 914Not adequate distance 306 73 067 1094Fail to yield right of way to downhill 313 75 066 1128

userBehaves in a discourteous manner 302 65 068 955Cuts others off 299 66 068 969Fails to yield to the less advanced 294 61 069 881

userUnacceptable skier behavior (out-group con ict)b 87

Fail to be aware of others 372 74 065 1140Not adequate distance 356 71 066 1083Fail to yield right of way to downhill 366 70 066 1052

userBehaves in a discourteous manner 340 74 065 1139Cuts others off 361 82 064 1284Fails to yield to the less advanced 342 66 067 993

user

aAll ts signi cant at p lt 001bVariables coded on a 5-point scale ranging from never (1) to almost always (5)

(Hypothesis 6) Both of these hypotheses were supported by the data (Table 5) The averagescores for the lifestyle tolerance index indicated that snowboarders were more likely thanskiers to perceive the two groups to be similar Compared with the snowboarders howeverthe skiers reported more unacceptable safety-related behaviors

Multivariate Analysis

The overall t of the skier and snowboarder models was assessed using ve indicators chi-square chi-squaredegree of freedom goodness-of- t index (GFI) comparative t index(CFI) and root mean square residual (RMR Table 6) Although both models produced asigni cant chi-square sample size tends to in ate this statistic Consequently Marsh andHocevar (1985) suggested that the chi-square should be evaluated in relation to the modelrsquosdegrees of freedom with a v 2 df ratio of 21 to 51 indicating an acceptable t This ratiofor both the skier ( v 2 df = 275) and snowboarder ( v 2 df = 168) models fell within thisrange Values for the GFI and CFI ranged from 91 to 93 also indicating an acceptable t forthe two models (Bollen 1989) Finally the RMRswhich measure the average discrepanciesbetween the observed and the model-generated covariances were less than or equal to 061for both skiers and snowboarders suggesting a close t of the data (Church amp Burke1994)

Skier Path ModelConsistent with Hypothesis 7 a signi cant and positive relationship between activity

style and out-group ( b = 023 p lt 05) and in-group ( b = 027 p lt 05) con ict wasobserved in the skier model4 (Figure 2) Resource speci city however did not signi cantly

306 J Vaske et al

TABLE 4 Snowboarder Indices for Activity Style Resource Speci city and LifestyleTolerance

Standardized CronbachStatement M factor loading SE ta a

Activity styleb 85Days per year snowboardc 482 72 064 1133Snowboards ownedd 229 76 060 1262Money invested in snowboardinge 286 61 065 941Number of years snowboarding f 226 76 063 1207Rating of snowboarding abilityg 352 84 058 1455

Resource speci cityh 76This area means a lot to me 379 78 067 1170Lots of my life is organized 292 67 069 967

around this areaThis area is best for what I like to do 313 53 072 736I identify strongly with this area 311 68 069 993

Lifestyle toleranceh Skiers and snowboarders have similar 75Lifestyles 293 60 072 839Education 327 52 074 701Income 312 69 070 980Attitudes toward the environment 273 50 074 673Feelings about the arearsquos value 364 68 070 976

aAll ts signi cant at p lt 001bBecause the items in the activity style index used different response scales all variables were

standardized before computing the indexcVariable coded on a scale ranging from 1ndash5 (1) to gt50 (8)d Variable coded on a scale ranging from 0 (1) to gt3 (8)eVariable coded on a scale ranging from 0 to 100 (1) to gt 3000 (8)f Variable coded on a scale ranging from 1 (1) to gt20 (8)gVariable coded on a scale ranging from beginner (1) to expert (8)hVariable coded on a scale ranging from strongly disagree (1) to strongly agree (5)

in uence beliefs about either out-group ( b = 008 ns) or in-group ( b = 006 ns ) unac-ceptable behavior as predicted by Hypothesis 8 Hypothesis 9 which predicted a positiverelationship between mode of experience and the two con ict constructs was only partiallysupported Similarly the predicted in uence of lifestyle tolerance on perceived con ictreceived only partial support (Hypothesis 10) In-group con ict increased for skiers whowere focused on their activity (b = 014 p lt 05) but the relationship between out-groupcon ict and mode of the experience was not signi cant ( b = 000 ns) Although sig-ni cant paths between the tolerance variable and out-group ( b = iexcl 028 p lt 05) andin-group ( b = 016 p lt 05) con ict were observed only the negative relationship waspredicted by theory Finally safety concerns (Hypothesis 11) in uenced out-group con ict( b = 042 p lt 05) but had no effect on beliefs about in-group unacceptable behavior( b = 010 ns )

Taken together three of the ve predictor variables (activity style lifestyle toler-ance safety) accounted for 44 of the variance in skiersrsquo reported unacceptable behaviorwith snowboarders (out-group con ict) Of these safety concerns had the largest relativein uence Three variables also explained skiersrsquo evaluations of con ict with other skiers

SkierSnowboarder Con ict 307

TABLE 5 Bivariate Analyses Comparing Skiers and Snowboarders

Skier SnowboarderVariable (M N = 383) (M N = 212) t p

Unacceptable snowboarder behaviora 320 305 258 010Unacceptable skier behaviora 304 356 890 001Activity styleb 445 411 322 001Resource speci cityc 295 324 423 001Mode of experienced 381 424 636 001Lifestyle tolerancee 279 315 616 001Safetyd 286 183 1047 001

aSix-variable index coded on a 5-point scale ranging from never (1) to almost always (5)bFive-variable standardized indexcFour-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)dSingle-variable coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)eFive-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)

(in-group) but it is important to note the following First although lifestyle tolerancesigni cantly in uenced in-group con ict the positive relationship was opposite that pre-dicted by theory Second the in-group model accounted for only 10 of the variance incon ict

Snowboarder Path ModelFigure 3 diagrams the ndings from the snowboarder path model5 Three variables (ac-

tivity style b = 022 p lt 05 resource speci city b = 023 p lt 05 lifestyle toleranceb = iexcl 033 p lt 05) signi cantly in uenced snowboardersrsquo beliefs about unacceptableskier behavior (out-group con ict) and accounted for 23 of the variance in the criterionconstruct All relationships were in the predicted directions Relative to snowboardersrsquo eval-uations of other snowboarders (in-group con ict) only activity style ( b = 040 p lt 05)had a signi cant in uence accounting for 21 of the variance

Overall these ndings are consistent with Hypothesis 7 but provide only partial supportfor Hypotheses 8 and 10 The mode of experience (Hypothesis 9) and safety (Hypothesis 11)relationships were not supported for either out-group or in-group con ict

TABLE 6 Goodness-of- t Statistics

Measure Skier model Snowboarder model

Chi-square 31866 19535Degrees of freedom 116 116Chi-squaredegrees of freedom 275 168Goodness-of- t index 92 91Comparative t index 91 93Root mean square residual 059 061

308 J Vaske et al

FIGURE 2 Skier con ict model Only signi cant paths (p lt 05) are shown

Discussion

Recent advances in technology such as snowboards and mountain bikes have changed theway people recreate in the outdoors and have introduced new challenges to land managementagencies and researchers (Hendricks 1995) When these new interest groups share the sameresource with traditional recreationists the potential for con ict increases This study hashighlighted the complexities in understanding and predicting both out-group and in-groupcon ict The ndings reinforce some aspects of previous research raise questions aboutother predicted relationships and suggest other constructs worthy of future investigation

FIGURE 3 Snowboarder con ict model Only signi cant paths (p lt 05) are shown

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

306 J Vaske et al

TABLE 4 Snowboarder Indices for Activity Style Resource Speci city and LifestyleTolerance

Standardized CronbachStatement M factor loading SE ta a

Activity styleb 85Days per year snowboardc 482 72 064 1133Snowboards ownedd 229 76 060 1262Money invested in snowboardinge 286 61 065 941Number of years snowboarding f 226 76 063 1207Rating of snowboarding abilityg 352 84 058 1455

Resource speci cityh 76This area means a lot to me 379 78 067 1170Lots of my life is organized 292 67 069 967

around this areaThis area is best for what I like to do 313 53 072 736I identify strongly with this area 311 68 069 993

Lifestyle toleranceh Skiers and snowboarders have similar 75Lifestyles 293 60 072 839Education 327 52 074 701Income 312 69 070 980Attitudes toward the environment 273 50 074 673Feelings about the arearsquos value 364 68 070 976

aAll ts signi cant at p lt 001bBecause the items in the activity style index used different response scales all variables were

standardized before computing the indexcVariable coded on a scale ranging from 1ndash5 (1) to gt50 (8)d Variable coded on a scale ranging from 0 (1) to gt3 (8)eVariable coded on a scale ranging from 0 to 100 (1) to gt 3000 (8)f Variable coded on a scale ranging from 1 (1) to gt20 (8)gVariable coded on a scale ranging from beginner (1) to expert (8)hVariable coded on a scale ranging from strongly disagree (1) to strongly agree (5)

in uence beliefs about either out-group ( b = 008 ns) or in-group ( b = 006 ns ) unac-ceptable behavior as predicted by Hypothesis 8 Hypothesis 9 which predicted a positiverelationship between mode of experience and the two con ict constructs was only partiallysupported Similarly the predicted in uence of lifestyle tolerance on perceived con ictreceived only partial support (Hypothesis 10) In-group con ict increased for skiers whowere focused on their activity (b = 014 p lt 05) but the relationship between out-groupcon ict and mode of the experience was not signi cant ( b = 000 ns) Although sig-ni cant paths between the tolerance variable and out-group ( b = iexcl 028 p lt 05) andin-group ( b = 016 p lt 05) con ict were observed only the negative relationship waspredicted by theory Finally safety concerns (Hypothesis 11) in uenced out-group con ict( b = 042 p lt 05) but had no effect on beliefs about in-group unacceptable behavior( b = 010 ns )

Taken together three of the ve predictor variables (activity style lifestyle toler-ance safety) accounted for 44 of the variance in skiersrsquo reported unacceptable behaviorwith snowboarders (out-group con ict) Of these safety concerns had the largest relativein uence Three variables also explained skiersrsquo evaluations of con ict with other skiers

SkierSnowboarder Con ict 307

TABLE 5 Bivariate Analyses Comparing Skiers and Snowboarders

Skier SnowboarderVariable (M N = 383) (M N = 212) t p

Unacceptable snowboarder behaviora 320 305 258 010Unacceptable skier behaviora 304 356 890 001Activity styleb 445 411 322 001Resource speci cityc 295 324 423 001Mode of experienced 381 424 636 001Lifestyle tolerancee 279 315 616 001Safetyd 286 183 1047 001

aSix-variable index coded on a 5-point scale ranging from never (1) to almost always (5)bFive-variable standardized indexcFour-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)dSingle-variable coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)eFive-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)

(in-group) but it is important to note the following First although lifestyle tolerancesigni cantly in uenced in-group con ict the positive relationship was opposite that pre-dicted by theory Second the in-group model accounted for only 10 of the variance incon ict

Snowboarder Path ModelFigure 3 diagrams the ndings from the snowboarder path model5 Three variables (ac-

tivity style b = 022 p lt 05 resource speci city b = 023 p lt 05 lifestyle toleranceb = iexcl 033 p lt 05) signi cantly in uenced snowboardersrsquo beliefs about unacceptableskier behavior (out-group con ict) and accounted for 23 of the variance in the criterionconstruct All relationships were in the predicted directions Relative to snowboardersrsquo eval-uations of other snowboarders (in-group con ict) only activity style ( b = 040 p lt 05)had a signi cant in uence accounting for 21 of the variance

Overall these ndings are consistent with Hypothesis 7 but provide only partial supportfor Hypotheses 8 and 10 The mode of experience (Hypothesis 9) and safety (Hypothesis 11)relationships were not supported for either out-group or in-group con ict

TABLE 6 Goodness-of- t Statistics

Measure Skier model Snowboarder model

Chi-square 31866 19535Degrees of freedom 116 116Chi-squaredegrees of freedom 275 168Goodness-of- t index 92 91Comparative t index 91 93Root mean square residual 059 061

308 J Vaske et al

FIGURE 2 Skier con ict model Only signi cant paths (p lt 05) are shown

Discussion

Recent advances in technology such as snowboards and mountain bikes have changed theway people recreate in the outdoors and have introduced new challenges to land managementagencies and researchers (Hendricks 1995) When these new interest groups share the sameresource with traditional recreationists the potential for con ict increases This study hashighlighted the complexities in understanding and predicting both out-group and in-groupcon ict The ndings reinforce some aspects of previous research raise questions aboutother predicted relationships and suggest other constructs worthy of future investigation

FIGURE 3 Snowboarder con ict model Only signi cant paths (p lt 05) are shown

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

SkierSnowboarder Con ict 307

TABLE 5 Bivariate Analyses Comparing Skiers and Snowboarders

Skier SnowboarderVariable (M N = 383) (M N = 212) t p

Unacceptable snowboarder behaviora 320 305 258 010Unacceptable skier behaviora 304 356 890 001Activity styleb 445 411 322 001Resource speci cityc 295 324 423 001Mode of experienced 381 424 636 001Lifestyle tolerancee 279 315 616 001Safetyd 286 183 1047 001

aSix-variable index coded on a 5-point scale ranging from never (1) to almost always (5)bFive-variable standardized indexcFour-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)dSingle-variable coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)eFive-variable index coded on a 5-point scale ranging from strongly disagree (1) to strongly

agree (5)

(in-group) but it is important to note the following First although lifestyle tolerancesigni cantly in uenced in-group con ict the positive relationship was opposite that pre-dicted by theory Second the in-group model accounted for only 10 of the variance incon ict

Snowboarder Path ModelFigure 3 diagrams the ndings from the snowboarder path model5 Three variables (ac-

tivity style b = 022 p lt 05 resource speci city b = 023 p lt 05 lifestyle toleranceb = iexcl 033 p lt 05) signi cantly in uenced snowboardersrsquo beliefs about unacceptableskier behavior (out-group con ict) and accounted for 23 of the variance in the criterionconstruct All relationships were in the predicted directions Relative to snowboardersrsquo eval-uations of other snowboarders (in-group con ict) only activity style ( b = 040 p lt 05)had a signi cant in uence accounting for 21 of the variance

Overall these ndings are consistent with Hypothesis 7 but provide only partial supportfor Hypotheses 8 and 10 The mode of experience (Hypothesis 9) and safety (Hypothesis 11)relationships were not supported for either out-group or in-group con ict

TABLE 6 Goodness-of- t Statistics

Measure Skier model Snowboarder model

Chi-square 31866 19535Degrees of freedom 116 116Chi-squaredegrees of freedom 275 168Goodness-of- t index 92 91Comparative t index 91 93Root mean square residual 059 061

308 J Vaske et al

FIGURE 2 Skier con ict model Only signi cant paths (p lt 05) are shown

Discussion

Recent advances in technology such as snowboards and mountain bikes have changed theway people recreate in the outdoors and have introduced new challenges to land managementagencies and researchers (Hendricks 1995) When these new interest groups share the sameresource with traditional recreationists the potential for con ict increases This study hashighlighted the complexities in understanding and predicting both out-group and in-groupcon ict The ndings reinforce some aspects of previous research raise questions aboutother predicted relationships and suggest other constructs worthy of future investigation

FIGURE 3 Snowboarder con ict model Only signi cant paths (p lt 05) are shown

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

308 J Vaske et al

FIGURE 2 Skier con ict model Only signi cant paths (p lt 05) are shown

Discussion

Recent advances in technology such as snowboards and mountain bikes have changed theway people recreate in the outdoors and have introduced new challenges to land managementagencies and researchers (Hendricks 1995) When these new interest groups share the sameresource with traditional recreationists the potential for con ict increases This study hashighlighted the complexities in understanding and predicting both out-group and in-groupcon ict The ndings reinforce some aspects of previous research raise questions aboutother predicted relationships and suggest other constructs worthy of future investigation

FIGURE 3 Snowboarder con ict model Only signi cant paths (p lt 05) are shown

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

SkierSnowboarder Con ict 309

As noted by Watson (1995) there has been little agreement regarding the de nitionof recreation con ict Jacob and Schreyerrsquos (1980) concept of goal interference is at leastpartially derived from motivation theory As de ned here con icts can also arise whengroups do not share the same norms (Ruddell amp Gramann 1994) In other words con ictmay occur when individuals perceive the behavior of others to be unacceptable Althoughthe ndings here are consistent with norm theory more work is needed to explore therelationship between traditional de nitions of con ict and the one used in this article

Contrary to past research (Adelman et al 1982 Gibbons amp Ruddell 1995 Watsonet al 1991 1994) data reported here do not support an asymmetrical relationship betweenuser groups Although skiers reported more unacceptable behavior with snowboarders thanwith fellow skiers snowboarders also identi ed more out-group than in-group con ict Assuggested earlier the visual differences in clothes language and on-slope behavior of theparticipants in these two activities may serve to magnify the potential for con ict

The analyses presented here were primarily based on variables suggested by Jacoband Schreyer (1980) 20 years ago These determinants along with other sources of con ict(eg safety) are still pertinent today The skiers and snowboarders in this sample variedin terms of activity style resource speci city mode of experience lifestyle tolerances andconcerns with safety The skiers for example attached more importance to the activitythan did the snowboarders Data reported by P W Williams et al (1994) indicated that theBritish Columbian skiers were more involved in their activity than were snowboarders onthe basis of years of participation ndings that probably re ect the recent emergence ofsnowboarding as an activity The snowboarders on the other hand were more involved onthe basis of frequency of participation per year The Colorado data produced exactly thesame pattern of ndings (see Tables 2 and 4) We hypothesized that when these items werecombined into a single activity style index the differences in years versus days would canceleach other out and there would be no differences between skiers and snowboarders relativeto activity style Contrary to the hypothesized relationship the results demonstrated activitystyle differences regardless of whether single-item or latent constructs are used Over timeif snowboarders continue to participate as frequently as noted here the importance theyattach to the activity may be greater than that reported by the skiers

The P W Williams et al (1994) study suggested few differences in resource speci- city between skiers and snowboarders leading to the hypothesis here that the two groupswould not differ on this concept In the Colorado data the snowboarders rated the resourcemore highly than the skiers did This difference in ndings between the two studies mayre ect the situational speci cs between the British Columbian and Colorado resorts Untilrecently relatively few Colorado ski areas allowed snowboarders Individuals who havebeen constrained from participating in their activity may place greater importance on theresource once such restrictions have been lifted

These ndings similar to those reported by Gibbons and Ruddell (1995) suggest ex-panding the scope of con ict research to include the notion of place attachment Placeattachment is typically operationalized using two conceptsmdashplace dependence and placeidentity (D R Williams amp Patterson 1999) Place dependence (a functional attachment)re ects the importance of the resource in providing amenities necessary for desired activ-ities This functional attachment is embodied in the arearsquos physical characteristics (eglength and steepness of ski trails) and can increase when the resource is close enough toallow frequent participation (Vaske amp Kobrin in press) Place dependence thus suggests anongoing relationship with a particular setting Place identity (an emotional attachment) onthe other hand is not a direct result of any one particular experience (Porshansky Fabianamp Kaminof 1983) but rather a psychological investment with the setting that has developedover time (D R Williams amp Patterson 1999) A history of repeat visitation due to place

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

310 J Vaske et al

dependence may lead to place identity (Moore amp Graefe 1994 Vaske amp Kobrin in press)Similar to the resource speci city hypothesis this suggests a relationship between placeattachment (dependencendash identity) and perceived con ict Theoretical and empirical workis needed in this area

Mode of experience a third determinant of con ict identi ed by Jacob and Schreyer(1980) was predicted here to be similar for skiers and snowboarders This hypothesis wasbased on ndings reported by P W Williams et al (1994) who found that skiers focusedon natural features of the environment whereas snowboarders focused on improving theirskills In essence the mode of experience for both groups was focused (rather than un-focused) but for different reasons As operationalized in this article mode of experienceemphasized the importance of developing snowboardingskiing skills Consistent with theBritish Columbia snowboarders the Colorado snowboarders were more focused on devel-oping their skills than were the skiers Unfortunately the data reported here do not permitan examination of the extent to which the skiers focused on other aspects of the experienceThis remains a topic for future study

The average scores for the lifestyle tolerance index indicated that snowboarders weremore likely than skiers to perceive the two groups to be similar Given that many of thesnowboarders (n = 194) were also skiers and only 13 of the skiers had snowboarded such ndings are as expected Research is needed however to understand this lifestyle toler-ance relationship with con ict Snowboarders on average tend to be younger than skiersConsequently some of the differences in tolerance observed here could be a function ofdifferences in age Similarly Thapa and Graefe (1998) showed differences both betweenand within skier and snowboarder groups on the basis of the participantsrsquo level of expertiseLow-skilled skiers and snowboarders experienced more con ict and were less tolerant thanhigh-skilled skiers and snowboarders

The two path models presented here suggest concepts that are more or less likelyto in uence beliefs about unacceptable behavior Of Jacob and Schreyerrsquos (1980) fourdeterminants of recreational con ict only activity style signi cantly in uenced both out-group and in-group con ict in both the skier and the snowboarder path models Resourcespeci city in uenced snowboardersrsquo perceptions of con ict with skiers (out-group) but hadno effect on con ict with other snowboarders (in-group) and did not enter either of theskier equations (out-group or in-group) Mode of experience predicted in-group con ict inthe skier model but not out-group con ict and had no effect in either of the snowboarderequations (out-group or in-group) For the skiers lifestyle tolerance predicted both out-group and in-group con ict but the positive path coef cient was opposite that predicted bytheory for the in-group con ict equation Finally safety only predicted skiersrsquo perceptions ofcon ict with snowboarders and had no in uence in the other three equations Taken togetherthese patterns of relationships highlight the complexity of predicting perceived con ict Itremains for future research to support or refute the generalizability of these ndings

The two path models also highlight the distinction between out-group and in-groupcon ict The snowboarder path model accounted for virtually the sameamount of variance inthe out-group (23)and in-group (21)equations Becausemany of the snowboarders werealso skiers the perception of out-group and in-group differences may have been minimizedThe skier path model however explained 44 of the variance in out-group con ict andonly 10 of that in in-group con ict These ndings may suggest that Jacob and Schreyerrsquos(1980) determinants of con ict are better suited to addressing con ict between activitiesrather than within activities Differences in activity style resource speci city mode ofexperience and lifestyle tolerancemay be minimalwithin a given activitybut relatively largebetween activities To address within group con ict research should focus on additionalfactors such as age gender and skill-level differences (Thapa amp Graefe 1998 1999)

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

SkierSnowboarder Con ict 311

The distinction between out-group and in-group con ict has implications for manage-ment For example in both the skier and the snowboarder path models lifestyle tolerancewas negatively related to perceived out-group con ict This suggests that physically sepa-rating the two activities by developing speci c trails or attraction areas (half-pipes) for eachgroup would help to reduce some of the negative interaction that occurs and may minimizesafety concerns

Coping with in-group con ict however raises a different set of issues to be resolvedIn part such problems may stem from skiersrsquosnowboardersrsquo level of expertise in theirrespective activities As noted above Thapa and Graefe (1999) found that relatively un-skilled skiers and snowboarders experienced more con ict than those with more expertiseThese results may be explained by novicesrsquo lack of experience in avoiding problems associ-ated with high-speed activities Designating trails for beginner intermediate and advancedskierssnowboarders and placing warning signs at trailheads about the dangers associatedwith more dif cult trails has helped reduce some of these problems Improved educationefforts emphasizing proper etiquette and behavior for both skiers and snowboarders mayserve to further minimize the potential for in-group con ict For example in response toincreasing complaints about snowboarder behavior from both snowboarders and skiers theCopper Mountain ski resort initiatedan education program calledldquoShrediquetterdquo (shreddingis slang for snowboarding) Along with the Skierrsquos Responsibility Code a variety of rulesand suggestions speci c to snowboarding were printed on brochures and distributed (Baird1993) These educational efforts address speci c issues and concerns so that recreationistsmay share the resource

Overall this study has attempted to identify the determinants of both out-group andin-group con ict among skiers and snowboarders The two structural equation modelshighlight the complexities involved in predicting the occurrence of reported unacceptablebehavior Although our ndings have helped to clarify some of the interrelationships amongthe variables the lack of previous reseach speci c to these two activities limits their gener-alizability Such a limitation however can only be addressed through future research

Notes

1 Ancillary analyses indicated that respondents from the ve different ski areas were similar onitems of concern to this article

2 Recognizing that recreationistsmay participate in multiple activities skiers were asked if theyhad ever snowboarded and snowboarders were asked if they had ever skied Each was then asked toindicate if they still participated in the other activity As the number of skiers that still snowboardedwas low (n = 13) and many (n = 194) snowboarders were once skiers all respondents were kept intheir assigned categories

3 When using LISREL the error variance must be assumed for single-item indicators of latentconstructs Following the recommendation of Hayduk (1987 pp 119ndash123) the error variance for themode of experience and safety concern constructs was set at 10 (reliability = 90) for all analysespresented here Use of other error variance values (05 15) provided similar solutions

4 The average correlation among the predictor variables in the skier model was 16 and rangedfrom 03 to 52 Only two of the correlations were above 15 The correlation between resourcespeci city and activity style was 33 and the correlation between safety and lifestyle tolerance wasiexcl 52 These ndings indicate that multicollinearity among the latent constructs was not an issue

5 In the snowboarder model the average correlation among the latent variables was 14 andranged from 01 to 41 Only three of the correlations were above 15 Similar to the skier analysesthe largest correlation was between safety and lifestyle tolerance (r = iexcl 41) Resource speci citywas correlated with activity style (r = 28) and with lifestyle tolerance (r = 26) ndings that againsuggest collinearity among the predictors was minimal

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

312 J Vaske et al

References

Adelman B J Heberlein T A amp Bonnicksen T M (1982) Social psychological explanations forthe persistence of a con ict between paddling canoeists and motor craft users in the BoundaryWaters Canoe Area Leisure Sciences 5 45ndash62

Aitkens M (1990) Have snowboard will soar The Physician and Sports Medicine 18 114ndash120Asher W amp Markels A (1992) When snowboards and skis collide Snow Country 5 22ndash23Baird W (1993) Recreation con ict between skiers and snowboarders Unpublished masterrsquos thesis

Colorado State University Fort CollinsBaron R S Kerr N L amp Miller N (1992) Group processes group decisions group actions

Belmont CA BrooksColeBlahna D J Smith K S amp Anderson J A (1995) Backcountry llama packing Visitor perceptions

of acceptability and con ict Leisure Sciences 17 185ndash204Bollen K A (1989) Structural equations with latent variables New York WileyCarothers P Vaske J J amp Donnelly M P (in press) Social values versus interpersonal con ict

between hikers and mountain bikers Leisure SciencesChavez D J (1999) Mountain bikingmdashA rapidly growing sport In K Cordell (Ed) Outdoor

recreation in American life A national assessment of demand and supply trends (pp 245ndash246)Champaign IL Sagamore

Church A T amp Burke P J (1994) Exploratory and con rmatory tests of the big ve and Tellegenrsquosthree- and four-dimensional models Journal of Personality and Social Psychology 66 93ndash114

Devall B amp Harry J (1981) Who hates whom in the great outdoors The impacts of recreationalspecialization and technologies of play Leisure Sciences 4 399ndash418

Finley B (1990 October 21) Snowboarding in safety ght Denver Post p P-1Gibbons S amp Ruddell E J (1995) The effect of goal orientation and place dependence on select

goal interference among winter backcountry users Leisure Sciences 17 171ndash183Hayduk L A (1987) Structural equation modeling with LISREL Baltimore Johns Hopkins Uni-

versity PressHendricks W W (1995) A resurgence in recreation con ict research Introduction to the special

issue Leisure Sciences 17 157ndash158Hoger J L amp Chavez D J (1998) Con ict and management tactics on the trail Parks and Recre-

ation 33 41ndash56Hughes K (1988 March 22) Surfboarding shifts to the ski slopes and cultures clash Wall Street

Journal p lIvy M I Stewart W P amp Lue C (1992) Exploring the role of tolerance in recreational con ict

Journal of Leisure Research 24 348ndash360Jackson E L amp Wong R (1982) Perceived con ict between urban cross-country skiers and snow-

mobilers in Alberta Journal of Leisure Research 14 47ndash62Jacob G R amp Schreyer R (1980) Con ict in outdoor recreation A theoretical perspective Journal

of Leisure Research 12 368ndash380Joreskog K G amp Sorbom D (1993) LISREL 8 Structural equation modeling with the SIMPLIS

command language Hillsdale NJ ErlbaumKnopp T B amp Tyger J D (1973) A study of con ict in recreational land use Snowmobiling versus

ski touring Journal of Leisure Research 5 6ndash17Lucas R C (1964) Wilderness perception and use The example of the Boundary Waters Canoe

Area Natural Resources Journal 3 394ndash411Marsh H W amp Hocevar D (1985) Application of con rmatory factor analysis to the study of self-

concept First and higher order factor models and their invariance across groups PsychologicalBulletin 97 562ndash582

Meyers C (1991 January 23) All aboard The battle over down time mounts between skiers andsnowboarders Denver Post p D-1

Moore R L amp Graefe A R (1994) Attachments to recreation settings The case of rail-trail usersLeisure Sciences 16 17ndash31

Porshansky H M Fabian A K amp Kaminof R (1983) Place identity Physical world and social-ization of the self Journal of Environmental Psychology 3 57ndash83

SkierSnowboarder Con ict 313

Ramthun R (1995) Factors in user group con ict between hikers and mountain bikers LeisureSciences 17 159ndash169

Ruddell E J amp Gramann J H (1994) Goal orientation norms and noise-induced con ict amongrecreation area users Leisure Sciences 16 93ndash104

Schneider I E (2000) Revisiting and revising recreation con ict research Journal of Leisure Re-search 32 129ndash132

Shelby B Vaske J J amp Donnelly M P (1996) Norms standards and natural resources LeisureSciences 18 103ndash123

Thapa B (1996) The role of tolerance in recreation con ict The case of adult skiers and snow-boarders Unpublished masterrsquos thesis Pennsylvania State University University Park

Thapa B amp Graefe A R (October 1998) Level of skill and its relationship to con ict and toleranceamong adult skiers and snowboarders Paper presented at the National Recreation and ParkAssociation Leisure Research Symposium Miami Beach Florida

Thapa B amp Graefe A R (1999) Gender and age group differences in recreational con ict and tol-erance among adult skiers and snowboarders (pp 219ndash226) In Proceedings of the 1998 North-eastern Recreation Research Symposium (Tech Rep NE-255) Radnor PA US Department ofAgriculture Forest Service Northeastern Research Station

Todd S L amp Graefe A R (1989) Level of experience and perception of con ict among canoeistson the Delaware River In Northeast Forest Experiment Station Proceedings of the 1989 North-eastern Recreation Research Symposium (Gen Tech Rep NE-132 pp 147ndash156) BurlingtonVT US Department of Agriculture Forest Service

Vaske J J Donnelly M P Wittmann K amp Laidlaw S (1995) Interpersonal versus social-valuescon ict Leisure Sciences 17 205ndash222

Vaske J J amp Kobrin K C (in press) Place attachment and environmentally responsible behaviorJournals of Environment Education

Vaske J J Shelby B Graefe A R amp Heberlein T A (1986) Backcountry encounter normsTheory method and empirical evidence Journal of Leisure Research 18 137ndash153

Watson A E (1995) An analysis of recent progress in recreation con ict research and perceptionsof future challenges and opportunities Leisure Sciences 17 235ndash238

Watson A E Niccolucci M J amp Williams D R (1994) The nature of con ict between hikers andrecreational stock users in the John Muir Wilderness Journal of Leisure Research 26 372ndash385

Watson A E WilliamsD R amp Daigle J J (1991) Sources of con ict between hikers and mountainbike riders in the Rattlesnake NRA Journal of Park and Recreation Administration 9 59ndash71

Watson A Zaglauer H amp Stewart S (1996) Activity orientation as a discriminant variable inrecreationcon ict research In Proceedings of the 1995 Northeastern Recreation Research Sym-posium (Gen Tech Rep NE-218) Saratoga Springs NY U S Department of Agriculture ForestService Northeastern Forest Experiment Station

White D (1990) Alpine systems come up to speed Snowboards Skiing 43 196ndash199WilliamsD R amp Patterson M E (1999) Environmental psychology Mapping landscape meanings

for ecosystemmanagement In H K Cordell amp J C Bergstrom (Eds) Integrating social sciencesand ecosystem management (pp 141ndash160) Champaign IL Sagamore

WilliamsD R amp Roggenbuck J W (October 1989) Measuring place attachment Some preliminary ndings Paper presented at the Symposium on Outdoor Recreation Planning and ManagementNRPA Symposium on Leisure Research San Antonio TX

Williams P W Dossa K B amp Fulton A (1994) Tension on the slopes Managing con ict betweenskiers and snowboarders Journal of Applied Recreation Research 19 191ndash213

Woodward B (1996) Sports Illustrated mountain biking The complete guide New York WinnerrsquosCircle Books