WPC  2 BJ c}urier3|os Roman ItalicHelvetica BoldXPTimes Romanmark Network PS LaserI4LASEPS.PRSx  @hhhhg2}X@"m^2CRddCCCdq2C28dddddddddd88qqqYzoCNzoozzC8C^dCYdYdYCdd88d8ddddCN8ddddY`(`lC2CC!CCCCCCCCCCd8YYYYYYzYzYzYzYC8C8C8C8ddddddddddYdddddodYYYYYYYdzYzYzYzYddddddddC8C8C8C8Ndz8z8z8z8z8ddddddCCCoNoNoNoNz8z8z8dddddddzYzYzYdz8dCoNz8dddddNF2[dCYddddd7>d<d<$YYdCCddooCYXhh4>>Im4>44hhhhhhhhhh>>mmmr}r4hr}}r}}}r>4>mh4hrhrh>rr44h4rrrrIh>rhhh]I4Im>4>>>>>>>>>>>>r4hhhhhh}h}h}h}h44444444rrrrrrrrr}hhrrr}hr}rhhhhhhhr}h}h}h}hrrrrrrrr44444444hhr4r4r4r4r4rrrrrrIII}h}h}h}hr>r>r>rrrrrr}hr]r]r]rr4rI}hr>}h}hrrrNA4hh>rhhhhhEDh>h>,]]h>>hhrr4]>}Times RomanTimes Roman ItalicHelvetica BoldHelveticaHelvetica ObliquehhhhYXPUSUK  X` hp x (#%'0*,.8135@8:@  U Table 1  Table 1  U 1  1  1  1  3'3'Standard6&A46&A4LaserI4LASEPS.PRSo\ } 2i 4 cX"m^ANoϨANN[ANAANN䨨A¨ܜNANA}r4hr}}r}}}r[A[NANN'NNNNNNNNNNAϨAAAAAAAAAAAAAAAAAAAAAܨ[[[NNNܶuuuA[NNRANVUÂNN8uuNNAuN"m^8C_oo8CCNu8C88ooooooooooCCuuuzÐz8ozzzC8Cuo8ozozoCzz88o8zzzzNoCzooodN8NuC8CC!CCCCCCCCCCz8oooooȲooooo88888888zzzzzzzzzoozzzozzooooooozoooozzzzzzzz88888888ooz8z8z8z8z8zzzzzzȽNNNoooozCzCzCzzzzzzozdzdzdzz8zNozCoozzzNF8ooCzoooooJIoCoC0ddoCCoozz8dC"m^2CTddCCCd2C28ddddddddddCCdzzzzCYozzdozzooN8NTdCddYdY8dd88Y8ddddNN8dYYYNP7PlC2CC!CCCCCCCCCCd8zdzdzdzdzdYzYzYzYzYC8C8C8C8dddddddddoYzddddoYdzdzdzdzdYYYYdzYzYzYzYddddddddC8C8C8C8YYo8o8o8o8o8ddddddzNzNzNdNdNdNdNo8o8o8ddddddoYoNoNoNdo8dzNdNo8oYoYdddNF2idNdddddd7>d<d<+oodCCddddCo<чnn8!BBnnnyyP7c1RyyXyycnnnѐ~nyRzczXzcyhCBnndhcnnonvyXzXshn~XyBBnss~|y~~~~~~~~~~~~~~~~~~~XXXXXXXyyyyyyyyyyyyyyyyyyyyBBBBBBBBBBBBnnnnnnnsssssssssssszCCn"^]o(]oo]o]]ooE];o]o]o]]](o]o]ooM8oooooooooo]M(]]]]]]]]]]]]]]]]]]]]]M;ooo;]oNu]o{zooOMooMM]Mo2k*X !XK%("^JY~JYYhJYJJYYJYJYJYJJJhYhKhYJYY -YYYYYYYYYYJ JJJJJJJJJJJJJJJJJJJJJ hhhYYYJhYN]JYcaYY? YY  J Y"m^88Goo,CCNu8C88oooooooooo88uuuo˅z8dozz888^o,oodoo8oo,,d,ooooCd8oddddC4CuC8CC!CCCCCCCCCCz8oooooȲdoooo88888888ooooooooodoozodoooooddddooooooooooooo88888,88ddo,o,o,o,o,ooooooȽCCCddddz8z8z8oooooodzdzdzdoo,oCdz8ddoooNF8koCzoooooJIoCoC&CCoCCoodd,CC`2p}wC>YJ 2p}wC YJhYN]JYcaYY? YY  J Y"^JY~JYYhJYJJYYJYJYJ;whKhYJYY -YYYYYYYYYYJ JJJJJJJJJJJJJJJJJJJJJ hhhYYYJhYN]JYcaYY? YY  J YzCCn"m^88Goo,CCNu8C88oooooooooo88uuuo˅z8dozz888^o,oodoo8oo,,d,ooooCd8oddddC4CuC8CC!CCCCCCCCCCz8oooooȲdoooo88888888ooooooooodoozodoooooddddooooooooooooo88888,88ddo,o,o,o,o,ooooooȽCCCddddz8z8z8oooooodzdzdzdoo,oCdz8ddoooNF8koCzoooooJIoCoC&CCoCCoodd,CC24[*Z,c.z1Times RomanTimes Roman ItalicHelvetica BoldHelveticaHelvetica ObliqueTimes Roman Bold}}r>4>mh4hrhrh>rr44h4rrrrIh>rhhh]I4Im>4>>>>>>>>>>>>r4hhhhhh}h}h}h}h44444444rrrrrrrrr}hhrrr}hr}rhhhhhhhr}h}h}h}hrrrrrrrr44444444hhr4r4r4r4r4rrrrrrIII}h}h}h}hr>r>r>rrrrrr}hr]r]r]rr4rI}hr>}h}hrrrNA4hh>rhhhhhEDh>h>,]]h>>hhrr4]>"m^**5SSd!22:X*2**SSSSSSSSSS**XXXSddlld\ul*KdS}luduld\lddd\***FS!SSKSS*SS!!K!}SSSS2K*SKlKKK2'2X2*222222222222\*dSdSdSdSdSlKdSdSdSdS********lSuSuSuSuSlSlSlSlSdKdSlSu\uSdKlSdSdSdSdSlKlKlKlKlSdSdSdSdSuSuSuSuSuSuSlSlS*****!**KdKS!S!S!S!S!lSlSlSlSuSuSl2l2l2dKdKdKdK\*\*\*lSlSlSlSlSlSldK\K\K\KlSS!lSl2dK\*dKdKlSuSlSN5*QS2\SSSSS87}}S2ooS}222S22SSKK!22"m^(1<YJ 2p}wC zKW!C(kSAC\  PChP,p>4Ip2p}wC06NA92p}wC CCn2O? c5r8 c/:s="m^(5APP|555PZ(5(,PPPPPPPPPP,,ZZZGtkktbYtt5>tbttYtkYbttttb5,5KP5GPGPG5PP,,P,|PPPP5>,PPtPPGM MW5(555555555555P,tGtGtGtGtGkkGbGbGbGbG5,5,5,5,tPtPtPtPtPtPtPtPtPtPtGtPtPtPtPtPYPtGtGtGkGkGkGkGtPbGbGbGbGtPtPtPtPtPtPtPtP5,5,5,5,>tPb,b,b,b,b,tPtPtPtPtPtPtk5k5k5Y>Y>Y>Y>b,b,b,tPtPtPtPtPtPttPbGbGbGtPb,tPk5Y>b,tPtPtPtPtPN8(HP5GPPPPP,2xxP0zzPx0GGP55PPYY5G0ZZXXr,Z55XXXr{rrZ``@Z,rO(Bn``{rrrrF{{{``iOXXXrrrtekX`BbObFbOt`wS55tXnXPtSgOtX{XYX_`FbFn\zStXn{neF`55X\\nec`reeeeeeeeeeeeeeeeeeeFFFFFFF````````````````````555555555555XXXXXXX\\\\\\\\\\\\nnnnnnnnnnnnnnnnnnnntbt5tn{5nX"m^**5SSd!22:X*2**SSSSSSSSSS**XXXSddlld\ul*KdS}luduld\lddd\***FS!SSKSS*SS!!K!}SSSS2K*SKlKKK2'2X2*222222222222\*dSdSdSdSdSlKdSdSdSdS********lSuSuSuSuSlSlSlSlSdKdSlSu\uSdKlSdSdSdSdSlKlKlKlKlSdSdSdSdSuSuSuSuSuSuSlSlS*****!**KdKS!S!S!S!S!lSlSlSlSuSuSl2l2l2dKdKdKdK\*\*\*lSlSlSlSlSlSldK\K\K\KlSS!lSl2dK\*dKdKlSuSlSN5*QS2\SSSSS87}}S2ooS}222S22SSKK!22"m^2CRddCCCdq2C28dddddddddd88qqqYzoCNzoozzC8C^dCtkktbYtt5>tbttYtkYbttttb`(`lC2CC!CCCCCCCCCCd8YYYYYYzYzYzYzYC8C8C8C8ddddddddddYdddddodYYYYYYYdzYzYzYzYddddddddC8C8C8C8Ndz8z8z8z8z8ddddddCCCoNoNoNoNz8z8z8dddddddzYzYzYdz8dCoNz8dddddNF2[dCYddddd7>d<d<$YYdCCddooCYAvBpqC"m^ )AAhN''-D ' AAAAAAAAAA DDDAvNNTTNG[T :NAaT[N[TNGTNnNNG 7AAA:AA AA:aAAAA': A:T:::''D' ''u''''''''''G NANANANANAuhT:NANANANA TA[A[A[A[ATATATATAN:NATA[G[AN:TANANANANAT:T:T:T:TANANANANA[A[A[A[A[A[ATATA  :N:AAAAATATATATA[A[AunT'T'T'N:N:N:N:G G G TATATATATATAnTN:G:G:G:TAATAT'N:G N:N:TA[ATAN) ?A'GAAAAA++aaA'VVAa'''Au''AAu::u'u'"m^*2GSSl*22:X*2**SSSSSSSSSS22XXX\lllld\ul*Sl\}luduld\lddd\2*2XS*S\S\S2\\**S*\\\\:S2\SuSSK:*:X2*222222222222\*lSlSlSlSlSlSdSdSdSdS********l\u\u\u\u\l\l\l\l\dSlSl\u\u\dSl\d\lSlSlSlSlSlSlSl\dSdSdSdSu\u\u\u\u\u\l\l\********SlS\*\*\*\*\*l\l\l\l\u\u\l:l:l:dSdSdSdS\2\2\2l\l\l\l\l\l\udS\K\K\Kl\\*l\l:dS\2dSdSl\u\l\N5*SS2\SSSSS87}}S2ooS}2$KKS22SS\\*K2a8DocumentgDocument Style StyleXX` `  ` a4DocumentgDocument Style Style . 2FkDk~DDvEa6DocumentgDocument Style Style GX  a5DocumentgDocument Style Style }X(# a2DocumentgDocument Style Style<o   ?  A.  a7DocumentgDocument Style StyleyXX` ` (#` 2I6F FG3Ha3DocumentgDocument Style StyleB b  ?  1.  a1DocumentgDocument Style Style Xqq   l ^) I. ׃  Times RomanHelvetica BoldTimes Roman ItalicHelvetica Bold ObliqueTimes Roman BoldTimes Roman Bold ItalicHelveticaHelvetica Oblique"m^88Goo,CCNu8C88oooooooooo88uuuo˅z8dozz888^o,oodoo8oo,,d,ooooCd8oddddC4CuC8CC!CCCCCCCCCCz8oooooȲdoooo88888888ooooooooodoozodoooooddddooooooooooooo88888,88ddo,o,o,o,o,ooooooȽCCCddddz8z8z8oooooodzdzdzdoo,oCdz8ddoooNF8koCzoooooJIoCoC&CCoCCoodd,CC2M"J K yL M"m^**5SSd!22:X*2**SSSSSSSSSS**XXXSddlld\ul*KdS}luduld\lddd\***FS!SSKSS*SS!!K!}SSSS2K*SKlKKK2'2X2*222222222222\*dSdSdSdSdSlKdSdSdSdS********lSuSuSuSuSlSlSlSlSdKdSlSu\uSdKlSdSdSdSdSlKlKlKlKlSdSdSdSdSuSuSuSuSuSuSlSlS*****!**KdKS!S!S!S!S!lSlSlSlSuSuSl2l2l2dKdKdKdK\*\*\*lSlSlSlSlSlSldK\K\K\KlSS!lSl2dK\*dKdKlSuSlSN5*QS2\SSSSS87}}S2ooS}222S22SSKK!22a1Right ParRight-Aligned Paragraph Numbers :`S@ I.  X(# a2Right ParRight-Aligned Paragraph Numbers C @` A. ` ` (#` a3Right ParRight-Aligned Paragraph Numbers L! ` ` @P 1. ` `  (# 2Q MNnO6Pa4Right ParRight-Aligned Paragraph Numbers Uj` `  @ a. ` (# a5Right ParRight-Aligned Paragraph Numbers_o` `  @h(1)  hh#(#h a6Right ParRight-Aligned Paragraph Numbersh` `  hh#@$(a) hh#((# a7Right ParRight-Aligned Paragraph NumberspfJ` `  hh#(@*i) (h-(# 2T8QRSSa8Right ParRight-Aligned Paragraph NumbersyW"3!` `  hh#(-@p/a) -pp2(#p Tech InitInitialise Technical Style. k I. A. 1. a.(1)(a) i) a) 1 .1 .1 .1 .1 .1 .1 .1 Technicala5TechnicalTechnical Document Style)WD (1) . a6TechnicalTechnical Document Style)D (a) . 2WQTTULVa2TechnicalTechnical Document Style<6  ?  A.   a3TechnicalTechnical Document Style9Wg  2  1.   a4TechnicalTechnical Document Style8bv{ 2  a.   a1TechnicalTechnical Document StyleF!<  ?  I.   2~]@WWLX/O\a7TechnicalTechnical Document Style(@D i) . a8TechnicalTechnical Document Style(D a) . PleadingHeader for numbered pleading paperP@n   $] X X` hp x (#%'0*,.8135@8:`a"m^'/BNN|e'//6R'/''NNNNNNNNNN//RRRVeeee]Vme'NeVuem]me]Ve]]]V/'/RN'NVNVN/VV''N'|VVVV6N/VNmNNF6'6R/'////////////V'eNeNeNeNeN|eN]N]N]N]N''''''''eVmVmVmVmVeVeVeVeV]NeNeVmVmV]NeV]VeNeNeNeNeNeNeNeV]N]N]N]NmVmVmVmVmVmVeVeV''''''''NeNV'V'V'V'V'eVeVeVeVmVmVe6e6e6]N]N]N]NV/V/V/eVeVeVeVeVeVm]NVFVFVFeVV'eVe6]NV/]N]NeVmVeVN1'NN/VNNNNN43uuN/ggNu/!FFN//NNVV'F/head3g 3heading 3=k<@ #:x2p}wC iX## o\  PC+XP#head2g 2Heading 2@?j0 #2p}wC9##o\  PC+XP#chaptheadChapter?@W'O X D#2p}wC9##o\  PC+XP#2e b!b"Xc#/dheadAChapter heading Ig0  X -#2p}wC ## o\  PC+XP#quoteslong quotes!b'Xhead4g 4heading 4"@Z? #:s2PkCXXP##o\  PC+XP#head1g 1heading 1#@0c?0[ #2p}wC ##o\  PCXP#2vc8ehXjcl"m^2CoddȧCCCdr2C28ddddddddddCCrrrdzNdzoȐC8CtdCdoYoYCdo8Co8odooYNCodddYO,OhC2CC!CCCCCCCCCCo8dddddȐYYYYYN8N8N8N8oddddooooddoddddzodddYYYYoYYYYddddddooN8N8N8N8do88888ooooddȐYYYoNoNoNoNCCCooooooȐdYYYo8oYoNCddodoNF2ldCddddddXhh4>>Im4>44hhhhhhhhhh>>mmmr}r4hr}}r}}}r>4>mh4hrhrh>rr44h4rrrrIh>rhhh]I4Im>4>>>>>>>>>>>>r4hhhhhh}h}h}h}h44444444rrrrrrrrr}hhrrr}hr}rhhhhhhhr}h}h}h}hrrrrrrrr44444444hhr4r4r4r4r4rrrrrrIII}h}h}h}hr>r>r>rrrrrr}hr]r]r]rr4rI}hr>}h}hrrrNA4hh>rhhhhhEDh>h>,]]h>>hhrr4]>"m^2NoddCCCdr2C28ddddddddddCCrrrdNdzzozzzC8CrdCddYdYCdo88d8odddNN8oYdYNF,FrC2CC!CCCCCCCCCCd8dddddYYYYYN8N8N8N8oddddoooozYddddzYdzddddYYYYdYYYYddddddooN8N8N8N8ddz8z8z8z8z8ooooddNNNoNoNoNoNz8z8z8oooooozYzNzNzNdz8oNoNz8zYzYddoNF2ddNdddddd54.(p2x4vC06NA.w2x4vC~FBi2s`vab*b|z:7oC2+o\  PCXP 4 F5nC2 Ӯn*f9 xCXXG6NA92p}wC HNo]2p}wCI>YJ 2p}wC 1G;`2p}wCJ>YJ 2p}wC zKW!C(kSAC\  PChP,p>4Ip2p}wC0L6NA92p}wC zt,Y5( gY\  PCP M7oC2+o\  PCXP N/xC8i:x2p}wCXDO1sC8.X:s2PkCXPP1sC88n:s2x(CXX7tC2J?t4  p(ACXz:7oC2+o\  PCXP 4 F5nC2 Ӯn*f9 xCXXG6NA92p}wC HNo]2p}wCI>YJ 2p}wC 1G;`2p}wCJ>YJ 2p}wC zKW!C(kSAC\  PChP,p>4Ip2p}wC0L6NA92p}wC zt,Y5( gY\  PCP M7oC2+o\  PCXP N/xC8i:x2p}wCXDO1sC8.X:s2PkCXPP1sC88n:s2x(CXXQ7tC2J?t4  p(ACXDRq%V2*.䡬V2PkCPSs#Z2*-Z2p}wCq%V2*8=V2x(CX2dZ z:7oC2+o\  PCXP 4 F5nC2 Ӯn*f9 xCXXG6NA92p}wC HNo]2p}wCI>YJ 2p}wC 1G;`2p}wCJ>YJ 2p}wC zKW!C(kSAC\  PChP,p>4Ip2p}wC0L6NA92p}wC zt,Y5( gY\  PCP M7oC2+o\  PCXP N/xC8i:x2p}wCXDO1sC8.X:s2PkCXPP1sC88n:s2x(CXXQ7tC2J?t4  p(ACXDRq%V2*.䡬V2PkCPSs#Z2*-Z2p}wCTq%V2*8=V2x(CX6pC2[kp9 xICX>YJ 2p}wC 1G;`2p}wC  X/ De+chapthead#2p}wC9#Chapter 3 Ochapthead #o\  PC+XP#ۃ  o  gheadA9e*#2p}wC#METHODOLOGY headA#o\  PC+XP#ۃ #0chead1  _ #2p}wC #3.1 Introduction#0[head1#o\  PC+XP# The previous chapter has reviewed the literature on recreational conflicts and the theoretical perspectives which help in their understanding. I argued that the development of recreational conflict is influenced both by environmental factors, such as increasing use, changing use patterns, proximity of incompatible groups, and by socialpsychological factors, such as how individuals interpret the situation. These different influences are best explored using a range of methods. This chapter describes and discusses the methods that were used in this research in the light of these epistemological assumptions. The process of analysis is also explained to elaborate the strengths and possible weaknesses of the research process. #0chead1  _" #2p}wC #3.2 Epistemology and Ontology]#0[head1#o\  PC+XP# Recreational conflict has previously been studied from a positivistic perspective using selfadministered questionnaires as a primary source of data. Positivists aim to find out  X' about the world using objective measurement of direct sensory experience. In a positivist  X) epistemology, truth in the social sciences is singular and external to the individual (Henderson, 1990:169). This perspective, however, does not address adequately the role of the researcher in the research process, nor does it acknowledge the cultural context in,Z0.. which research is carried out (Berno, 1995). Knowledge is a human construction, thus it changes, changing our view of the world in the process. Our theories about how the world works, not only come out of the observations that we make and the experiences  X that we have, they also influence what we see and how we experience an event. Language and culture are also collections of symbols or theories that shape our interpretation of the world. We name things that have meaning for us. That there are words in other languages with no equivalent in English shows differences in the importance of those named objects in different cultures. The English language, like other Germanic languages, separates subjects and objects, which fosters the impression that the world is made up of discrete entities (Pirsig, 1974), thus it is easier for English speakers to see a researcher as separate from the research. It is not surprising, that positivism has been a strong force in our approach to the sciences, and that scientific thought originated in cultures with similar linguistic structures. head2  W #2p}wC9#3.2.1 A Humanist Frameworkj0head2#o\  PC+XP# Humanistic approaches broaden the world view of positivism because they account for the role of human perceptions and interpretations. Accepting the role of the researcher in research does not preclude the use of the objective techniques of the natural sciences. Instead it highlights both the limitations and advantages of an objective science. Humanism, and more specifically, idealism, permits an appreciation of the way human beings construct a view of the world, which then influences their actions in that world. It acknowledges different types, or ways, of knowing, thus providing a better framework from which to investigate the world, particularly in the social sciences. Belenky, Clinchy, Goldberger and Tarule (1986) distinguish a number of different ways  X* of knowing. However, of particular interest in the research context are separated  X+ knowing and connected knowing. Separated knowing is the type of knowing that arises from the use of objective techniques, where the researcher is seen as being separatedb-Z0.. from the subjects of study. The researcher imposes her/his own interpretations on the  X data, thus the separated knowing implies some level of mastery over the actors involved. In comparison, connected knowing, or understanding, implies;  V3 !bquotespersonal acquaintance with an object (usually but not always a person). Understanding involves intimacy and equality between self and object, while knowledge implies separation from the object and mastery over it. Understanding  X ... entails acceptance ...p!'quotes (p.101)" and therefore some accommodation of another person's perspective. Ethically, when one's research involves people, it should be aimed at connected knowing and  X understanding rather than separate knowing and mastery over the actors. In the human sciences, those under study are influenced by the same processes of language, culture and interpretation as the researcher. The interpretations of respondents, or the way they construe a situation, will direct their actions (Ross and Nisbett, 1991). A connected  Xk approach is, in some respects, more objective because the subjects of the study explain their own actions, while in a positivist approach the researcher imposes her/his own  X meaning on the recorded actions of others (Belenky et al., 1986). A humanist framework helps us see that no research methods are free of the imperfections and limitations of the human condition. It is very unlikely that we will  XE ever know the absolute truth. Such a thing may be just another figment of our language. This does not excuse sloppy research techniques, however, it does allow us to look at  X# research from a pragmatic perspective. Truth is that which helps us understand the  X6% world better, and objectivity is achieved when another researcher could be expected to elicit similar data using the same methods (Simmons, 1985; Smith, 1984).  X#* Pragmatists see theory from an instrumental perspective. A theory does not represent truth, but instead represents ideas that best fit the facts as they are currently known (Smith, 1984). Therefore, using a range of theoretical perspectives is better thanm-Z0.. adhering to any single perspective. Denzin (1989) takes a pragmatic approach to  X research when he recommends triangulation of methods, theoretical approaches and also of researchers as a means to obtain the best possible result from research. As argued in the previous chapter, conflict occurs only when the actors construe themselves as being in conflict. Different actors in the same situation may see it in a variety of ways and thus feel very differently about it (Ross and Nisbett, 1991). Conflict depends on an individual's feelings and perceptions of what is going on. Similarly, the satisfaction gained from a recreational activity depends on the perceptions of the participants. Understanding conflict in any recreational setting, therefore, requires  Xb methods which allow the researcher to understand how, when and why respondents feel negatively about a recreational setting. However, the development of conflict is also influenced by factors in the wider social and physical environment. A more separated approach may be useful for exploring these factors. head2  W #2p}wC9#3.2.2 Quantitative Methodsl"j0head2#o\  PC+XP# Survey questionnaires are best suited to testing theory that has already been formulated (Glaser and Strauss, 1967), and therefore they are useful for investigating descriptive questions. One of the aims of this research, for example, was to look at the differences between mountainbikers and trampers. Past research indicates that descriptions of gender, age, education, participation in other activities, and motivations may be of interest (Manning, 1986). Survey methods allow for a large number of people to be asked these simple questions. Analysis of survey data can also draw attention to patterns and links that exist within the sample population, and thus may give clues to changes occurring at a macrosociological level. However, quantitative methods have a number of weaknesses. Complex communication is impossible in a survey questionnaire. Not only does one have to account for the+Z0..  X reading ability and aptitude of respondents, but both questions and answers can be interpreted differently from what either writer originally meant. Closed (tickthebox) questions elicit opinions that people may not have held previously.  X In comparison, open questions may result in shallow and off the top of the head answers that do not reflect what that person really feels (Babbie, 1989). In my own experience, open questions are answered much less often than closed questions, possibly because people are not inclined, or able, to frame short answers to what might be complex questions. Neither type of question can be rephrased or clarified when used in selfadministered questionnaires, and nuances of context and qualification cannot be addressed adequately. The construction of the questionnaire will influence answers to questions that vary with context. Open questions may elicit relatively few answers, where closed questions influence the content of those answers. As Babbie (1989:255) notes:  V !bquotesSurvey research is generally weak on validity and strong on reliability.?+!'quotes" Closed questions are reliable because, given the same range of closed answers, a similar spread of replies could reasonably be expected over and over again. However, if the replies change with the range of offered answers, then the validity of the results must be in question. If the choice must be between open questions (that elicit shallow answers), and closed questions, then it seems that the researcher is caught in a double bind. Of course, if the researcher is interested only in finding out how individuals within the population are spread across a given range of options, then the results of closed questions will result in valid data. Validity, therefore, is heavily reliant on the context of what is already known and of the question the researcher is asking. Survey data are unlikely to reflect how people feel, think and act with respect to recreational settings, especially if the questions are about experiences, or things which are9,Z0.. not normally put into words. If a person is being questioned about their motivations for a particular recreational activity, for example, it may take considerable thought to frame an answer. In fact, it seems that quantitative methods are most suited for developing the  X separated knowing of Belenky et al. (1986), but are unsuitable for developing connected knowledge or understanding. Neither is survey research an effective way to generate new theories or to understand the subjective worlds of the actors in the research setting. The subjective nature of recreational experience implies that the subjective experiences of the actors are important in recreation research. head2  W #2p}wC9#3.2.3 Qualitative Methods2j0head2#o\  PC+XP# Qualitative methods are more suited to generating theory and understanding meaning in a subjective context. Indepth interviews and participant observation allow researchers to investigate complex issues and to develop connected understanding. Communication that encourages intimacy and conveys personal meaning is necessarily a two way process that takes considerable time and is possible only in unstructured or semistructured settings. With time to get to know the respondents, it is possible to clarify points, be sure of meanings and to ask the same question many times in different contexts so as to understand how each individual qualifies her/his answers. Indepth interviews allow a depth of enquiry, but the researcher cannot effectively question a large number of people because of the time required. The indepth approach over a large number of respondents would seem to be the best possible method, however it is debatable whether the results of such a study would be any more reliable and valid than one using a range of methods. As the research progresses, one begins to hear the same things from respondents. It is this repetition that forms the basis of the findings. As the interviews become more repetitious, it becomes difficult to listen carefully enough to justify the large amounts of time spent in indepth interviewing. Thus, the nature of the communication in indepth interviews limits the ability of the researcher to contact a large sample in a consistently meaningful way. Z-Z0..ԌOnce the stories become familiar through repetition, methods such as participant observation, searching the popular media (e.g. newspapers and cycling magazines) and carrying out survey research become more useful as one searches for confirmation that these stories are similar throughout the wider population. head2  Wu #2p}wC9#3.2.4 Section Summary:j0head2#o\  PC+XP# Indepth interviews provide very high internal validity over a small part of the population, while survey questionnaires provide high external validity (or reliability). Indepth interviews, allow researchers to explore meaning and context in detail. Because of the open nature of the communication, qualitative research may allow the development  X of new theory or what Glaser and Strauss (1967) call grounded theory. Grounded theory is formulated inductively from the interview data while, in comparison, survey research uses logical deduction and hypothesis testing to test existing theories. Survey research and indepth interviewing thus have different and complementary functions at the theoretical level. This study is effectively pragmatic in nature, assuming that no single approach to research methods will be adequate in itself. Each method complements the others in some way, so that by using a range of methods the social researcher can generate data that is more reliable and valid than if they use only one method. Denzin (1989) advocates this triangulation of methods as a way of eliminating some of the shortcomings in research. The remainder of this chapter outlines the methods that were used, the way that they were used and some of the problems that were encountered. #0chead1  _& #2p}wC #3.3 Data Collection@#0[head1#o\  PC+XP# Indepth interviews, participant observation, focus groups and survey methods were utilised in this research. This section is an account of the way in which the methods were used. It also includes some of my observations on the relative strengths and weaknesses of the methods for collecting data.- Z0..Ԍhead2  W #2p}wC9#3.3.1 In Depth Interviews-Cj0head2#o\  PC+XP# Twenty seven semistructured interviews were conducted with 32 people (five interviews involved two people) over the course of the study period which ran from December 1991 to March 1993. Most of them were undertaken in early 1992. Interviews ranged in length from 45 minutes to two hours with most lasting one to one and a half hours. Most respondents chose to be interviewed in their homes, however, this was negotiable because it was important that they felt comfortable. In one case, the respondent preferred to meet in a cafe and in two other cases the interviews were at the interviewees' places of work. In every case, the interviews were recorded on tape for later transcription. Interviewees were asked if they would allow me to record the conversation on tape, and noone refused. However, on two occasions it was obvious that the respondent was uncomfortable with the taperecorder going and, on those occasions, the conversation was much easier after the recorder was turned off. Two interviewees requested that the tape recorder be turned off during certain parts of the conversation, and in some cases, the interviews ran for longer than the available tape. If there were breaks in the tape, or if the conversation continued after the tape was turned off, I made notes to record the direction of the conversation. As soon as possible after the interview, I sat down and recorded my thoughts either on  X# tape or in writing (whichever was more convenient). These notes on notes enriched the data of the interview by noting any interesting ideas that had come to me during the course of the interview, and any other relevant feelings or facts that came to mind before, during or after the interview. Questions that came to mind after an interview were also recorded here and used in subsequent interviews. Thus, the interview process was constantly evolving. My final interview was very different to my first one because of the+ Z0.. ongoing process of analysis, and the development of my own theories throughout the course of the research. Interviewees were contacted in a range of ways. My earlier contacts were made at public meetings that had been called by the Canterbury Department of Conservation to discuss aspects of mountainbiking in the Canterbury area. These included both tramperonly respondents (that is, those trampers or walkers who had no experience mountainbiking) and mountainbikers. One or two early interviews were conducted with people I met incidentally and who expressed strong views on the subject of mountainbiking.  X Further contacts were made using the snowballing technique outlined in Lofland and Lofland (1984). At the end of each interview, respondents were asked if they knew anybody that I should talk to. Invariably, I received a list of names and phone numbers. Other interviewees were chosen from the many questionnaire respondents who had indicated they would be prepared to be interviewed. Respondents were chosen to maximise the range of age, experience and attitude represented in the interview sample. Interviewees ranged in age from 15 years to over 70 years. They included mountainbikers and trampers with a wide range of experience in their respective activities, people who participated regularly in both activities, people with a wide range of opinions about the advent of mountainbiking, and managers of the recreation resource. In general, I contacted one respondent at a time to arrange an interview but sometimes partners or friends that were present also listened in and contributed. Sessions with more than one interviewee proceeded in a number of different ways. Sometimes a partner would not join in the conversation except to add to an answer or to challenge the accuracy of a statement the interviewee made. In other interviews, both answered the questions on an equal basis. At all times, when there was more than one person present, interview respondents appeared more relaxed and the interviews lasted around two hours. 5, Z0..ԌThe research included two focus group sessions. In both cases the participants knew everyone else in the group, hence these groups were excellent for developing ideas and exploring more complex issues. Whenever there was more than one interviewee present, it was easier to elicit descriptions of feelings and motivations, and to compare the experience of walking with the experience of riding. These are things that many people were not used to explaining, and having a friend or friends with whom they could try out ideas appeared to help them put words to experiences with very little input from myself. This interaction highlights one of the difficulties of indepth interviewing. It is very easy for the researcher to inadvertently ask loaded questions or to put ideas into the heads of those being interviewed. While there are many ways of asking questions which avoid these problems, it is much easier if the discussion takes its course with a minimum of input from the interviewer. During focus group sessions, individuals in the group often questioned the assertions of others in ways that an interviewer could not afford to do! Additionally, new questions and perspectives were raised to provide many new insights that could not have come from interviewing each individual in turn. The chance of the interviewer invalidating data by asking a leading question was also decreased in these situations. I was constantly impressed with the time people were prepared to put into these interviews, which contrasted sharply with the time they were prepared to spend answering a questionnaire. In asking people to fill in questionnaires, through watching people filling in questionnaires and through talking to other researchers about their experiences in similar situations, I noticed that people often appear uncomfortable when they cannot qualify an answer or are unsure what a question is asking. Because  X* understanding requires equality with people rather than mastery over them, it seems that individuals might feel less threatened when they feel understood. This difference may be5, Z0.. one of the reasons that respondents seemed happier to talk face to face for two hours rather than to fill in an impersonal questionnaire which took only 15 minutes to complete. It seems that far from being unwilling to impart information, respondents are happy to talk and to spend much time doing so, if they feel understood. The decision to stop indepth interviewing came when it seemed that the stories I was hearing, and the theories I was developing, had remained the same for several interviews and for that reason, it was becoming more difficult to listen to each respondent with the appropriate intensity. It therefore seemed more profitable to look at data from other sources. head2  WG #2p}wC9#3.3.2 Participant Observationk_j0head2#o\  PC+XP# Participant observation was used as a complement to the indepth interviews and the questionnaires. Observations were done in a range of settings: such as attending evening meetings and day trips with the Canterbury MountainBike Club, and walking on some of the Port Hills tracks. I also went along to public meetings that were discussing mountainbike use and talked to many people informally, often at unexpected times. Conversely, when I was out looking for people to talk to about mountainbiking, there was sometimes noone there! Thus, I was unable to separate the research process from other aspects of my life. At no time did I ever hide my role as researcher, but neither did I tell people without their asking me. I was often thrown unexpectedly into the researcher role when people found out that I was studying conflict between bikers and trampers, and wanted to get their view across. Participant observation proved particularly useful for consolidating information gathered from interviews and questionnaires and for questioning my interpretation of that information. For example, it was my conversations with people at a range of social occasions that convinced me that the stories I was hearing during my interviews were reflected throughout a much larger portion of the population than I had managed toV- Z0.. interview formally. It was thus the continuous reiteration of interviewees' stories during  X these casual conversations that convinced me that it was time to leave the field and focus my attention on analysing the data. head2  Wy #2p}wC9#3.3.3 Questionnairesfj0head2#o\  PC+XP# Questionnaires were pretested by students at Lincoln University and by 31 members of the Canterbury MountainBike Club who attended a monthly meeting early in 1993. As a result of the pretest, some of the questions were rephrased or simplified. No question was removed. In order to try to obtain a good response rate, the questionnaire was kept as short and as simple as possible. In general it seemed to take respondents somewhere between 10 and 30 minutes to complete. Questionnaires were sent both to mountainbikers and trampers. Obtaining a sample that would not affect any of the variables in the questionnaire was especially difficult for the mountainbiker group. None of the bike shops I contacted kept address lists of those who have bought bikes from them. The lists some shops did have were not freely available and, additionally, were compiled from race lists, so using them would affect estimates of offroad use and race participation. Similarly, club riders are, by definition, offroad riders which would affect estimates of offroad use and of race participation. In addition, the mountainbike owners I talked to informally, who did not ride offroad, did not identify themselves as mountainbikers and therefore did not feel qualified to answer a questionnaire about mountainbiking. This last observation makes it seem unlikely that even a random population survey would provide an accurate estimate of offroad use as a percentage of mountainbike owners. Eventually, bikers were contacted though the Canterbury MountainBike club (46 sent),  X+ through handing out 100 questionnaires at The Big Coast (an offroad recreational ride, organised by Tourism Resource Consultants) in Wellington in March 1993, and 72^-Z0.. questionnaires were handed out to students standing in registration queues at Lincoln University in 1993. Sampling of trampers was easier because they are more organised in clubs and so I sent questionnaires out through tramping clubs. In all, 400 questionnaires were sent out to trampers. Many clubs were not willing to release their address lists but were prepared to put questionnaires in with their newsletters. Additionally, one tramping club initially refused to send out my questionnaires because they believed that they did not have enough mountainbikers in their club to warrant it. They reversed their decision when I explained the questionnaire was specifically designed to find out what trampers thought about mountainbiking. This story highlights the need to design the questionnaire cover so that potential respondents did not immediately discard them thinking they were designed for mountainbikers. Wherever possible, a reminder notice was sent out seven to ten days after the questionnaire. Because I did not always have access to people's addresses, it was not always possible to send out the reminders, and this was reflected by the response obtained from the various samples. head2  W_ #2p}wC9#3.3.4 Response Rate\sj0head2#o\  PC+XP# Overall, the response rate was 71 percent. Response rates across various sample groups (see table one) indicate that reminder notices have a significant effect (Chi square = 91, pearson's rho =<< 0.001). Reminder notices were not sent to the Hospital Tramping Club, the Catholic Tramping Club, the Big Coast, all of which show significantly lower return rates than those of the other samples (see RESPONSE RATESTable I). The Lincoln sample shows a high return rate even though there was no follow up, however in this sample, people were handed questionnaires in person after being asked whether they had a mountain X+ bike. Anyone who did not want to fill in a questionnaire could have said no to the initial+Z0.. question. These people were also in student registration queues, and had much time to fill in after they were given the questionnaires, hence the high response rate.  Table 1 # Table 1 !#pddd,d X Ԏ ^ ddx !ddx &&& ^     Sample Number sent Reminder Notices? Returns Response Rate  q  Over Forties TC 100 yes 77 77%q q  Peninsula TC 77* yes 69 90%q q  Christchurch TC 80 yes 57 71%q q  Women Outdoors, N.Z.n 29n yesn 23n 79%q q  Canterbury Mt. Bike Club 46 yes 38 83%q q n Card (tramper questionnaires)P 34P yesP 28P 82%q q   Card (biker questionnaires) 43 yes 38 88%q q P  Hospital TC2 502 no2 292 58%q q   Catholic TC 30 no 14 47%q q 2 Lincoln University 72 no 54 75%q    The Big Coast 100 no 43 43%    * eighty originally sent but three returned unopened. $""F"" !"$RESPONSE RATES As the results came in, it appeared that the sample may have been skewed in favour of older people answering the tramper questionnaire and younger ones answering the biker questionnaire. In order to check whether this was due to sampling error, freepost cards, containing a small number of questions, were placed in a variety of huts in Arthur's Pass and on the St. James Walkway. Again, this sample was not random, as respondents self selected according to their willingness to fill in, carry out and post the cards. It seems unlikely that individuals would select themselves in such way as to influence the age distribution of the sample. As it turned out, the relative age distribution of bikers and nonbikers in the card sample were similar to that represented in the questionnaire samples. Further questionnaires were sent out to card respondents who said they would be prepared to fill in a questionnaire. Extensive informal discussions with bike shop0,Z0.."C!0 owners, trampers, mountainbikers and all my participant observations also confirmed the significant age difference between mountainbikers and walkers who do not mountainbike. #0chead1  _ #2p}wC #3.4 Data Analysis$#0[head1#o\  PC+XP# head2  W #2p}wC9#3.4.1 Analysing Questionnaire Dataj0head2#o\  PC+XP# In the results that follow, there are a number of different sample sizes and sample names. The two different questionnaires had some common questions for trampers, who also owned a mountainbike, and a large group of respondents participated in both mountain X biking and tramping. Most of my figures distinguish only between bikers and nonbikers (people who said they had been mountainbiking and people who said that they had not).  X A smaller group of figures distinguish instead between trampers and bikers which are, respectively those that filled in a tramper questionnaire and those that filled in a biker questionnaire. As a result, some of the biker samples contain around 173, while others contain around 217, depending on whether the questions were contained in both questionnaires (and whether they were answered adequately by respondents). Questionnaires were analysed using SPSS (statistics package for the social sciences). Open questions were coded and entered into a spreadsheet, while closedended questions were entered directly into the same spreadsheet with no processing. Many respondents answered only the closed questions. SPSS generated frequency tables, or crosstabulations and Chi square was used to test for significant differences. Wherever a result  X$ is indicated as significant in this thesis, it at most has a probability of 0.01 of occurring  X& accidentally (i.e. at least 99% confidence levels). head2  W%) #2p}wC9#3.4.2 Analysing Qualitative Data j0head2#o\  PC+XP# The indepth interviews were transcribed verbatim along with the additional notes on notes made by the researcher at the time of the interview. Notes taken during participant,Z0.. observations were also typed up along with any comments, insights and thoughts that  X arose from the observation. These notes on notes gradually grew with each interview and provided the themes that make up a significant part of the results section of this thesis. Each interview and observation was coded according to these general themes, and then filed using NUDIST a computer package designed for analysing textual data. NUDIST allows the researcher to reference and cross reference interview data into whatever themes seem useful. From here the data can be further divided up into subthemes if appropriate. NUDIST replaces the manual methods of cutting up printed matter and pasting it under the chosen headings. #0chead1  _ #2p}wC #3.5 LimitationsĒ#0[head1#o\  PC+XP# The method of sampling used in this research was purposive, rather than random, and some of the samples, particularly those of mountainbikers may have skewed the questionnaire results. Club riders, for example, were mostly offroad enthusiasts, while the riders on the Big Coast would also have just finished an offroad ride, thus these two samples would probably make my estimates of offroad use higher than those for the general population. Similarly, the Lincoln students were likely to be younger than the general population of mountainbikers, and thus may affect the overall age spread of mountainbikers. However, there were also mountainbikers in the sample of trampers that I used, so that, for those questions that were common to both questionnaires, there was an additional sample with a potentially different background. Relying on tramping clubs also may have affected some of the results. Club members may have had more exposure to opinions about mountainbikers than those walkers who go out individually and who, therefore, may not discuss the matter with other walkers. Many of the clubs were involved in submissions on mountainbiking, therefore, it is possible that club members might have stronger views on the subject than nonmembers. These sampling problems will be discussed further in the results chapters. -Z0..ԌThe main focus of this research was on gaining a profile of mountainbikers (that is those bikers that take their bikes offroad) and understanding the conflict between the two groups. For these two purposes the sampling is seen to be adequate. The study includes 217 offroad riders and therefore should have built up a good, general picture of these people. Similarly, in investigating the conflict, the views of those who are most vocal and most prepared to answer questionnaires are especially relevant. Additionally, questionnaires were answered by a wide range of people, including respondents who regularly participated in both activities, respondents who appeared to be unconcerned about mountainbiking, as well as respondents who felt strongly on both sides of the conflict. There may be some question as to whether the sample is representative of the relative number of people with the various standpoints in the general population. However, these results were not contradicted by any of the methods used. Each of the methods confirms the findings of the others where similar questions were investigated. Additionally, much informal conversation over the last three years in a wide variety of situations indicates  X0 that these figures are indicative of the population as a whole and are therefore reliable and valid. In retrospect, it would have been better to have sent out the same questionnaire to all groups with some sections that were for mountainbikers and some for trampers. A single questionnaire would have given a more accurate profile of the group that participates in both tramping and biking a group which proved to be very important in understanding the conflict. Analysis of a single questionnaire would also have been much easier than analysing two different questionnaires and trying to integrate the two. #0chead1  _) #2p}wC #3.6 Chapter Summaryb#0[head1#o\  PC+XP# Indepth interviews, participant observation and survey questionnaires were all used to investigate the conflict situation that has developed between mountainbikers and-Z0.. trampers in New Zealand. This is a pragmatic approach based on the idea that there is  X no absolute truth and no perfect way of doing research, given the complexity of social systems and the range of theoretical perspectives that underlie our understandings of conflict within those systems. Each method has its own strengths and weaknesses but when used together, they complement each other and, therefore, increase the internal and external validity of the overall research process. The research was carried out in the Canterbury area, however, respondents were not limited to talking only about the Canterbury area. Much mention was made of the Central North Island, Central Otago and the West Coast of the South Island, for example when bikers began to speak of their favourite rides. The Port Hills near Christchurch provided a good venue for participant observations, because of the large number of different users in the area. It also proved an interesting area of focus in terms of the familiarity for most respondents and the long associations some people had with the area.