<%BANNER%>

Integration of Sense of Place into Recreation Planning and Management in Ocala National Forest, Florida

Permanent Link: http://ufdc.ufl.edu/UFE0023652/00001

Material Information

Title: Integration of Sense of Place into Recreation Planning and Management in Ocala National Forest, Florida
Physical Description: 1 online resource (175 p.)
Language: english
Creator: Kil, Nam-Yun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: attachment, behavior, meanings, place, recreation, settings
Tourism, Recreation, and Sport Management -- Dissertations, Academic -- UF
Genre: Health and Human Performance thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The value of recreation settings has been well documented in recreation planning frameworks such as the recreation opportunity spectrum (ROS) as a land-classification system for recreation in which the quality of recreation experiences is associated with diverse recreation opportunities. Tangible natural resource setting attributes can be modified and adapted as needed to satiate visitors? desired experiences. This commodity-driven relationship between visitors and natural resources is framed as the commodity metaphor. This study sought to contribute to filling gaps in place theory as well as provide an enhanced example of a recreation planning process. That is, this study attempted to employ setting indicators for sense of place dimensions, connect recreation experience preferences with sense of place dimensions, and associate sense of place dimensions with place-protective behaviors. These linkages were primarily based on expectancy valence theory, theory of reasoned action and theory of planned behavior. On-site data was collected from the Ocala National Forest users between March and May during the spring of 2008. A random sampling technique was stratified to obtain a representative sample of all users that visited major recreation areas in the forest. Mail-back surveys were implemented as follow-up surveys. Structural equation modeling and spatial diagnostic tests were utilized to examine the linkage between sense of place constructs and recreation planning constructs. Results showed significant relationships between recreation experiences, a sense of place construct, and place-protective behavior. For instance, the level of importance of recreation experience preferences were not related to place meanings but were negatively related to place satisfaction, and positively related to overall place attachment. Overall, place attachment was influenced by place satisfaction and predictive of greater intentions toward place-protective behavior for the forest. More notably, setting perceptions were associated with attainment of recreation experiences, place satisfaction and place attachment. Attainment of recreation experiences predicted place satisfaction and place attachment, while it was not related to place-protective behavior. These results demonstrated a partial support of expectancy valence theory as well as theory of reasoned action. Additionally, although spatial diagnostic tests demonstrated similar findings, future studies may seek to identify additional variables influencing the sense of place construct.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nam-Yun Kil.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Holland, Stephen.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0023652:00001

Permanent Link: http://ufdc.ufl.edu/UFE0023652/00001

Material Information

Title: Integration of Sense of Place into Recreation Planning and Management in Ocala National Forest, Florida
Physical Description: 1 online resource (175 p.)
Language: english
Creator: Kil, Nam-Yun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: attachment, behavior, meanings, place, recreation, settings
Tourism, Recreation, and Sport Management -- Dissertations, Academic -- UF
Genre: Health and Human Performance thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The value of recreation settings has been well documented in recreation planning frameworks such as the recreation opportunity spectrum (ROS) as a land-classification system for recreation in which the quality of recreation experiences is associated with diverse recreation opportunities. Tangible natural resource setting attributes can be modified and adapted as needed to satiate visitors? desired experiences. This commodity-driven relationship between visitors and natural resources is framed as the commodity metaphor. This study sought to contribute to filling gaps in place theory as well as provide an enhanced example of a recreation planning process. That is, this study attempted to employ setting indicators for sense of place dimensions, connect recreation experience preferences with sense of place dimensions, and associate sense of place dimensions with place-protective behaviors. These linkages were primarily based on expectancy valence theory, theory of reasoned action and theory of planned behavior. On-site data was collected from the Ocala National Forest users between March and May during the spring of 2008. A random sampling technique was stratified to obtain a representative sample of all users that visited major recreation areas in the forest. Mail-back surveys were implemented as follow-up surveys. Structural equation modeling and spatial diagnostic tests were utilized to examine the linkage between sense of place constructs and recreation planning constructs. Results showed significant relationships between recreation experiences, a sense of place construct, and place-protective behavior. For instance, the level of importance of recreation experience preferences were not related to place meanings but were negatively related to place satisfaction, and positively related to overall place attachment. Overall, place attachment was influenced by place satisfaction and predictive of greater intentions toward place-protective behavior for the forest. More notably, setting perceptions were associated with attainment of recreation experiences, place satisfaction and place attachment. Attainment of recreation experiences predicted place satisfaction and place attachment, while it was not related to place-protective behavior. These results demonstrated a partial support of expectancy valence theory as well as theory of reasoned action. Additionally, although spatial diagnostic tests demonstrated similar findings, future studies may seek to identify additional variables influencing the sense of place construct.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Nam-Yun Kil.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Holland, Stephen.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0023652:00001


This item has the following downloads:


Full Text

PAGE 1

1 INTEGRATION OF SENSE OF PLACE INTO RECREATION PLANNING AND MANAGEMENT IN OCALA NATIONAL FOREST, FLORIDA By NAMYUN KIL A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Namyun Kil

PAGE 3

3 ACKNOWLEDGMENTS I would like to thank m y parents, brothers a nd sisters, and brothers and sisters-in law (esp. my father, mother, and brother) for their su pport of my educational car eers. I am grateful to Dr. Stephen Holland for his kind guidance and a dvice on my dissertation and other projects I worked with him on during my doctoral program. I am also grateful to Dr. Brijesh Thapa for his advice and support of my disserta tion processes. In addition, I w ould like to thank Dr. Taylor Stein and Division of the Forestry, Florida fo r their support of my di ssertation processes, including financial resources for data collection in the Ocala Nati onal Forest. I also would like to thank Dr. Timothy Fik for his guidance on spatial an alysis. I thank all students at the University of Florida and Ocala NF recreati on managers who helped me collect data in the forest. I also thank Drs. Robert Burns, John Confer, and Myron Floyd for their assist ance in developing my research skills through numerous projects I work ed with them on. I thank all the staff in the Department of Tourism, Recreation, and Sport Management, especially Donna Walker, Nancy Gulic, and Emer Ramos (esp. for Parks & Recrea tion Master Plan Proj ect for Putnam County, Florida). Lastly, I thank all my other colleagues in the Department as well.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 3 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .........................................................................................................................8 ABSTRACT ...................................................................................................................... ...............9 CHAP TERS 1 INTRODUCTION .................................................................................................................. 11 Significance of the Study ........................................................................................................13 Purpose of the Study .......................................................................................................... .....18 Hypotheses .................................................................................................................... ..........19 Additional Hypotheses ......................................................................................................... ...21 2 LITERATURE REVIEW .......................................................................................................24 Theoretical Frameworks of Recreation .................................................................................. 24 Recreation Opportunity Spectrum (ROS) ....................................................................... 24 Expectancy-Valence Theory ...........................................................................................26 Disconfirmation Models ..................................................................................................28 Predicting Behavior from Sense of Place ........................................................................ 29 Overview of Sense of Place ....................................................................................................31 Places ...............................................................................................................................32 Place Attachment ............................................................................................................. 34 Place Dependence ............................................................................................................34 Place Identity ................................................................................................................ ...35 Domains of Place Attachment or Sense of Place ............................................................36 Place Satisfaction ............................................................................................................ .37 Place Meanings ................................................................................................................39 Summary of Sense of Place Constructs ........................................................................... 40 Spatial Autocorrelation ....................................................................................................41 Empirical Literature Review ................................................................................................... 41 3 METHODOLOGY ................................................................................................................. 62 Description of the Study Site ................................................................................................. .62 Data Collection .......................................................................................................................63 Description of Variables and Instruments .............................................................................. 64 Importance and Attainment of REP ................................................................................ 65 Place Meanings ................................................................................................................67 Place Attachment ............................................................................................................. 68

PAGE 5

5 Place Satisfaction ............................................................................................................ .71 Place-Related Behavior and Acceptability of Potential Manag ement Actions ............... 73 Pilot Study ..............................................................................................................................75 Data Analysis ..........................................................................................................................76 Procedures of Confirmatory Factor Anal ysis and Structural Equation Modeling .......... 76 Procedures of Spatial Diagnostic Anal ysis of Place-Protective Behavior ......................78 Autocorrelation Statistics ................................................................................................80 Autocorrelation and Regression Models .........................................................................81 4 RESULTS ....................................................................................................................... ........84 Response Rate .........................................................................................................................84 Demographics of Respondents ...............................................................................................84 Non-Response Bias Test .........................................................................................................84 Differences in Behavioral Cons tructs am ong Different Users ............................................... 85 Respondents Overall Rating of Each Construct ....................................................................87 Measurement Model Analysis ................................................................................................ 89 Results of Hypothesis Testing ................................................................................................91 Structural Analysis of Model 1 ........................................................................................91 Additional Findings: Indirect Effects .............................................................................. 94 Structural Analysis of Model 2 ........................................................................................94 Findings of Direct Effects Same as Revealed in Model 1 ............................................... 96 Additional Findings: Indirect Effects .............................................................................. 96 Structural Analysis of Model 3 ........................................................................................97 Findings of Direct Effects Same as Revealed in Model 1 ............................................... 98 Additional Findings: Indirect Effects .............................................................................. 99 Additional Results of Spa tial Diagnostic Tests .............................................................100 5 DISCUSSION .................................................................................................................... ...124 Summary of Study Objectives and Major Findings ............................................................. 124 Structural Models .......................................................................................................... 124 Spatial Diagnostic Tests ................................................................................................131 Discussion .................................................................................................................... .........133 Structural Model 1 ......................................................................................................... 133 Structural Model 2 ......................................................................................................... 136 Structural Model 3 ......................................................................................................... 138 Spatial Diagnostic Tests ................................................................................................140 Delimitations ................................................................................................................. ........141 Limitations ................................................................................................................... .........141 Management Implications .................................................................................................... 142 Descriptive Behavioral Constr ucts among Different Users .......................................... 142 Respondents Overall Rating of Each Construct ...........................................................143 Structural Models .......................................................................................................... 144 Spatial Diagnostic Tests ................................................................................................145 Future Research ....................................................................................................................146 Structural Models .......................................................................................................... 146

PAGE 6

6 Spatial Diagnostic Tests ................................................................................................148 SURVEY INSTRUMENT ...........................................................................................................150 On-Site Survey Questionnaire .......................................................................................150 Mail-Back Survey Questionnaire ..................................................................................155 LIST OF REFERENCES .............................................................................................................162 BIOGRAPHICAL SKETCH .......................................................................................................175

PAGE 7

7 LIST OF TABLES Table page 4-1 Demographics of respondents ..........................................................................................103 4-2 Differences between re spondents and non-respondents .................................................. 104 4-3 ANOVA to determine differences in beha vioral constructs am ong different users ........104 4-4 Descriptive statistics for the m easurement model ........................................................... 105 4-5 Final fit indices of measurement m odels before structural models ................................. 109 4-6 Final factor loadings for the measur em ent model before structural models .................... 109 4-7 Direct and indirect eff ects in structural m odel 1 .............................................................. 112 4-8 Direct and indirect eff ects in structural m odel 2 .............................................................. 114 4-9 Direct and indirect eff ects in structural m odel 3 .............................................................. 116 4-10 Results of regression model 1 for spatial diagnostic test .................................................118 4-11 Results of regression model 2 for spatial diagnostic test .................................................118 4-12 Results of regression model 3 for spatial diagnostic test .................................................119 4-13 Results of spatial autocorrelation tests .............................................................................119

PAGE 8

8 LIST OF FIGURES Figure page 1-1 Hypothesized st ructural model 1 .......................................................................................22 1-2 Hypothesized st ructural model 2 .......................................................................................22 1-3 Hypothesized st ructural model 3 .......................................................................................23 2-1 Framework of ROS planning system .................................................................................60 2-2 Expectancy theory: Motivation for nature-based recreationists ........................................ 60 2-3 Effect of gap score on satisfaction .....................................................................................60 2-4 Predicting behavior from sense of place ............................................................................60 2-5 Model of place relationships ..............................................................................................61 3-1 Ocala National Forest, FL ................................................................................................ ..83 4-1 Final structural model 1 ................................................................................................. ..113 4-2 Final structural model 2 ................................................................................................. ..115 4-3 Final structural model 3 ................................................................................................. ..117 4-4 Results of spatial autocorrelation test for model 1 ........................................................... 120 4-5 Results of spatial autocorrelation test for model 2 ........................................................... 121 4-6 Results of spatial autocorrelation test for model 3 ........................................................... 122 4-7 Results of spatial analysis of place-protec tive behavior ..................................................123

PAGE 9

9 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy INTEGRATION OF SENSE OF PLACE INTO RECREATION PLANNING AND MANAGEMENT IN OCALA NATIONAL FOREST, FLORIDA By Namyun Kil December 2008 Chair: Stephen Holland Major: Health and Human Performance The value of recreation settings has b een well documented in recreation planning frameworks such as the recrea tion opportunity spectrum (ROS) as a land-classification system for recreation in which the quality of recreation e xperiences is associated with diverse recreation opportunities. Tangible natura l resource setting attri butes can be modified and adapted as needed to satiate visitors desired experiences. This commodity-driven re lationship between visitors and natural resources is framed as the commodity metaphor. This study sought to contribute to filling gaps in place theory as well as provide an enhanced example of a recreation planning proc ess. That is, this study attempted to employ setting indicators for sense of place dimensions connect recreation experience preferences with sense of place dimensions, and associate sens e of place dimensions with place-protective behaviors. These linkages were primarily base d on expectancy valen ce theory, theory of reasoned action and theory of planned behavior. On-site data was collected from the Ocala National Forest users between March and May during the spring of 2008. A random sampling techni que was stratified to obtain a representative sample of all users that visited major recreati on areas in the forest. Mail-back surveys were implemented as follow-up surveys. Structural equation modeling and spa tial diagnostic tests

PAGE 10

10 were utilized to examine the linkage between se nse of place constructs and recreation planning constructs. Results showed significant relationships betw een recreation experience s, a sense of place construct, and place-pro tective behavior. For instance, the level of importance of recreation experience preferences were not re lated to place meanings but were negatively related to place satisfaction, and positively related to overall place attachment. Overall, place attachment was influenced by place satisfaction and predictive of greater intentions to ward place-protective behavior for the forest. More notably, setting pe rceptions were associated with attainment of recreation experiences, place sa tisfaction and place attachment Attainment of recreation experiences predicted place satisf action and place attachment, while it was not related to placeprotective behavior. These results demonstrated a partial support of expectancy valence theory as well as theory of reasoned action. Additionally, although spatial diagnostic tests demonstrated similar findings, future studies may seek to identi fy additional variables in fluencing the sense of place construct.

PAGE 11

11 CHAPTER 1 INTRODUCTION It is im portant to understand recreation settings to improve outdoor recreation management. The primary reason is that recreation takes place in settings, and recreation activities and the quality of recreation experi ences are contingent upon the setting in many outdoor recreation cases. Therefore, most outdoor recreation research has been related to better understanding the settings suitable for specific ac tivities or desired experiences (Manning, 1999; Williams, Patterson, Roggenbuck, & Watson, 1992). The value of recreation settings has been well documented in recreation planning framew orks such as the R ecreation Opportunity Spectrum (ROS) as a land-classification system for recreation in which th e quality of recreation experiences is associated w ith diverse recreation opportunitie s. Other resource management frameworks such as Limits of Acceptable Cha nge (LAC) and Visitor Experience and Resource Protection (VERP) were devel oped to evaluate and manage impacts of recreational use on natural resource settings (Moore & Scott, 2003). Along with such frameworks, a multiple-use utilitarian philosophy was traditionally accepted among natural resource recreation manage rs throughout most of the last century (Smaldone, Harris, Sanyal, & Lind, 2005; Wellman, 1987). Following this idea, natural recreation settings, much like cons umer products, are viewed as a collection of attributes or characteristics to satisfy visitors recreation goal s. That is, tangible na tural resource setting attributes can be modified and adapted as needed to satiate visitors de sired experiences. This commodity-driven relationship between visitors and natural resources is framed as the commodity metaphor. However, such a commodity view has also been problematic. Sometimes, the tangible property aspects of recreation settings cannot be designed or reproduced (Williams et al., 1992).

PAGE 12

12 If settings are manipulated or are interchangeab le, the values or meani ngs of a special place recreationists seek may be lost. As a consequenc e, recreationists may not attain their desired experience preferences and benef its, and they may be displaced (Warzecha & Lime, 2001). Some have emphasized that more value should be put on emotional and symbolic views of natural resources rather than a commodity view (S maldone et al., 2005; Williams et al., 1992). Furthermore, it has been suggested that when recreation resource planning frameworks neglect the intangible characteris tics of place, outdoor recreation planne rs need to consider integrating place attachment/sense of place as useful c oncepts into conventional recreation planning frameworks to better meet the needs (preferences and attitudes) of recr eationists for recreation settings (Mitchell, Force, Carroll, & McLaughlin, 1993; Warzecha & Lime, 2001; Williams et al., 1992). As part of the effort to combine recreati on planning frameworks with place attachment, numerous studies have been conducted to identify the effects of place attachment on environmental factors such as recreation setting preferences (Warzecha & Lime, 2001), acceptability of recreation resource setti ngs (Kyle, Graefe, Manning, Bacon, 2004a, 2004b; Warzecha & Lime, 2001), and environmentally re sponsible behavior (Vaske & Kobrin, 2001). Furthermore, variables that foster the developmen t of place attachment have also been identified. For instance, they can include recreational motivations (K yle, Mowen, & Tarrant, 2004), frequency of visits, duration of visits, and length of association (Moore & Graefe, 1994; Moore & Scott, 2003). In the meantime, the symbolic and emoti onal meanings assigned to a place have been documented in numerous ways across various academ ic disciplines. Place attachment is defined as a positive emotional bond with a specific place (Low & Altman, 1992) and has been widely

PAGE 13

13 used as a place-attached c oncept and has been commonly re ported to have two distinct dimensions; place dependence and place identity. Mo re recently, the term that encompasses all place-attached concepts is sense of place (Shamai, 1991; Stedman, 2000, 2002, 2003a, 2003b) consisting of three major constructs, place m eaning, place attachment and place satisfaction. These constructs have also been linked to place-related behavior in more recent studies. Significance of the Study This study sought to contribute to filling gaps in place theory as well as provide an enhanced example of a recreation planning process. First of all, the theoretical perspective is based on a belief that most previous studies on emotional and symbolic relationships between individuals and a place have been conducted us ing the term, place attach ment. Many researchers have made efforts to identify how attachment to a place is developed. They have identified numerous variables that foster place attachment and other variab les that are affected by place attachment, some of those variables are mentione d briefly later in this section. Unlike previous studies, however, Stedman (2002) used sense of place as an overarching concept of place encompassing three dimensions: place meani ngs, place satisfaction, a nd place attachment. Additionally, those dimensions were linked to pr edict future place-protective behavior. More importantly, the construct of place meaning is bett er able to support the c oncept of sense of place by predicting place attachment, place satis faction, and place-related behavior. However, Stedman (2002) raised the ques tion of how cognitions are created. Reporting that favorite activities and residence pattern s did not adequately explain place meanings, Stedman (2002) suggested that more research should be conducted on identifying the causal factors of place meanings and that three dime nsions of sense of place (place meaning as cognitions, place attachment as id entities, place satisfaction as at titude) should be explored for identifying those factors. Place meanings would be formed through shaping the physical

PAGE 14

14 landscape, through interpretation of the landscape, and in fa voring certain experiences over others (p. 577). As Stedman s uggested, this study aimed to c ontribute to identifying those factors to fill gaps in sense of place theory. As suggested by Kyle, Graefe, and Manning (2003/2004), there is a need to examine the relationship between the importan ce of recreation setting preferen ces and place attachment (esp. place identity). Length of association with reso urce settings and the importance of recreation activities were positively related to place iden tity. Furthermore, as Mesch and Manor (1998) indicated, social settings (e .g., number of friends, types of people in neighborhood) were important to foster the development of place atta chment. The linkage between recreation settings and place attachment can also be identified by the definition of place attachment (e.g., place dependence and place identity). Wh en recreational settings are cons istent with recreational goals and needs, place dependence may be developed. What remains to be done is an empirical examination of the relationship between recreation setting perceptions and place attachment or place satisfaction because recreati on settings may affect attainme nt of desired recreation goals. In addition, Kyle, et al. (2004) conducted a study on the relationship between motivation (the importance of recreation experience pref erences) and place attachment, using Lawlers (1975) motivation theory. Since my study used th ree major dimensions of sense of place (place meanings, place satisfaction, and pla ce attachment), a more inclusive analysis of the relationship between motivation and sense of place dimensions is one of the foci of this present study. This study employed expectancy valence theory to te st the relationship. Establishing a significant relationship between motivation and place meanings would lead to a determination on whether meanings are formed not only through individual s direct experiences with a setting (Tuan, 1974, 1977) but also through indirect experience before their physical experien ce with a setting (Hunt,

PAGE 15

15 1975; Echtner & Ritchie, 1993; Gartner & H unt, 1987; Pearce, 1982; Young, 1995; Walmsley & Young, 1998; Young, 1999a). Furthermore, one of the assumptions of the Recreation Opport unity Spectrum (ROS) framework (Clark & Stankey, 1979; Driver & Brown, 1978) is that th e attainment of individuals recreation experience preferences (REP) leads to their recreati on satisfaction which sets up a need for this study to examine the relationship be tween the attainment of REP and sense of place aspects. This expanded the current literature on sense of place (e.g., Kyle, et al., 2004; Stedman, 2000, 2002, 2003a, 2003b) as well as on recreatio n satisfaction models (Whisman & Hollenhorst, 1998). Overall, this would contribute to the concep tual linkage between ROS and sense of place constructs, supporting justifications that tangible properties of recr eation settings should be protected and preserved. In the meantime, the concept of place atta chment is also important when recreation planners and managers manage na tural resources for recreation. It has been a valid means of predicting human behavior in numerous studi es (e.g., Kaltenborn, 1998; Kyle, et al., 2004a, 2004b; Payton, Fulton, & Anderson, 2005; Stedma n, 2002; Vaske & Kobrin, 2001; Warzecha & Lime, 2001; Williams & Vaske, 2003). As Kyle, Mowen, and Tarrant (2004) also suggested, it would be helpful to link sense of place studies to conflict, visitor preferences for resources, and visitor evaluations of management actions. Furthermore, Stedman (2002, 2003a, 2003b) initiated a linkage between sense of place constructs and place-protective behavior. But there is still a need to examine the relationship between sens e of place constructs (place meanings, place satisfaction, place attachment) and place-protective behavior becau se lack of evidence on the relationship exists. Additionally, examining other possible causal factors (e.g., the attainment of REP, place setting perceptions) of place-relate d behavior was another focus of this study.

PAGE 16

16 Overall, this study attempted to connect recreation experien ce preferences with sense of place dimensions and to associate sense of place dimensions with place-protective behaviors. This study could contribute to better understanding of sense of place by identifying the factors that affect place attachment as well as the effect of sense of place on at titudes toward recreation resource management policies and participation in the planning process (Williams & Vaske, 2003). From a managerial perspec tive, this study could be help ful for recreation resource planning managers to better understand how various recreation settings are perceived and selected by recreationists when pursuing recreational opportunities that sa tiate their recreation experiences (Moore & Graefe, 1994). Sense of place dimensions would likely vary across recreation setting preferences and recreation experien ce preferences or attainments. As suggested by Warzecha and Lime (2001), identifying the re lationships would facilitate satisfying the recreational needs of visitors with a strong sense of place. Instead of focusing on the historically do minant utilitarianism management system (Williams & Stewart, 1998), resource planning managers should recognize the uniqueness and spirit of places in the land planning process by centering on attachment-oriented values that are undervalued in the planning proces s, although attachments for places are not easier to plan for than uses (Mitchell, et al., 1993) Sense of place helps resource managers identify and decide desired future conditions of re sources in both ecological and hu man terms. It strengthens the relationship between individuals a nd a place. Therefore, it is deemed important to include every individuals particular sense of place in the resource plan ning process and decision making (Williams & Stewart, 1998). Calls for integrating sense of place into recreation planning frameworks such as ROS or LAC have been made in sense of pla ce studies by planning

PAGE 17

17 researchers and managers (e.g., Mitchell, et al., 1993; Moore & Graefe, 1994; Warzecha & Lime, 2001; Williams & Stewart, 1998). As mentioned earlier, this study employed ROS structures (e.g., importance of recreation experiences preferen ces, attainment of REP, gap scores of REP), sense of place dimensions (e.g., place satisfac tion variable composed of ROS settings). and place-protective behaviors (e.g., items composed of ROS settings) to shed light on identifying where management actions could be implemente d to satiate individuals and stakeholders (Williams, et al., 1992), especially when different opinions over management objectives need to be resolved (Cheng, Kruger, & Daniels, 2000). Additionally, further rese arch on recreation behavior should be conducted using Geographic Information Systems (GIS) to make better predictions of th e spatial aspects of recreation. Numerous calls have been made for in tegrating GIS into sense of place studies. For instance, landscape values have been mapped on scales to identify a nd manage any potential issues on the various values of the same res ources (Brown, Reed & Harris, 2002). Additionally, place attachment to natural resour ce areas is predicted by numerous variables such as distance and frequency of visitation (Moore & Graefe, 1994; Moore & Scott, 2003). From a geographical statistics perspectiv e, however, what has been overlooked in predicting place attachment or othe r place-related constructs is spat ial analyses that utilize the concept of spatial autocorrelation in residuals from regression models. That concept has been widely used in other academic disciplines su ch as geography, ecology and anthropology for predicting certain behavior or phenomenon. Spatial autocorrelation has proven to be useful in creating suitable regression models that predict spatial phenomenon. GIS applications with spatial autocorrelation analyses have also been used for an environmen tal justice study that examines the spatial relationship between th e level of access to parks and socioeconomic

PAGE 18

18 variables (Talen & Anselin, 1998). Further application of such sp atial autocorrelation analyses, however, should be also conducted in regression models that pr edict place-related constructs. This application would con tinuously contribute to behavi oral recreation research. Purpose of the Study The primary purpose of this study was to exam ine three major structural models, each of which consists of recreation experience, sense of place constructs, and place-related behaviors. The first structural model included three major themes: 1) recreation experience preferences (REP) within a recreation opportunity spectrum (ROS) framework, 2) place meanings (PM), place satisfaction (PS), and place a ttachment (PA) within percepti ons of sense of place (SOP), and 3) place-protective behaviors (PB) associated with pl aces. The second structural model focused on two additional themes: a) the impacts of PM on the attainment of REP, and b) the impacts of REP attainment on PS, PA, and PB. The third structural model focused on two fold: a) the impacts of PM on GAP scores, and b) th e impacts of GAP scores (values obtained after subtracting importance of REP values from attainment of REP) on PS, PA, and PB. The formulation of research questions for this study centered on identifying antecedent factors that influence sense of place as well as fa ctors that influence place-protective behaviors. In addition, the formulation was based on integrat ing three major constructs (ROS structures, sense of place, and place-related behaviors) as developed by previous co nceptual and empirical research. In a broad sense, the conceptual di fferences among models 1, 2 and 3 are based on the different ways visitor satisfac tion is measured. They include importance, attainment, and gap scores of recreation experiences; each of them predicts place satisfaction. In general, this study represented an attemp t to examine how sense of place dimensions (place attachment, place satisfaction, place meani ngs) are influenced by recreation experience which is one of the ROS components and how th ey affect individual acceptance of potential

PAGE 19

19 future resource management policies and practices in the Ocala National Forest (NF), Florida. Motivation theory (Lawler, 1973) and expectan cy-valence theory (Ajz en & Fishbein, 1980; Fishbein, 1967) led to a conceptual linkage between REP and sense of place dimensions (e.g., motivation linked to place satisfaction or place attachment). E xpectancy-valence theory also contributes to a conceptual re lationship between sense of plac e dimensions, especially the relationship between place meanings and place satis faction. Lastly, the theory of reasoned action (Ajzen & Fishbein, 1980) and the theory of pl anned behavior (Ajzen, 1991a, 1991b) purportedly foster an attitude-behavior relationship and an identity-behavior relationship which was also explored in this study. One last important purpose of the study was related to spatia l diagnostic tests, using the concept of spatial autocorrelation. The purpose wa s to examine conventional linear regression model residuals in order to provide more accurate estimates of the variables that could have significant impacts on place-related behaviors. The causal variables were the factors from each construct used for structural models to estima te the direct effect of the variables on placeprotective behaviors. This sought to sustain consistency of pred icting place-protective behaviors from the same variables, although different analys es were utilized. Therefore, three different spatial diagnostic tests were examined. Hypotheses This study examines three major structural models: Model 1: The first structural model includes th ree major concepts: 1) recreation experience preferences (REP) within a recreation opportunity spectrum (ROS) framework, 2) place meanings (PM), place satisfaction (PS), and place attachment (PA) within perceptions of sense of place, and 3) place-protectiv e behaviors (PB) (Figure 1-1):

PAGE 20

20 The relationship between the importance of RE P and sense of place constructs (PM, PS, and PA): H1: There is a positive relationship be tween the importance of REP and PM. H2: There is a positive relationship between the importance of REP and PS. H3: There is a positive relationship between the importance of REP and PA. The relationship between PM and other se nse of place constructs (PS and PA): H4: There is a positive relati onship between PM and PS. H5: There is a positive relati onship between PM and PA. The relationship between PS and PA: H6: There is a positive relationship between PS and PA. The relationship between the importance of REP and place-protec tive behavior (PB): H7: There is a positive relationship between the importance of REP and PB. The relationship between sense of place constr ucts and place-protective behavior (PB): H8: There is a positive relati onship between PM and PB. H9: There is a positive relationship between PS and PB. H10: There is a positive relati onship between PA and PB. Model 2: The second structural model focuses on two parts: 1) the impacts of PM on the attainment of REP, and 2) the impacts of RE P attainment on PS, PA, and PB (Figure 1-2). H11: There is a positive relationship betw een PM and the attainment of REP. H12: There is a positive relationship betw een the attainment of REP and PS. H13: There is a positive relationship betw een the attainment of REP and PA. H14: There is a positive relationship betw een the attainment of REP and PB. Model 3: Similarly, the third structural model focuses on two parts: 1) the impacts of PM on GAP scores, and 2) the impacts of GAP sc ores on PS, PA, and PB (Figure 1-3). H15: There is a positive relationshi p between PM and GAP scores. H16: There is a positive relationshi p between GAP scores and PS. H17: There is a positive relationshi p between GAP scores and PA. H18: There is a positive relationshi p between GAP scores and PB.

PAGE 21

21 Additional Hypotheses H19: There is no spatial autocorrelat ion in regression model residuals in predicting the effects of the importance of REP, PM, PS, and PA on PB. H20: There is no spatial autocorrelat ion in regression model residuals in predicting the effects of the attainment of REP, PM, PS, and PA on PB. H21: There is no spatial autocorrelat ion in regression model residuals in predicting the effects of gap scores of REP, PM, PS, and PA on PB.

PAGE 22

22 Figure 1-1. Hypothesize d structural model 1 Figure 1-2. Hypothesize d structural model 2 H9a H8a H12 H6a H14 H10a H13 H11 H4a H5a Place Meanings Place Satisfaction Attainment of REP PlaceProtective Behavio r Place Attachment Note. aThe same h yp otheses are shown in Model 1 and not examined in Model 2. H9 H7 H4 H6 H8 H10 H5 H1 H2 H3 Importance of REP Place Satisfaction Place Meanings PlaceProtective Behavio r Place Attachment

PAGE 23

23 Note. aThe same hypotheses are shown in Model 1 and not examined in Model 3. Figure 1-3. Hypothesize d structural model 3 H9a H8a H16 H6a H18 H10a H17 H15 H4a H5a Place Meanings Place Satisfaction GAP Scores PlaceProtective Behavio r Place Attachment

PAGE 24

24 CHAPTER 2 LITERATURE REVIEW Theoretical Frameworks of Recreation This section provides theoretical fram eworks of recreation to support hypotheses stated in Chapter 1. The theoretical frameworks include recreation opportunity spectrum, expectancyvalence theory, disconfirmation models, predicti ng behavior from sense of place. Additional concept includes spatial autocorrel ation that contribute to determ ination of whether conventional regression models are correctly specified. Recreation Opportunity Spectrum (ROS) It is a difficult task to p lan and manage outdoor recreation environments because it requires complex information such as resource capab ilities, visitor demands, agency policies, and so on. To develop rational approaches to recreation planning and management, user needs should be matched to various types of recrea tion settings (Heywood, Christensen, & Stankey, 1991). Recreational experience is regarded to be the set of psychological outcomes individuals achieve when engaging in an activity in a spec ific recreation setting or resource (Clark & Stankey, 1979; Driver & Brown, 1978). One conceptu al model that links those factors is known as the Recreation Opport unity Spectrum (ROS). The ROS model was developed in the late 1970s by researchers associated with the Forest Service for an applied planning and allo cation system (Clark & Stankey, 1979; Driver & Brown, 1978) and was initially utilized by the US Forest Service and the Bureau of Land Management. ROS attempts to link the relationships between settings that offer opportunities for recreation activities, experiences and outcomes. The settings and re creation activities are supplied by resource managers and diverse recr eational experiences ar e sought and achieved by recreationists (Clark & Stanke y, 1979). Diverse recr eation opportunities ar e categorized by a

PAGE 25

25 number of opportunity classe s (zones), crossing from primitiv e natural settings to more developed settings that visitors select from. Each ROS class c onsists of physical, social and managerial indicators as operational definitio ns. Physical indicators can include access, remoteness, naturalness (degree of human modification of the natural environment), and size of an area. Social indicators can in clude encounters with other visito rs and the presence of visitor impacts. Managerial indicators can include the degree of developed recreation facilities and sites, and the amount of on-site regulation (Clark & Stankey, 1979). Specific standards are set for each indicator, thereby dis tinguishing between varied recreation opportuni ties (Clark & Stankey, 1979; Driver, Brown, Stankey, & Gregoire, 1987). A water recreation opportunity spectrum (WROS), centering on water-based recreation resources, was modeled after ROS to set appropriate management objectives and provide diverse aquatic recreational opportunities to the public and protect natural resources (Haas, A ukerman, Lovejoy, & Welch, 2004) (Figure 2-1). The ROS conceptual framework is based on expectancy theories borrowed from and most common in social psychology (Ajzen & Fishbe in, 1980; Iso-Ahola 1980; Lawler, 1973). Again, recreationists seek out recreationa l experiences by engaging in part icular recreation activities in particular recreation settings provided by resource managers and achieve desirable and expected recreation experiences that cont ribute to human benefits (Dri ver, Brown, & Peterson, 1991). This model has been reported to be rational and supports a commodity metaphor that settings are managed for social and economic benefits. However, this commodity view has been criticized, due to not accounting for intangibl e characteristics (e.g., meanings) of natural resources or places (Schreyer 1989; Williams & Patte rson 1994; Williams, et al., 1992). One of the underlying questions about ROS assumptions is whether some recreation experiences are heavily setting-dependent while others are setting independent. For instance,

PAGE 26

26 social relationship experiences such as family to getherness and being with other people are less likely to be setting-dependent, wh ile more isolated experiences such as seeking solitude and connections with nature tend to be more settin g-dependent. It would be helpful to understand the types of recreation experiences that are setting/activity depend ent so that such recreation experiences can be categorized across ROS cla sses (Driver, Brown, St ankey, & Gregoire, 1987; Kaltenborn, 1999). Furthermore, experienced-based manageme nt emphasizes the importance of physical, social, and managerial settings to obtain de sired psychological recreation experiences. Hammitt and Patterson (1991) observed that campers w ith low encounter norms seemed to be less sensitive to higher density setti ngs. Similarly, Virden and Knopf (1989) observed that individuals with socially related expected outcomes were able to with stand higher encounter norms. In addition, whether individuals can attain desired recreation experiences could be contingent upon the presence of setting features such as the presence of wildlife, crowds, tranquility or litter (Floyd & Gramann, 1997). Floyd and Gramann (1997) observed that recreationists (hunters) seeking high-challenge experi ences preferred more difficult access and areas with low management control. Heywood, Christensen and Stankey (1991) observed the linear and nonlinear re lationship between biophysical and social settings. Manning (1999) mentioned that motivations were found to be a si gnificant predictor of vi sitor expectations for and attitudes about wilderness condi tions. Overall, these indicate th at integrated future research may link the importance of recreation experiences to the construct of place attachment that is influenced by the level of importance of recreation experiences (Kyle, et al., 2004). Expectancy-Valence Theory Most wilderness users value the natura l environment (Kaplan & Talbot, 1983), but whether or not they find the quality of the enviro nment they desire depends on the quality of the

PAGE 27

27 area they visit. When the environment they encoun ter during their visit is the same as what they are seeking, the quality of the vi sitor experience is likely increas ed. Otherwise, their recreation experience in natural settings might be negatively affected (Flood, 2002). Recreation managers and researchers are incr easingly concerned about the motivations of visitors to more nature-based recreation areas their evaluation of onsite conditions, and the effect of perceived onsite conditi ons on the quality of visitor experiences in natural settings (Flood, 2002). The most well-known model that assi sts our understanding of the relationship is Lawlers (1973) expectancy model. Expectancy th eory suggests that outc omes are attractive to an individual because of some drive or need that the individual possesses. That is, individuals engage in behaviors that they believe will pr oduce positive outcomes (Lawler, 1973). It has been adopted by many scholars (Cole, 1996; Hall & Sh elby, 1993) from the fields of psychology and sociology, to measure human behavior in more natu re-based recreation sett ings, specifically to represent the influence of visitor expectati ons (motivations) on perceptions of resource conditions, and the role of perceived resource cond itions on determining visitor satisfaction with nature-based recreation experi ences (Figure 2-2). Studies (e .g., Rossman & Ulehla, 1977) observed the effect of visitor e xpectations on the perceptions of resource conditions as well as the nature-based recreation experience. This also has been explained by other earlier researchers such as Fishbein (1967) and Ajzen and Fishbein (1980) that initiated exp ectancy-valence theory. The expectancy-valence theory framework also sugge sts that more visitors at nature-based areas choose resource conditions, expecting that they will achieve their desire d recreation experience in those settings (Flood, 2002). In addition, expectancy-valen ce theory has led to the de velopment of other leisure theories and management frameworks such as th e theory of reasoned action (Ajzen & Fishbein,

PAGE 28

28 1980), the theory of planned be havior (Ajzen, 1991a, 1991b), experience-based management (Driver & Brown, 1983), and benefits-based management (Driver, 1996). Benefit-based management centers on better understanding what vi sitors are pursing in re creational experiences (Bruns, Driver, Lee, Anderson, & Brown, 1991). Managers can manage re creation settings to provide opportunities for specifi c experiences and benefit out comes (Stein & Lee, 1995). Disconfirmation Models In the m eantime, there are different ways to predict visitors satisfaction in relation to recreation settings and experiences. The approach es include expectations, and performance or gap scores, which are borrowed from marketing disc iplines. In particular, gap scores were used by many scholars (e.g., Parasuraman, Zeithaml, & Berry, 1985, 1988; Zeithaml, Berry, & Parasuraman, 1988; Zeithaml, Parasuraman, & Berr y, 1985) in marketing di sciplines to predict service quality and satisfaction and have been applied by many scholars (e.g., Absher, Howat, Crilley, & Milne, 1996; Backman & Veldka mp, 1995; Hamilton, Crompton, & More, 1991; Holland, 1979; MacKay & Crompton, 1988, 1990) in parks and recreation areas (Figure 2-3). Expectancy and discrepancy theo ries suggest that i ndividuals satisfaction is determined by differences between their expected recreati on experience and actual recreation experience. Discrepancies exist in many areas including the number of encounters, mode of travel, the number of people in other groups, and so on (Schreyer & Roggenbuck, 1978). Only performance items were used in national forest recreation set tings (Absher, 1998) as we ll as service marketing areas (Churchill & Suprenant, 1982; Oliver & DeSa rbo, 1988). Expectation scores can be used in measuring satisfaction levels although it is no t necessary (Babakus & Boller, 1992; Carman, 1990). In this sense, there is st ill continuing debate as to the selection of optimal satisfaction measures (Crompton & Love, 1995).

PAGE 29

29 In addition, there are few guidelines or agreement on what dimensions of satisfaction should be included in the contex t of recreation management (Burns, Graefe, & Absher, 2003). A recreational performance framew ork with four tangible sett ing domains (e.g., facilities sufficiency, facilities operations, services, and information) was developed and has been used by recreation researchers (e.g., Absher, et al., 1996; Howat, et al., 1996). More recently, a recreational experience domain (less tangible) that was not used previously was added to instrumental setting domains (Burns, et al., 2003). The recreational e xperience domain that Burns, et al. (2003) added in their research was related to crowding and conflicts between different types of users which could affect the quality of recreation experiences (Manning, 1999). In addition to crowding and conflict items, Bu rns, et al. (2003) suggested that future studies should add more recreation experience domains (e.g., social gathering, escape from daily life), when measuring individuals satisfaction with recreation experi ences. Suggestions for including other recreatio n experience domains were also i ndicated by other researchers. For instance, Crompton and Mackay (1989) observed that recreation experien ce plays an important role in explaining recreation sa tisfaction. Connelly (1987) observed that recreati onal experience (expressive) items contributed more to vis itor satisfaction than instrumental items. Predicting Behavior from Sense of Place The Theory of Reasoned Action (Ajzen & Fi shbein, 1980) suggests that attitude and subjective norm s predict behavioral intentions and actual behaviors. A measure of perceived behavioral control as non-voliti onal behavior was incl uded in the Theory of Planned Behavior (Ajzen, 1991a, 1991b). Basically, the Theory of Reasoned Action and Theory of Planned Behavior predict relationships between attitude and future behaviors. Examination of this relationship is common in social psychology. In general, there is a positive relationship between attitude and behavior. That is people who hold positive attitude s should engage in behaviors

PAGE 30

30 that approach, support, or enhance the attitude object, and people who hold negative attitudes should engage in behaviors that avoid, oppose, or hinder the object (Eagly & Chaiken, 1993, p.155). Multiple indicators of be haviors are considered more valid means of measuring the relationship between attitude and behavior than single indicator of behaviors (Eagly & Chaiken, 1993; Fishbein & Ajzen, 1975). In the context of sense of place studies, it is suggested that attitude theory can provide a basis for conceiving place attachment as pot entially encompassing c ognitive, affective and conative reactions to a spatia lly based object (Jorgensen & Stedman, 2001, p. 244). The effect of various domains of sense of place (meanings, a ttachment, and satisfacti on) on place-protective or pro-environmental behaviors has also been examined within the Theory of Reasoned Action and the Theory of Planned Behavior by sens e of place researchers (e.g., Stedman, 2002). The relationship between attachment and behavior is also reported by ot her researchers (e.g., Kaltenborn & Williams, 2002; Vaske & Kobrin, 2001; Walker & Chapman, 2003). For instance, Walker and Chapman (2003) obser ved that sense of place positively predicted pro-environmental intentions such as picking up litter, not littering, support of higher entrance fees to prevent poaching, and not visiting environmentally dama ged places undergoing recovery in parks. Kaltenborn and Williams (2002) observed that local residents with higher place attachment were more likely to have positive attitudes toward e nvironmentally favorable management priorities, compared to those with lower place attach ment. Protection of cultural monuments and wilderness and untrammeled nature was considered to be more important to highly attached residents, while facilitation of hunting and fish ing was less important to them. However, not much evidence on the impacts of the sense of place domains on behavior has been reported in the literature (Stedman, 2003a). Place attachment has a significant and positive impact on

PAGE 31

31 willingness to protect a place, while place satisf action has a significant and negative impact on it. In addition, pristine meanings of a place were po sitively related to pla ce-protective behaviors. But more developed meanings of a place were negatively related to behaviors (Stedman, 2002) (Figure 2-4). Overview of Sense of Place Contributors to the concept of sense of place hail from various academic fields such as environmental psychology, sociology, geography, anthropology, landscape and urban design, and architecture, illustrating the multidisciplinary utility of the concept. Therefore, meanings ascribed to places by individual people have been conceptualized under many different terms such as sense of place (Buttimer & Seam on, 1980; Chawla 1992; Relph, 1976; Tuan, 1980), place attachment (Altman & Low, 1992; Proshansky, Fabian, & Kaminoff, 1983), rootedness (Chawla 1992; Tuan, 1980), t opophilia (Tuan, 1974), place belongingness (Proshansky, et al., 1983), place identity (Pro shansky, et al., 1983), and community sentiment and identity (Hummon, 1992). The fundamental concepts are that a sense of place is generated when people ascribe meanings to what would be called space (Moore & Scott, 2003). All these different terms, however, have led to no clear theoretical relationship between the concepts. For instance, place attachment en compasses or is encompassed by all of those various terms stated earlier (Stedman, 2003b). Sense of place has been defined in many differe nt ways by numerous scholars. It refers to the meaning and importance of a setting held by an individual or group, based on an individuals and groups experience with the setting (p. 822, Stedman, 2003b). It can be considered as the collection of meanings, belief s, symbols, and feelings that individuals or groups associate with a particular locality (p.19, W illiams & Stewart, 1998). It is also viewed as a more practical process created by the setting co mbined with what a person brings to it. In

PAGE 32

32 other words, to some degree we create our own pl aces, they do not exist independent of us (p. 9, Steele, 1981). Tuan (1974), a human and environm ental geographer, conceives sense of place as the relationship between people a nd the land their environmental perceptions of the relationship between a place and themselves. It covers symbolic and emotional perspectives. Differentiating terms, a place, a sense of place and region, Sham ai (1991) conclusively considered a sense of place as an overarching concept encompassing a ll other concepts such as place attachment, national identity, and awareness or perception of region, all of wh ich describe the relationship between people and spatial settings. As an umbrella concept, sense of place refers to the entire set of cognitions and affective sentiments on a particular place (Altman & Low, 1992; Jorgensen & Stedman, 2001) as well as meanings ascrib ed to the place (Kaltenborn, 1998; Relph, 1976; Stedman, 2003a). Places Places are based on directly experienced phenomena of the lived world, full of meanings, with real objects, ongoing activities.. and become important sources of individual and community identity often profound centers of human existence with deep emotional and psychological ties (p. 141, Rel ph, 1976). Similarly, Low and Altman (1992) stated that a place is an individual and social experience, where by a place has meanings to people at individual and collective levels. Tuan (1974, 1975, 1977) mentioned that place, as a geographic term, is a hub of meaning built by experience. Only if a place is understood ab stractly or theoretically, and we get to know a place as we are familiar with close friends, can we say that the place is fully known to us. There is a difference between a small place a nd a large place in the way people know places. The former is known through all the modes of experi ence (e.g., home, fireplac e) and the latter is known through indirect and abstract experiences (e.g., a nation/state known by national anthem

PAGE 33

33 or geography or history). Thus, what begins as undifferentiated space becomes place as we get to know it better and endow it with value (p. 6, Tuan, 1977). That is, through human experience, spaces are more general, while pla ces are more specific and concrete (Tuan, 1977) and represent more than physical environments (Tuan, 1974). As Tuan (1977) stated, a forest landscape may begin as undifferentiated spaces and then be given personal meanings through accumulated e xperiences with the forest landscape. More direct or longer experiences tend to contribute to str onger attachment of indi viduals to the forest setting. Furthermore, these transforming experien ces are diverse across di fferent interactions such as work, recreation, or daily life intera ctions with the forest landscape (Stedman, 2003b). Meanings vary between people that interact with forest settings. A setting may have a specific meaning or various meanings (e.g., Grieder & Garkovich, 1994; Lync h, 1960; Meinig, 1979; Relph, 1976). For instance, a particular forest se tting may mean a primitive wilderness, a place for daily living and/or a place for active recreation, depending on th e types of direct experiences with that setting. Urbanites with less outdoor ex periences may ascribe wilderness meanings to even a small pine plantation. Additionally, symbolic meanings are forged through experiences with social relationships as well as physic al settings (Stedman, 2003b). A place is usually generated through a direct interaction between physical and cultural set tings and an individuals behavior in the setting. Such interaction with a setting is related to me mory of an event or a person. These interactions lead to special meanings for a place and as well as place attachment (Steele, 1981). However, an indi viduals interaction w ith a public land setting is controlled to some extent by management actions. Overall, as diagramed by Greene (1999), an individuals experience with physical, social a nd managerial components in a pla ce, lead to a sense of place (Figure 2-5).

PAGE 34

34 Place Attachment Place attach ment is one of the commonly cite d terms within the area of sense of place studies. According to Williams and Vaske (2003), place attachment in environmental psychology is the same as sense of place in human geography Place attachment is often used to encompass a wide range of place-related concepts such as place dependence, place identity, rootedness and satisfaction (Kaltenborn, 1998). However, it has b een used as a narrower concept. For instance, Kyle, Graefe, Manning, and Bacon (2003) pointed out the extent to which an individual values and identifies with a particular setting (p. 250) and that those emotions are necessary for place attachment. Altman and Low (1992) mentioned th at place attachment is multidimensional and inclusive. Place attachment tends to be among the attachments humans experience. In this sense, it would be appropriate to concep tualize place attachment within a larger context of sense of place concepts. In general, earlier works on the domains of emotional, symbolic, and functional attachment of places in natural resource recreation studies exist (Mitchell, et al., 1993; Moore & Graefe, 1994; Williams, et al., 1992; Williams & Roggenbuck, 1989) as well as in other studies for the past few decades (e.g., Proshansky, et al., 1983; Stokols & Shumaker, 1981). Place attachment is generally reported as consisting of two separate dimensions: place dependence, and place identity. Place Dependence Place depen dence was defined as a functional form of attachment an individual has to a particular place or setting where the needs and goals of an individual can be met (Shumaker & Taylor, 1983; Stokols & Shumaker, 1981; Williams & Roggenbuck, 1989). If a particular place where an individual recreates is considered to be better, compared to other available places in terms of their needs and goals to be satiated, an individual tends to have attachment to it

PAGE 35

35 (Shumaker & Taylor, 1983; Stokols & Shumaker 1981). This suggests the significance of a place in offering a particular sett ing necessary for desired activi ties (Stokols & Shumaker, 1981; Vaske & Kobrin, 2001; Williams & Roggenbuck, 1989). In other words, individuals assess the quality of a particular place to some degree, ba sed on the ability of a place to satisfy their activities or goals (Warzecha & Lime, 2001). For instance, some activities such as mountain climbing and whitewater rafting require more res ource specificity or close physical contact with the resource than other passive activities such as hiking and viewing scenery, so that an individual reduces the number of other alternativ e places to achieve their recreational experience in mountain climbing or whitewater rafting (Warzecha & Lime, 2001). The concept of place dependence is similar to that of resource specif icity referring to the importance an individual attaches to the use of a particular recreati on resource (p. 373). The latter term is commonly described in recreation studies on conflict (Jacob & Schreyer, 1980). An earlier view of place identity is from the work of Proshansky (1978). Place Identity Place identity as emotional attachment refers to those dimensions of the self that define the individuals personal identity in relation to the physical environment (p. 155, Proshansky, 1978). Place identity is also defined as a set of individuals memories, conceptions, interpretations, ideas, and relate d feelings about specific physical settings as well as types of settings (p. 60, Proshansky, et al., 1983). The emotional attachment to a particular place or setting may bring a sense of be longing that assigns meanings to an individuals life (Tuan, 1980). Overall, place identity means the emotional and symbolic connection an individual has to a particular place or setting (Williams & Roggenbuck, 1989; Williams, et al., 1992). Stedman (2002) also described the importance of symbo lic interaction between an individual and a

PAGE 36

36 particular setting. As place dependence grows, more frequent visits to a particular place may contribute to an increase in plac e identity (Moore & Graefe, 1994). Domains of Place Attachment or Sense of Place Williams and Roggenbuck (1989) and Williams, et al. (1992) initially developed place attachment scales and found that the primar y components of place attachment are place dependence, place identity, and pl ace indifference. Numerous ear lier quantitative researchers (e.g., Bricker & Kerstetter, 2000; Moore & Graefe, 1994; Warzecha & Lime, 2001) used the first two dimensions and found them to be distinct. More recent quantitative researchers (e.g., Kyle, Graefe, & Manning, 2005; Williams & Vaske, 2003) demonstrated that the validity and reliability of these two distinct dimensions prove d to be adequate. In addition, Mitchell, et al. (1993) also identified the two constructs usi ng grounded theory in their qualitative study. However, other researchers (e.g., Jorgen sen & Stedman, 2001; Moore & Scott, 2003; Stedman, 2003a) supported a unidimensional construct of place attachment. Jorgensen and Stedman (2001) and Stedman (2003a) observed a si ngle domain of sense of place that included attachment (affective), dependence (conative) and identity (cognitive). Stedman (2002) used the single domain of place attachment that was conc eptualized as personal identification with a setting. Moore and Scott (2003) emphasized that one probable reason for th e single construct is that fewer items (8 items) were used in their studies, compared to 28 items developed and used by Williams and Roggenbuck (1989). Overall, there appears to be more evidence to view them as distinct constructs as implied by quantitative and qualitative studies. Unlike other previous studies on place att achment or sense of place, Stedman (2002) conceived of sense of place as a collection of symbolic meanings, attachment, and satisfaction with a spatial setting held by an individual or group (p. 563). Stedman (2003a, 2003b) also took those three wider dimensions of sense of pl ace: place meanings, place attachment, and place

PAGE 37

37 satisfaction. This is based on a social-psyc hological perspective of human-environment relationships and offers much clearer terms a nd empirically more specifiable relationships between variables (Stedman, 2002). This concept of sense of place has not been used in placerecreation management studies except for recen t works by Stedman (2000, 2002). This concept should be also used and tested more in the future (Stedman, 2003b). Place attachment is described as a positive emotional relationship between people and their environment (Low & Altm an, 1992; Moore & Graefe, 1994; Williams, et al., 1992). It is viewed as a core concept of sense of place (Stedman, 2002, 2003a) and is regarded as identity similar to the social psychologica l construct of identity, ones personal location within social life (p. 563, Stedman, 2002). Additionally, Stedman (2002) noted that attachment is based on an emotional bond as well as interac tion between cognition and affect. However, the subdomains of place attachment (dependence, identity) or sense of place (attachment, identity, dependence) are useful for an understanding place-related constructs. The domains are conceptually different from one a nother but they are empirically correlated and do not consider other evaluative domains such as place satisfaction and place meanings (Stedman, 2003b). As stated earlier, Stedman (2002, 2003a, 2003b) suggests that atta chment, satisfaction and meanings should be included in sense of pl ace studies to examine the overall evaluative domains of interaction between in dividuals and a spatial setting. Place Satisfaction Place satisfaction is a multid imensional constr uct because it is a summary judgment of multiple-attributes of a setting such as physical social, and cultural characteristics (Mesch & Manor, 1998; Stedman, 2000). Place satisfaction eval uates the quality of a spatial setting. The place satisfaction construct as attitude is we ll-established in community sociology but not widespread in sense of place studies (Stedman, 2002). It is reported that satisfaction with

PAGE 38

38 physical characteristics or social relationships is positively rela ted to overall satisfaction with community (Fried, 1982; Herting & Guest, 1983). Community satisfaction is different from community attachment, although no clear differen tiation between them is provided yet. For instance, individuals can be satisf ied with their residential area but are not attached to the place in particular. Attachment is positive sentiment that develops toward place while satisfaction is the evaluation of features of the physical and social environment (p. 509, Mesch & Manor, 1998). Clearly, Mesch and Manor (1998) stated that those who express a high level of satisfaction with the physical and social attributes of the local environment also might tend to express a higher level of attachment (p. 506). That is, people with higher positive community satisfaction tend to create higher positive comm unity attachment and reside longer in the community (Mesch & Manor, 1998). This indicates that attachment is a little deeper and symbolically more meaningful than satisfacti on (p. 62, Fried, 1984). Overall, it seems that satisfaction develops attachment at least in th e context of community settings (Mesch & Manor, 1998). In the meantime, Stedman (2003b) made an initial distinction be tween attachment and satisfaction in the context of sense of place studi es in relation to natu ral resource recreation management. Attachment is the importance of a setting to an individua l corresponding to the social psychological construct of identity, whereas satisfaction is an evaluation of the quality of a spatial setting corresponding to the con cept of attitude (p. 825). Basically, Stedman (2003b) argues that place attachment is concep tually and empirically different from place satisfaction. In other words, they are not substa ntially correlated, each is predicted by different variables, and each affects behavior independe ntly. Stedman (2002) observed that those with

PAGE 39

39 high place attachment but low place satisfaction would be more willi ng to protect that place. This also suggests the utility of integrating place satisfaction in to place-related concepts. Place Meanings Another construct that has been ignored in more traditional recreation literature on sense of place studies is meanings about a place. Meanings and attachment, commonly cited in environmental perception studies (Brandenburg & Carroll, 1995; Rel ph, 1976; Tuan, 1975), are not the same but different c onstructs (Stedman, 2003b). Based on social psychology, meanings about a place tend to be descriptive belief, centering on examining what a place means to an individual, instead of how much it means to an individual. Empi rical research has not treated place meanings and attachment for a spatial setting distinctly (Stedman, 2003b, 2002). The separable measures of both constructs s hould not be overlooked. People may have strong attachment to a place but diverse meanings for it. For instance, a forest may mean a wilderness to some individuals, while it may mean a place to make a living to other individuals. This may cause conflict over recreation uses and timber management policies. Therefore, it is crucial to examine descriptive meanings about a place to ameliorate resource conflict (Stedman, 2003). Recent works (Stedman, 2000, 2002) operationalized meanings and attachment for a lake separately and observed different meani ngs that predicted place satisfaction. A setting can contain a particular meaning or multiple meanings. A setting may take on as many various meanings as different people use the setting (Relph, 1976) or common meanings because there are people using a place as fishermen, not as real estate developers (Stedman, 2000). Meanings are created through direct experience in a place, and social relationships in the place. The physical settings or characteristics of environment contribute to generating a sense of place. Individuals appreciate their environmen t through the senses color, texture, slope,

PAGE 40

40 quality of light, the feel of wind, the sounds and scents carried by that wind (p. 38, Ryden, 1993, cited in Stedman, 2000). Shumaker and Taylor (1983) point out the physical amenities of a place that satiate individuals ne eds and attachment to the place. In addition, experiences with place are not only those with landscape but also t hose with other people in the place. Meanings and memories of times spent together with othe r people can lead to a se nse of place (Stedman, 2000). Eisenhauer, Krannich, and Blahna (2000) also found that reasons respondents take on special meanings for a place are related more to social relationships (e.g., family and friend related) than to physical sett ings (e.g., environmental charac teristics). The importance of neighborhood social relationships was also obs erved in a study on community attachment (Mesch & Manor, 1998). Stedman (2000, 2003a) mentioned that physical attributes (e.g., mount ains, lakes) of a place can contribute to attachment and satisfac tion. Stedman also mentioned that environmental features of a place generate meanings for a place that influence attachment and satisfaction. For instance, wilderness experiences are generally not ascribed to deve loped areas. Wilderness meanings are generated by observing natural sc enery but also by experiences in pristine conditions. Summary of Sense of Place Constructs Overall, sense of place as an overarching umbrella concept refers to both affective (attachment) and cognitive (meaning) components of place. It consists of three constructs: place attachment, place satisfaction and place meanings. Place attachment refers to affective bonds between people and places. It typically em ploys two domains, place identity and place dependence or uses items representing emotional ties (place identity) and functional bonds (place dependence). Additionally, as Stedman (2003) suggested, place meaning as a cognitive

PAGE 41

41 dimension, and place satisfaction as an attitude-l ike dimension, should be used and tested in additional sense-of-place studies. Spatial Autocorrelation When similar values of a variable under st udy are closely located to each other on a map, representing that there is a regular pattern from one location to a neighbor ing location in terms of values, a spatial pattern of correlation is present. If the values are interrelated over space, then spatial autocorrelation exists (O dland, 1988). Spatial autocorrelat ion is positive when similar values are clustered together from one loca lity to neighboring local ities and the spatial autocorrelation is negative when dissimilar values alte rnate between adjacent localities (Odland, 1988). In general, spatial autocorrelation measur es the degree to which objects or activities at some place are similar to those at nearby places which reflects Toble rs (1970) first law of geography that everything is related to everything else, but near things are more related than distant things (Goodchild, 1987). Analysis of spatial pattern s has been of great interest to geographers. Geographers measure spatial patterns in numerical ways and test hypotheses about regu lar patterns in space. Other academic disciplines such as ecology, archaeology, epidemiology, sociology and geology have considered analyses of spatial patterns to be important (Odland, 1988). Empirical Literature Review This section provides an em pirical literatu re review to support hypotheses stated in Chapter 1. Empirical findings from the academic a nd managerial literature are introduced in the order of the hypotheses. Objective 1(Model 1): The first structural model includes three major concepts: 1) recreation experience preferences (REP), 2) place meanings (P M), place satisfaction (PS), and

PAGE 42

42 place attachment (PA) within a sense of place framework, and 3) place-protective behaviors (PB). H1: There is a positive association betw een the importance of REP and PM. Evaluative beliefs about places were rega rded as important by Stedman (2000, 2002) and additionally Schreyer and R oggenbuck (1978) observed that individuals who placed higher importance and expectation on recreation experiences (e.g., stress release and solitude), tended to be more sensitive to crowding, while individuals who placed higher importance on more socially-related recreation experiences (e.g., ac tion/excitement, autonomy/achievement, affiliation-being with other people) did not si gnificantly express a perception of crowding. In addition, individuals who rated the importance of wilderness values higher were more likely to be sensitive to crowding. Over all the findings are that perceived crowding varies based on expected recreational experiences. Place meanings are socially constructed and vary in places based on the perspective or motivation of individual visitors (Sack, 1992; Urry, 1990b, 1995). Basically, place meanings are influenced by the importance visitors assign various motivations. Young (1999a) investigated the relationship between tourist motivations and place meanings in the Daintree and Cape Tribulation area in North Queensland, Australia. Young hypothesized that place-specific meanings would be interpreted in several ways by different motivational gr oups in the context of tourism. Conducting a series of semi-structured, stratified-sample interviews, Young identified four clustered motivational groups of participants: interest in nature, escape and relax, social (spending time with friends or relatives, mee ting new people) and pla ce novelty (something to do, hear about the place and want to see it for my self). The meanings attributed to the area by respondents were five themes: natural values (e.g. ecologically important, important to preserve, and unique), aesthetics (e.g., tra nquil, scenic, and wilderness), remote/pristine (e.g., remote,

PAGE 43

43 unspoiled, authentic), cultural value (e.g. importa nt for Aboriginal cu lture, historical, and ancient), and human impacts. These meanings ar e closely similar to symbolic meanings and evaluative beliefs about a place described by Stedman (2000, 2002). Young (1999a) observed significant relationships between motivation and place meaning. More specifically, Young found th at people most interested in natural environments expressed natural values for the tourist area, Daintree and Cape Tribulation area higher than other motivational cluster groups. In addition, people who sought escape and relaxation in a natural environment, ascribed the area aesthetics (scen ery) higher, compared to other motivational groups. Contrary to the finding of Prentice, Witt, and Hamer (1998), the findings in this research support that gazers as tourists, view a nd define the object of place differently. Another aspect of significant relationships between expected motivation and meanings is that places are constructed or invented before they are visited. Tourists tend to decide which meanings are important, interpreting various plac es. That is, meanings tend to be implicitly credited with specific places that are suitable fo r behavior (Echtner & Ritchie, 1993; Hunt, 1975; Gartner & Hunt, 1987; Pearce, 1982; Young, 1995; Walmsley & Young, 1998, 1999a). In this sense, Young (1999a) argues that a significant as sociation between motivation and meanings demonstrates partial construction of place meanings before actual experiences. Due to indirect experience through mass media, visitors may be fa miliar with the Daintree and Cape Tribulation area, although they spen t less time in the area. Interestingl y, this is opposite to Tuans (1974, 1977) suggestion that place meanings are constructe d by individuals direct experience in places. More evidence on the relationship between motiva tion and meanings should be documented in future research.

PAGE 44

44 H2: There is a positive association betw een the importance of REP and PS. There is little empirical evidence available to support a relationship between the importance of recreation experiences and place satisfaction. Marketing fields used either performance (satisfaction) scores or gap scores to measure over all satisfaction of service quality. In the outdoor recreation field, Burns, et al. (2003) employed importance and satisfaction (performance) scores of recreat ion settings (facilities, services information) and recreation experiences (crowding, compatibil ity, and conflict) to measure visitors overall satisfaction (quality of recreation ex perience) in ten U.S. Army Corps of Engineers lakes. Burns, et al. observed that the satisfaction-only measure was a stronger predictor than gap scores. They suggested that future studies should add additional recreation experience domains (e.g., social gatherings, escape from daily lif e), when measuring satisfacti on with individuals recreation experiences. Expectation (or importance) scores can be used in measuring satisfaction levels, although it is not necessary (Babakus & Boller, 1992; Carman, 1990). Therefore, a part of the purpose of this study was to examine the effect of the importance of recreation experiences on visitors satisfaction with places. H3: There is a positive association betw een the importance of REP and PA. Using the expectancy value model of motivation, Kyle, et al. (2004) hypothesized relationship between recreation motivation a nd place attachment. The model is based on Lawlers (1973) theory of mo tivation developed for work or ganizations. It posits that expectations of positive outcomes (e.g., psychol ogical, social, and physiological) push people toward specific natural settings Kyle, at al. (2004) examined wh ether expectations or pursuit of such outcomes would foster humans attachment to natural settings. Performing exploratory factor analyses (EFA) for grouping REP scal e items and place attachment scale items, confirmatory factor analyses (CFA) for measurem ent models, and structural models, the authors

PAGE 45

45 observed that all dimensions of motivation di d not predict place attachment. Their findings demonstrated partial support for the hypothesize d model, indicating that the expectation of personal benefits or outcomes c ontributed to the partial develo pment of attachment to place. Regarding significant predictors of place att achment dimensions, place dependence (functional attachment) was positively predicted by health, a ffective attachment was predicted by autonomy, nature, and health, place identity was predic ted by learning and autonomy, and social bonding was predicted by social and nature. Warzecha and Lime (2001) studied the relationships between user motives and place attachment of Colorado River and Green River (m ore natural and more restricted, compared to Colorado River). Warzecha and Lime observed th at visitors with higher place attachment considered nature-oriented motives or experiences to be important reasons for visiting the rivers. Those motives included experiencing wildlands, escaping physical pressure (e.g., experiencing solitude, being away form other people), enjoying nature (e.g., being close to nature), learning, and experiencing introspection. Respondents with low levels of attachment expressed more importance to socially-oriented experiences (e .g., family togetherness). Contrary to these previous studies, Mesch and Manor (1998) studying community-based attachment, observed that as the number of friends living in neighborhoods increased and the relationship with neighbors got closer, residents felt more attached to their neighborhood-feeling more pride in the neighborhood. H4: There is a positive association between PM and PS. Place meanings that are considered as cogni tive domains of sense of place consist of symbolic meanings and evaluative beliefs. Symbo lic meanings represent what types of places a forest represents. Relatively specific evaluative be liefs about specific place or setting attributes were treated as evaluative meanings for a pla ce (Stedman, 2000, 2002). As lakes meant pristine

PAGE 46

46 (e.g., evaluative abundance of wildlife and plants, forested shore) and u p north (e.g., symbolic escape from civilization, pristine wilderness) meanings to respondents, respondents satisfaction with the lakes increased. But if resp ondents considered thei r lakes to be impacted (e.g., evaluative more houses or cabins on shore, many lake recreationists), their satisfaction with the lakes decreased (Stedman, 2002). Typically, overall recreation satisfaction is predicted by tw o major factors: situational factors and subjective factors. Si tuational factors refer to specific recreation setting or activity attributes and includes presence of litter, biting insects, number of other users, and conditions of whitewater rapids, which affected satisfaction (Dorfman 1979, Graefe & Fedler, 1986; Herrick & McDonald, 1992; Peterson 1974; Whisman & Hollenho rst, 1998). Subjective variables refer to evaluative variables such as perc eptions of crowding, fulfillment of motives, and perception or evaluations of resource impacts, which affected visitor satis faction (Graefe & Fedler, 1986; Whisman, & Hollenhorst, 1998; Williams, 1989). Usi ng data collected from commercial and private whitewater boaters right after their ri ver trips on the Cheat River of West Virginia, Whisman, and Hollenhorst (1998) observed that us e level influenced crowding perceptions that negatively predicted overall satisfaction. Additionally positive environmental conditions of forest and land were reported to be significa nt predictors of overall satisfaction. Similarly, earlier researchers (e.g., Floyd, 1997) on satisfaction emphasized that it is important for recreation managers to be info rmed or aware of the physical, social, and managerial conditions of ROS-re lated recreation sett ings that contribut e to park visitor satisfactions. For instance, managerial conditions such as low level of facility development, pleasant behavior of others, and good trail c onditions were positively related to visitor satisfaction (Shelby, 1980). Other diverse mana gerial variables (e.g., upkeep of facilities,

PAGE 47

47 information signs, upkeep of the grounds, information sources, park staff) also influenced visitor satisfaction (McGuire, OLeary, & Dottavio, 1989). More recently, us ing the data collected from 156 diverse state park visitors, Fletcher and Fl etcher (2003) found maintenance of the park (e.g., cleanliness, absence of litter, natural environment), beha viors of personnel (e.g., prompt and courteous staff), and information (e.g., signs an d information) to be important managerial predictors of park visitor sa tisfaction. Additionally, annoyances (bothering pets, noisy people, crowded) as a social factor was an important predictor of park visitor satisfaction. According to Shelby (1980), however, density, interaction, and perceived crow ding in wild and backcountry areas did not seem to affect respondents overall tr ip measured as a single factor of satisfaction. Overall, the two evaluative factors (managerial, social) were more strongly related to park visitor satisfaction. H5: There is a positive association between PM and PA. There is not much literature to support a possible relationship between place meanings and place attachment except earlier works of Stedman (2000, 2002, 2003a). Place attachment as an identity-like dimension is considered to be based on meanings. Stedman (2002) pointed out a lack of empirical research on m eanings and the abundance of resear ch on identities. That is, the beliefs one has about a spatial setting should be measured (Stedman, 2002, 2003a). Stedman (2003a) observed that spatial meanings of escape from civilization for lakes, are positively related to attachment and social or more develo ped meanings for lakes also fostered attachment. Respondents who considered the lake as wilderness were more likely to be attached to the lake than those who considered the lake as a place for sociability. H6: There is a positive association between PS and PA. Mesch and Manor (1998) argued that the conc ept of place attachment is different from that of place satisfaction. The former refers to a positive emotional bond toward place. The latter

PAGE 48

48 refers to a subjective evaluation of the physical and social environment and is based on a relative perception of a setting or place. In this sens e, neighborhood satisfacti on is a multidimensional concept encompassing a variety of subjective per ceptions of place attributes such as physical, social and economic characteristics. Residential or neighborhood satisfaction is an antecedent of place attachment (Mesch & Manor, 1998). This means that place attachment is treated as a larger construct than place satisfaction and is formed through direct experience with and satisfaction with physical and social setti ngs of the local environment. Adapting two theoretical perspectives, comm unity of limited liability, and liberated community, and using previous studies on re sidential satisfaction, Mesch and Manor (1998) observed that as respondents were more positiv ely satisfied with the quality of the physical environment in a neighborhood (e.g., noise level, air pollution level) a nd satisfaction with the open space in the neighborhood (e.g., parks, playgr ounds, parking spaces), they were more likely to be attached to that place. In addition, Mesch and Manor (1998) found social characteristics (e.g., type of people and number of people livi ng in neighborhood) affected place attachment. Overall, Mesch and Manor (1998) showed that place attachment is formed through direct experiences with the neighborhood environment, and a higher level of place attachment is predicted by a positive judgment (evaluation) of the local environmen t (place satisfaction). H7: There is a positive association betw een the importance of REP and PB. It is expected that the importance of r ecreation experience preferences would predict place-protective behavior. That is, it was hypothesi zed that visitors who have higher preferred recreation experiences in natural resource areas would be more wi lling to show place-protective behavior against potential future changes in recreation settings.

PAGE 49

49 H8: There is a positive association between PM and PB. There is little evidence to support the re lationship between place meanings and placeprotective behavior. Smaldone, et al. (2005) utilized a pooled cr oss-sectional mixed approach (closed-quantitative and open-ended/more qualitative) and asked respondents why they have special reasons or meanings for places in Gr and Teton National Park (GTNP). Analyzing data from 493 responses to assess and analyze place attachment and critical issues in GTNP, Smaldone, et al. (2005) found that visitors holding a special place (place attachment) were more likely to express a negative impact for a gr eater number of activities and growing built environment (e.g., hunting, grazing, a commercial air port, private property and houses, and the existence of a dam). In general, more than one place meaning is reported by visitors. That is, a variety of meanings for places in GTNP are gr ouped into the most frequent meanings (e.g., physical setting, outdoor recreation activities, emo tional connections, wild life viewing, getting away, social aspects, and solitude). Visitors wi th higher levels of such place meanings (esp. getting away, solitude, and emotional connections) were more likely to be groups that perceived those issues to be negative (Smaldone, et al. 2005). One earlier and more specific study found that symbolic meanings predicted place-protective behavior. In detail, respondents considering their lakes to be environmentally pristine (up nort h), were more willing to participate in placeprotective behavior, but respondent s believing the lake to be a community of neighbors were less likely to express their willingness to participate in such behavior. The dimensions of specific evaluative beliefs (pristine and impacted factors) did not predict such behavior (Stedman, 2002). These findings indicate that th ere is an empirical relationshi p between place meanings and behavior toward places. This is also implied by Davenport (2005). In addition, Davenport (2005) explored the pl ace-based meanings that local community members have to the Niobrara River and exam ined their perceptions of landscape change,

PAGE 50

50 especially towards developmen t in the river valley. Unlike many previous studies, Davenport used an interpretive research approach to get in-depth meanings of places and a network (snowball) sampling method. Davenport found that re spondents had four categorized dimensions of meanings for the river sustenance (water, economic), tonic/dependence (enjoyment, solitude, freedom), nature (ecology, undisturbed), and identity (individual, family, community). Sustenance and nature meanings were considered as additional dimensions of sense of place. Davenport seemed to use place-based meanings as place attachment within sense of place constructs because the terms were used inte rchangeably. As Davenport mentioned, additional research should clarify respondent s perception of or attitudes to wards development at the river, and identify the relationship between perceived belie fs and future intentions. Therefore, a part of the purpose of this study was to test the relationship between place meanings and placeprotective behavior. H9: There is a positive association between PS and PB. Place satisfaction is an attitude-like dimens ion in sense of place studies. Not much recreation literature exists about the effect of place satisfaction on respondents behavior towards places they visit. However, the linkage between attitude and behavi or is very common in social and psychological research. Peopl e who hold positive attitudes should engage in behaviors that approach, support, or enhance th e attitude object, and people who hold negative attitudes should engage in behaviors that avoi d, oppose, or hinder the object (Eagly & Chaiken 1993, 155). This predictive relationship between th e two is also supported by the th eory of reasone d action (Ajzen & Fishbein, 1980; Fishbein & Aj zen, 1975) and the theory of planned behavior (Ajzen, 1991a, 199b). Using the predictive behavioral measures with attitude measures, Stedman (2002) observed that respondents satisfied with place at tributes such as scenery and water quality (physical), solitude and number of users (social) and level of development (managerial) did not

PAGE 51

51 show an intention to participate in place-protecti ve actions, while respondents strongly attached to a place would participate in such actions. Additional evidence to support the relationship of place protective behavior to recreation-related behavi or would be helpful. Therefore, a part of the purpose of this study was to test the relationshi p between place satisfac tion and place-protective behavior. H10: There is a positive association between PA and PB. Many recent empirical studies have documented support for including place attachment when examining different types of environmen t-related impacts (Patterson & Williams, 1991; Smaldone, et al., 2005), social-related impact s (Smaldone, et al., 2005; Patterson & Williams, 1991; Williams, et al., 1992), attitudes toward s development (Vorkinn & Riese, 2001), acceptability of social and environmental condit ions (Kyle, et al., 2004b), and place-protective behavior (Stedman, 2002). Wilderness users with higher levels of place attachment were more inclined to express greater sensitivity to diverse impacts such as ecological, visual, and audio (Patterson & Williams, 1991). That is, place attachment is related to users sensitivity to ecological impacts, social impacts (horse or hiker encounters), and si ght/sound intrusions (Williams, et al., 1992). Attachment to areas influenced by a propos ed development (e.g., hydropower development plans) was observed to be one of the stronger pr edictors of attitudes towards this development (Vorkinn & Riese, 2001). Similarly, Smaldone, et al. (2005) observed that Grand Teton National Park users, expressing it as a spec ial place (place attachment), were more inclined to be aware of critical issues (e.g., a number of activities and growing built environment such as hunting, grazing, a commercial air port, private property and houses, and the existence of a dam) in the park. In addition, visitors with a special place attachment were mo re likely to consider critical issues as negative impact issues than those that did not have a sp ecial place attachment.

PAGE 52

52 Furthermore, place attachment has been linked to willingness to protect natural resources (e.g., Kaltenborn, 1998), environmentally responsi ble behavior (e.g., Vask e & Kobrin, 2001) or other similar constructs. As indi viduals developed an emotional c onnection to their local natural resources, they appeared to act more responsibly in day-to-day environm ental activities as well as toward the environmental setting (Vaske & Kobrin, 2001). For instance, Kaltenborn (1998) observed that respondents with a strong sense of pl ace tended to be more willing to contribute to solutions to environmental problems such as crude oil along parts of the shoreline and insufficient supervision of natura l areas, compared to their counterparts. The former respondents would tend to have pro-environmental intentions such as picking up lit ter, not littering, not visiting environmentally damaged places undergoi ng recovery in park, and even support for higher entrance fees to prevent poaching (W alker & Chapman, 2003). Similarly, management actions favoring protection of environmental resour ces such as cultural monuments, wilderness and untrammeled nature were reported to be more important to highly attached residents, whereas facilitation of hunting and fishing was less important to them (Kaltenborn & Williams, 2002). In the context of more specific recreati on settings, Warzecha and Lime (2001) studied place attachment of users of the Colorado and Green rivers (more natural, less motorized, compared to Colorado River), in Colorado, in relation to acceptability of encounter norms. Users with high place identity consider ed the lower maximum number of watercraft in rivers more acceptable to them. Furthermore, visitors with high place identity in a more natural and less motorized place (or setting, Green River) rate d lower maximum encounter levels to be acceptable than those in a more developed and less restrictive place (or setting, Colorado River). Using social judgment theory that guides an unde rstanding of how people ev aluate or react to

PAGE 53

53 new stimuli (e.g., setting density), and two typica l dimensions of place attachment (dependence and identity) plus social boding, Kyle, et al. (2004a) observed that place identity was more associated with negative percepti ons of or less acceptability to setting density (e.g., acceptability of the number of encounters, le vel of feeling crowded, agreement on the right number of people on the trail) along the Appalachian Trail, while place dependence was less associated. The latter finding indicates that for place dependent responde nts, social impacts do not hinder the settings capacity to provide users desired recreation e xperiences. Interestingl y, contrary to their hypothesis, social bonding with othe rs was not a significant factor in predicting perceptions of setting density. The authors suggest the clarification of who others are (e.g ., within group or not within group) as a possible explanation. Using the same theory and two dimensions of place attachment (dependence and identity), Kyle, et al. (2004b) observed that visitors with higher place identity tended to be more critical of social conditions (e.g., pe rceived crowding, use conflict, and depreciative behavior) and environmental conditions (e.g., use impact, trail development, and human urban encroachment). Overall, place identity respondents raised support for place-protective mana gement conditions or actions. On the other hand, respondents with high place dependence tended to support development of facilities such as amenities a nd trails (Bricker, 1998). Wickham and Kerstetter (2000) observed that place attachment was positivel y associated with visitors attitudes toward crowding in an event setting. Similarly, Mowen, Graef e, and Virden (1998) also observed that an individual with place attachment had more posit ive evaluations of the setting and recreation experience. Considering theoretical aspect s, it should be noted that Kyle, et al. (2004b) examined the relationship between place attach ment and acceptability of recrea tional conditions, using social

PAGE 54

54 judgment theory. Additionally, Stedman (2002) borrowed the theory of reasoned action and planned behavior (Ajzen, 1991a, 1991b; Fishbein & Ajzen, 1975), handled place attachment as an identity-like dimension, and explained the relationship between place attachment and placerelated behavior. Stedman observed that respond ents who were strongly attached to a place expressed their active willingness to support the place. Objective 2 (Model 2): The second structural model focu ses on two parts: 1) the impacts of PM on the attainment of REP and 2) the impacts of REP attainment on PS, PA, and PB. H11: There is a positive association betw een PM and the attainment of REP. No previous literature has reported a dire ct linkage between place meanings and the attainment of recreation experien ce. As mentioned earlier, however, place meanings in sense of place studies are categorized into two major cognitive dimensions: symbolic meanings and evaluative beliefs. The former is related to broa d meanings for a place an d the latter appears to evaluate how respondents perceive a place or site in relation to current conditions of physical, social, and managerial elements (Stedman, 2000, 2002), if the elements are considered a part of recreation management settings. This measurement perspective tends to link place meanings with attainment of recreation e xperience. As suggested by e xperience-based management, biophysical, social, and managerial settings are related to recreat ion experiences. The conditions of those settings can facilitate or hinder the attainment of desired experiences. Floyd (1997) mentioned that it is important for recreation managers to be in formed about the physical, social, and managerial conditions of recreation settings that influence visitors desired experiences and park visitor satisfaction. As mentioned earlier, situational and evaluative variables are typically used in recreation literature that measures overall sa tisfaction of visitors. For instance, using the data collected from

PAGE 55

55 commercial and private whitewater boaters right afte r their river trips on th e Cheat River of West Virginia, Whisman and Hollenhorst (1998) observed th at water flow levels (situational variable) was positively related to the attainment of recreational experiences such as adventure-related aspects (e.g., challenge, excitement, skill testing) that also affected overall visitor satisfaction. These findings supported their proposed model that shows linkage between situational variables and overall visitor satisf action indirectly through evaluative variables. H12: There is a positive association betw een the attainment of REP and PS. According to Schreyer and Roggenbuck (1978), visitors engage in recreation with some expectation that they will fulfill desirable c onditions and outcomes. Borrowing Lawlers (1973) expectancy-value model of motivation, Kyle, et al. (2004) observed the partial support for the contribution of expectations of personal benefi ts/outcomes to the development of visitors attachment to natural settings. Within the same framework of ROS adopting expectancy valence theory, however, it is assumed that if these outco mes are not attained, indi viduals are dissatisfied with their recreation experience (Driver & Brown, 1978). Shelby (1980) observed the positive relationship between recreation be nefits and satisfacti on with wild and backcountry recreation on Grand Canyon river trips. As re spondents achieved personal bene fits (e.g., learning, personal growth) from recreation experiences they tended to be satisfied with their overall trip measured as a single factor of satisfaction. Using a genera l recreation satisfaction model with situational and subjective variables, Wh isman and Hollenhorst (1998) al so observed that recreation satisfaction was affected by the fulfillment of r ecreation experiences such as adventure-related factors (e.g., challenge, excitement, skill tes ting), escape from life demands and social experiences. Similarly, in the analysis of a disc repancy model measuring satisfaction with overall experiences, Burns, et al. (2003) observed that satisfaction with recreational experiences (e.g., opportunity to recreate without feeling crowded, opportunity to recreate without interference

PAGE 56

56 from other visitors, compatibility of recreati on activities at the area, and places to recreate without feeling conflict from other visitors) posi tively affected levels of visitors satisfaction with overall visits in natural areas. Burns, et al. (2003) also suggested that more recreation experience domains (e.g., social gathering, and es cape from daily life) should be included in measuring satisfaction. In these contexts, a part of the purpose of this st udy was to examine the relationship between the fulfillment of expressi ve recreation experience preferences and place satisfaction in the context of sense of place studies. H13: There is a positive association betw een the attainment of REP and PA. No previous literature has documented em pirical support for the direct relationship between the attainment of recr eation experience outcomes and place attachment. Using Lawlers (1973) motivation theory, Kyle, et al. (2004) obser ved a direct linkage be tween the expectations of recreation experiences and place attachment to natural settings. That is, based on expectations, respondents showed their increasing attachment to natural settings. However, as ROS framework adopting expectancy valence theory assumed, if respondents do not fulfill their desired recreation outcomes, they would be dissatisfied (Drive r & Brown, 1978). Additiona lly, individuals who attained their desired recreation experience in a place, would display their increasing attachment to the place, calculating that the place fits th eir recreation pursuits. On the contrary, their counterparts would not identify the place to be su itable for their recreation pursuits. Along this line, constraints interfering with attaining recreation experiences would not foster the development of attachment to natural settings. Therefore, a part of the purpose of this study contends a possible link of expect ancy disconfirmation theory to pl ace attachment concepts. That is, this study was to explore the association of in dividuals attainment of REP with the development of setting attachment.

PAGE 57

57 H14: There is a positive association betw een the attainment of REP and PB. No recreation literature exists about th e prediction of the attainment of REP on respondents behavior towards places they visit. However, when individuals fulfill their desired experiences in a place, they ar e satisfied with their visits or place attributes (Shelby, 1980; Whisman & Hollenhorst, 1998). If they are satisfie d with the place, they are less willing to engage in place-protective behavior. This is supported by the theory of reasoned action and planned behavior. In a similar sense, if res pondents whose primary motivation is related to engaging in more nature-oriented or social experiences did not at tain their desired experience, they would be expected to support the conditions of place settings that could facilitate their desired recreation experience. For instance, if vi sitors seek solitude in a place but they did not achieve that experience, they would strive for so litude and engage in management actions that reduce the number of visitors in the place. If their primary motivati on is to bond with other friends, family or other visitors, then they may no t consider such actions to be favorable to their recreation pursuits. There should be an explorat ory study to support the re lationship between the attainment of recreation experien ces and place-related behavior. Th erefore, a part of the purpose of this study was to test the relationship between them. Objective 3 (Model 3): Similarly, the third structural m odel focuses on two parts: a) the impacts of PM on GAP scores, and b) the impacts of GAP scores on PS, PA, and PB. As mentioned earlier, there are different ways to predict visitor satisfaction from recreation settings and experien ces. Most research has used performance scores only, and difference in scores between importance and pe rformance (attainment) called as GAP, while importance can be used. Performance scores were better predictors th an disconfirmed scores when predicting overall satisfa ction (Burns, et al., 2003). This section employs gap scores

PAGE 58

58 between the importance of recreation experiences and the fulfillment of recreation experiences, examining the effects of place satisfaction, place attachment, and place-related behavior. H15: There is a positive association between PM and GAP scores. Similar to the hypothesized relationship between place meanings and attainment of REP, it was expected that place meanings are positively re lated to gap scores of recreation experiences. H16: There is a positive associatio n between GAP scores and PS. It was expected that gap scores of recreation experiences are positively related to place satisfaction. Individuals who atta in their recreation experiences more than they expected may display place satisfaction, which is similar to the expected relationship between the attainment of recreation experiences and place sa tisfaction. Similarly, Burns, et al. (2003) observed that gap scores of recreation experiences (e.g., compatibility of recreation activ ities, no conflict from other visitors), along with facilities, services a nd information domains were positively related to satisfaction. H17: There is a positive association between GAP scores and PA. It was also expected that if attained scores are higher than expected scores in terms of recreation experiences, then it would also foster the development of attachment to a place. H18: There is a positive association between GAP scores and PB. By the same token, higher scores on the atta inment of recreation experience than on the importance of recreation experience may be more positively related to pl ace-related behaviors. No literature exists on the influence of GAP scor es on PA and PB, each. A part of the purpose of this study was to explore the relationship between them.

PAGE 59

59 H19: There is no spatial autocorrelation in reg ression model residuals in predicting the effects of the importance of REP, PM, PS, and PA on PB. It was expected that if there is no spatial autocorrelation in regression model residuals, variables in the model would be specified correctly to interpret the parameters in the model. H20: There is no spatial autocorrelation in reg ression model residuals in predicting the effects of the attainment of REP, PM, PS, and PA on PB. It was expected that if there is no spatial autocorrelation in regression model residuals, variables in the model would be specified correctly to interpret the parameters in the model. H21: There is no spatial autocorrelation in reg ression model residuals in predicting the effects of gap scores of RE P, PM, PS, and PA on PB. It was expected that if there is no spatial autocorrelation in regression model residuals, variables in the model would be specified correctly to interpret the parameters in the model.

PAGE 60

60 Recreation activities (fishing, canoeing, etc) + Settings (physical, social & managerial) = Experiences (seek solitude, learn about nature, etc.) => => Benefits/outcomes (individual, community, economy & environment) Note. From Haas, Aukerman, Lovejoy, & Welch (2004) Figure 2-1. Framework of ROS planning system Figure 2-2. Expectancy theory: Motiva tion for nature-based recreationists Figure 2-3. Effect of gap score on satisfaction Figure 2-4. Predicting behavi or from sense of place PM: Meanings ( belief ) PB: Place protective b ehavio r PS: Place satisfaction PA: Place Attachment Satisfaction Expectation Satisfaction Gap scores Satisfaction Performance Level of visitor satisfaction Visitor expectations (motivations) Perception of resource conditions

PAGE 61

61 Figure 2-5. Model of place relationships Physical Setting Social Setting Managerial Settin g Setting Sense of Place Person

PAGE 62

62 CHAPTER 3 METHODOLOGY Description of the Study Site The Ocala National Forest, located in central Florida, is the southernmost National Forest (NF) within the continental United States. It was established in 1908. It is estimated that almost 8 million recreation visitor days occur each year in the Ocala NF. The forest encompasses about 383,000 acres of sand, longleaf and slash pine tr ees along with cypress and hardwoods. The forest provides more intimate, natural and remo te recreation experiences in major recreation areas such as Lake George Wild erness, Juniper Prairie Wilderne ss, Billies Bay Wilderness, and Alexander Springs Wilderness within the Ocala NF. Diverse recreational activities such as hiking, canoeing, and swimming are offered in f our major spring areas: Alexander Springs, Juniper Springs, Silver Glen Springs, and Salt Springs. Camping is offered at all the major springs except Silver Glen Springs ( http://www.fs.fed.us/r8/florida/). Off-Highway Vehicles (OHV) riding is allowed along a bout 80 miles of designated mixed-use trails stretching within the forest with other areas off limits to protect some sensitive areas and endangered species. The major designate d areas for OHV use are located at Big Scrub areas (campground, trailhead), Wandering Wi regrass Trailhead, Lake Delancy areas (campground, west trailhead, east trailhead), and Rodman Trailhead (Figure 3-1). Big Scrub areas and Wandering Wiregrass Trailhead are locate d in the southern part of the forest, while Lake Delancy areas and Rodman Trailhead are locat ed at the northern part of the forest. Lake Delancy campground tends to provide more natural settings, compared to Big Scrub campground. OHV users need to comply with certai n rules and regulations such as adhering to speed limits and paying recreation fees. Recrea tion fees have been charged for OHV use on the forest since October 1, 2007 ( http://www.fs.fed.us/r8/florida/).

PAGE 63

63 Data Collection This study employed two phases of surveys. Phas e 1 used an on-site survey as an initial contact and Phase 2 used a m ail survey as a fo llow-up. The primary reason for implementing two phases was due to length of the questionnaire in a one survey form at and the types of questions asked in the survey discussed later in this methodology section. Furthermore, the follow-up mail survey mode took into consideration users that did not have access to computers, was more personal than an internet survey, and was more reasonable in terms of cost, compared to a telephone survey. On-site data for this study was collected fr om users of the Ocala National Forest (NF) between March and May, during the spring of 2008. A random sampling technique was stratified to obtain a representative sample of users th at visited four major OHV designated trailheads (Rodman Trailhead, Lake Delancy Trailheads, Big Scrub Trailhead, Wandering Wiregrass Trailhead) and OHV campgrounds (Big Scrub, Lake Delancy), and four major springs with campgrounds and swimming areas (Alexander Springs Juniper Springs, Silver Glen Springs, Salt Springs), including a canoe pi ck-up site at Juniper Springs (Babbie, 1995). The sampling of OHV users and spring users was proportionately st ratified by recreat ion sites, based on general use estimates provided by forest recreation managers. Stratified random sampling increased the likelihood of representativeness of all users within the selected r ecreation areas of Ocala NF if a sample is not large because characteristics should be represented proportionately in the sample. More importantly, stratified sampling should resu lt in smaller standard errors, especially compared with a simple random sampling. Users were intercepted at OHV trailheads, campgrounds, swimming areas, and canoe trail exit areas by trained and paid staff from the Department of Tourism, Recreation, and Sport Management and the School of Forest Resources and Conservation at the University of Florida

PAGE 64

64 and users were requested to fill out an onsite survey voluntarily. The format was a selfadministered paper-pencil surve y. After users had completed the onsite survey, they were asked to participate in the follow-up mail survey. If th ey agreed to do so, they were asked to provide their name and address to be sent a mail surve y. The format was a self-a dministered paper-pencil survey of mail participants. The on-site users survey included the follo wing sections: 1) forest use profile, 2) importance of recreation experiences, 3) demographic profile, and 4) contact information (e.g., names, mailing and email addresses). The survey instrument included typical evaluation type questions that should be non-threatening to the re spondents (See Appendix, p.150). For the follow-up mail survey, the research er contacted users three times, having an initial contact letter with questionnaire sent ou t in a business-reply mail envelope (about two weeks after the onsite contact), a reminder/thank you postcard (2 weeks after the first contact mailing), and a cover letter with a final copy of the replacement questionnai re (2 weeks after the reminder) to non-respondents, following modifi ed Dillmans (2000) recommendations to increase response rates. The follow-up mail questionnaire included the following sections: 1) attainment of recreation experiences, 2) place meanings, pla ce attachment, place satisfaction, and 3) placeprotective behavior. The survey instrument included typical recr eation experience type questions that should be non-threatening to the respondents (See Appendix, p.150). Description of Variables and Instruments This section introduces the major variables used in this study. The variables included REP, place meaning, place attachment, place satis faction, and place-prot ective behavior. REP

PAGE 65

65 and place attachment are major measurement scales developed and tested by previous researchers. Importance and Attainment of REP Individuals engage in recrea tion activities to pursue certain psychological and physical goals. This is conceptualized as motivation theo ry (Driver & Tocher, 197 0), thereby contributing to developing the Recreation Experience Preferen ce (REP) scales to measure the dimensions of individuals recreation e xperiences (Driver, 1983). A master li st of 328 items with 19 domains of REP scales was developed by Driver (1983). To make REP scales valid and reliable, a list of items were made by reviewing the motivation literature and open-e nded discussion with recreationists and recreation literature. The c oncurrent validity of the REP scales was demonstrated by earlier studi es (e.g., Beard & Ragheb, 1983; Tinsley, Driver, & Kass, 1982). Beard and Ragheb (1983) identified an initial pool of more than 150 items derived from existing literature and checked with students and professo rs to evaluate whether each of the items was relevant to leisure motivation and to clarify th e items. After this step, 106 items remained. The authors administrated the 106 item in strument to 65 students to eval uate any flaws in items with 5-point Likert scales and instructions. After this step, 174 studen ts were asked to rate the 103 items. Using exploratory factor analysis, the authors observed accep table alpha reliabilities of 4 extracted factors ranging from .90 to .92. The average inter-item correlation of .4 or greater was acceptable and Cronbachs alpha of .60 or greater was acceptable (Driver, Tinsely, & Manfred o, 1991). Weissinger and Bandalos (1995) reported acceptable internal reliability levels of these items administered to diverse populations in the United States and Canada. More recently, Manf redo and Driver (1996) used a meta-analysis of data that were collected from 36 different studies conducted in the 1970s and contained most of the REP scales similar to those included by Driver (1983). Implementing

PAGE 66

66 confirmatory factor analysis, the authors demonstrated the reliability and construct validity of the REP scales. Respondents might be asked to indicate how important each of the REP scales were when they decided to visit or engage in an activity, while they can be asked to indicate how much each REP item added to their trip if the study aime d to measure recreation experience attainment (Manfredo & Driver, 1996). It is recommended th at experience attainment should be measured immediately after the end of their trip, while recreation e xperience preferences should be measured before their trip (Holland, 1979; Manfredo, 1984; Manfredo & Driver, 1996). The reason is that if recreation experience preferen ces are asked after individuals experience, their answers may be confounded by experience attainment. This study included 24 REP items adopted from Driver (1983) and Manfredo and Driver (1996) to measure recreation experience prefer ences and attainment. Manfredo and Driver (1996) suggested that all REP s cales should be included in the pretest and the items that are important and show high variabil ity should be included in the fi nal instrument. More than one item from each factor should be used to reduce item sampling error and increase generalizations of recreation experiences. However, the numbe r of items included in this study took into consideration the length of the original REP scales develope d by Driver (1983) and use of past research in similar contexts. Smaller sets of item s or certain domains pertinent to their respective research have been typically chosen by resear chers (Graefe, Thapa, Confer, & Absher, 2000). Recreation experience preferences were rated on a 5-point Likert scale ranging from not at all important to extremely important and recreati on experience attainment was rated on a 5-point Likert scale ranging from did not attain to totally attained. REP scales were initially developed to measure outdoor recreation experiences in natural settings so, using REP scales for

PAGE 67

67 this study was expected to provi de reasonable content validity (Manfredo & Driver, 1996). A gap score variable was obtained afte r subtracting importa nce of recreation experience preferences from attainment of recreation experiences. Appendi x provides a full list of items that were be used in this study. Place Meanings A place meaning scale consisting of both sym bolic meanings and evaluative beliefs was initially developed and used by Stedman (2000). Eight items were used to assess symbolic meanings (what kind of place is your lake?) such as a place for family, a place to escape from civilization, and a pristine wilderness et c. Lake property owners in Northern Wisconsin were asked to rate each of 8 items for the meani ngs they held for their lake on a 5-point Likert scale anchored from strongly disagree to stron gly agree. After retain ing items with factor loadings of more than .30, factor analysis produced two-factor solutions that shared 55% of the variation. One factor is called up north with 4 items referring to the naturalness of the area and the other factor is called comm unity with 2 items related to residential areas. Reliability analysis revealed good reliable standardized Cro nbachs alpha coefficients of .824 for up north and .680 for community. Additionally, there was a moderately negative correlation between the two ( r = -.135). To measure the evaluative beliefs about the la ke (what is true a bout the lake), Stedman (2000) used 14 items of lake elements such as scenic lake and peaceful lake referring to the environmental quality of the lake. The items were rated on a 5-point bipolar scale anchored such as extremely scenic to not at all scenic. After retaining items with factor loadings of more than .30, factor analysis produced two-factor so lutions that shared 58% of the variation. One factor is called pristine with 6 items referring to a more environmentally sound area and the other factor was called impacted with 5 items referring to more developed areas. Reliability

PAGE 68

68 analysis revealed high reliable standardized Cronbachs alpha coefficients of .806 for pristine and of .828 for impacted. Additionally, there was a negative correlation between the two (r=.279). For the study here, the survey included place meaning items initially developed and used by previous research (Stedman, 2000) on a 5-point Likert scale ranging from strongly disagree to strongly agree to measure abstract symbolic meanings and on a 5-point bi-polar scale for specific evaluative beliefs about place attributes. In addition, this study included a few more items from other related studies (e.g., Relph, 1976; Smaldone, et al., 2005) that relate to the current studys contexts to provi de more variation in place meani ngs, compared to percentage of scale variation (55%) initially observed by Sted man (2000). The appendix provides a full list of items that were used in this study. Place Attachment Place attachment was initially measured and developed by Williams and Roggenbuck (1989). After reviewing the place attachment lite rature in environmental psychology (Proshanky et al., 1983; Stokos & Shumaker, 1981) and activity involvement lite rature in leisure behavior (Wellman, Roggenbuck, & Smith, 1982) and consulti ng with other researchers, Williams and Roggenbuck ended up compiling 27 items, 11 items of which were related to place dependence and 16 items which were related to place identity. 129 students from four different universities were asked to respond to 27 items designed to measure place attachment such as This place means a lot to me and I find that a lot of my life is organized around this place. Respondents were asked to indicate how they felt about various na tural settings (e.g., wilderness, backcountry, roadless or natural area) that they had most recently visited. Each item was rated on a 5-point Likert scale that was anchored from strongly agree to strongly disagree. Rotated principal factor analysis produced three factors. The fi rst factor with 6 items showed an internal

PAGE 69

69 consistency of .86 which represented place identit y. The second factor with 7 items showed an internal consistency of .82 which represented pl ace dependence. The third factor consisting of 9 items showed an internal consistency of .78 wh ich represented place indifference. The first two combined factors (place identity and dependence) with a total of 13 items showed an internal consistency of .85. Following the work of Williams and Rogge nbuck (1989), Williams et al. (1992) used 13 place attachment questions on a five-point Like rt scale to measure th e place attachment of visitors to four wilderness areas, each of which wa s located at different U.S. national forests. The extent of the relationship between place dependence and place identity was not reported in their study. However, responses to 13 items were combin ed into a place attachment scale. This scale was found to be internally consistent and high (Cronbachs alpha = .92 to .93) across the four wilderness areas. Following the initial scale developm ent (Williams & Roggenbuck, 1989), Moore and Graefe (1994) measured the place a ttachment of users at three recr eational trails (Heritage Trail near Dubuque, Iowa; St. Marks Trail near Tallah assee; Lafayette/Moraga Trail, east of San Francisco, CA) that were purposively selected. They were reported to be slightly different in terms of physical features, length and use levels. The author used 15 items which were slightly reworded to trails rather than wilderness areas. Principal factor analysis for each trail separately identified two dimensions of place attachment to recreational trail settings. The dimensions consisted of place dependence based on 6 items and place identity using 4 items while a third factor was not considered because it did not lo ad across all study areas. Standardized internal consistency for each dimension was high (Cronb achs alpha = .85). Correlation between two dimensions ( r = .623, p < .001) empirically showed some ove rlapping constructs between them.

PAGE 70

70 Vaske and Kobrin (2001) used 4 items for place dependence and 4 items for place identity drawn from a pool of items used by Williams and Roggenbuck (1989) to measure place attachment. The data were collected from student s of 14-17 years of age that participated in natural-resource-based work programs. Explorator y factor analysis produ ced the two dimensions of place attachment and high internal reliability coefficients for place dependence (Cronbachs alpha = .82) and place identity (Cronbachs alpha = .83). Using 12 items that have supported good inter-item reliability in other studies (Williams et al., 1995), Williams and Vaske (2003) tested whether two dimensions of place attachment were adequately valid. Confirmatory factor anal ysis demonstrated that a two-dimensional model was more predictive than a single dimension mode l and that the two dimensions were distinct. Additionally the convergent va lidity of place dependence an d place identity proved to be significant, with the frequency of visitation, fami liarity with a place, and awareness of a special place responses. In another study, Jorgensen and Stedman ( 2001) used 12 items borrowed from previous studies (Stedman, 1997; Williams & Roggenbuck, 1989) that were slightly different from items used by Williams and Vaske (2003) and asked lake property owners in Northern Wisconsin to rate the items. Jorgensen and Stedman observed three dimensions of place attachment (called sense of place in their study) : place dependence, place attachme nt, and place identity. Their further study demonstrated multidimensional constructs of sense of place as well as some common variation across those thr ee constructs. Sense of place c onstructs tend to be distinct, which is consistent with the findings (Williams & Roggenbuck, 1989). But there was some overlapping variation among the constructs.

PAGE 71

71 Using nine items borrowed from previous studies (Moore & Graefe, 1994; Williams, et al., 1992) to measure place attachment on a 7-po int Likert scale, Stedman (2002, 2003a) found a unidimensional and reliable measure of place at tachment (Cronbachs alpha = .937), which was consistent with a previous study (Jorgensen & Stedman, 2001) and used a summated place attachment scale in rela tion to other variables. More recently, Kyle, Mowen, and Tarrant (2004) used 11 items consisting of place dependence, place identity, and affective attach ment (place attachment) developed by Williams and Roggenbuck (1989), along with 3 items taken from Gruen, Sommers, and Scitos (2000) organizational commitment scale representing respondents social bond ing with a Cleveland Metroparks setting to measure pl ace attachment. Confirmatory f actor analysis in their study demonstrated that internally reliable consis tency for each of the dimensions in their study. In the current study, to measure place attachment of visitors to Ocala National Forest, the survey included place attachment items modified from previous research (Jorgensen & Stedman, 2001; Kyle, et al., 2004; Moore & Graefe, 1994; Williams et al., 1992; Williams & Roggenbuck, 1989) on a 5 point-Likert scale ranging from s trongly disagree to strongly agree. For instance, an original item, My lake property says very little about who I am (Jorgensen & Stedman, 2001) was reworded into This place says very little about who I am. This study asked visitors to rate items about the place they vis ited at the time they were interviewed in Ocala National Forest. The appendix provides a full list of items that were used in this study. Place Satisfaction Place satisfaction was initia lly developed and accessed by Stedman (2002) as an overall attitude measure consisting of three different me asures: satisfaction with place attributes, place quality, and perceived condition of ones place. Nine items of lake elements such as the number of users, the level of solitude /peacefulness, or quality of fish ing were rated by lake property

PAGE 72

72 owners in Northern Wisconsin to measure satisfaction with lake attributes. The items were anchored on a 4-point Likert scale from extremely dissatisfied to extremely satisfied. Factor analysis generated a single dimension. An intern al reliability analysis demonstrated a very reliable unidimensional place satisfaction (Cronb achs alpha = .846) factor. The second measure assessed lake quality using a single item (How would you rate your lake overall?) on a 6-point Likert scale anchored from poor to perfect. The third measure assessed perceived conditions of lake using a single item (How do you feel about the condition of your lake?) on a 4-point Likert scale anchored from str ongly dislike to str ongly like. Stedman (2002) standardized and summed those three separate measures into a single place satisfaction scale and observed very reliable internal consistency of the place satisfaction scale (C ronbachs alpha = 0.822). Stedman (2003a) used 8 items representing different elements of the respondents lake that were used by a previous study (Stedman, 2002), which asked res pondents to rate the items on a 5-point scale ranging from extremely dissatisfied to extre mely satisfied. This study found the internal reliability of a composite measure to be very reliable (Cronbachs alpha = .822). Similarly, Kyle, Graefe and Manning (2005) employed 5 items on a 5-point Likert scale ranging from strongly disagree to strongly agree to measure satisfaction of hikers at the Appalachian Trail. For instance, the items included I thoroughly enjoyed my trip on the Appalachian Trail, and my trip on the Appalachian Trail was well worth the money I spent on it. The items were summated into a single comp osite measure of satisfaction that was reported to be reliable. The Cronbachs alpha score was mo re than .60 across different types of hiking use at the Appalachian Trail. The single measure wa s predicted by two place attachment dimensions and three leisure involvement dimensions. However, such types of satisfaction items were treated in the context of a more leisureoriented concept of satisfaction.

PAGE 73

73 To measure visitors satisfaction with Ocal a National Forest, this study used items that have been recently developed and used (Stedm an, 2002; 2003a) but have not been employed in other studies. Additional items were added. Overa ll, this study included 10 items to measure satisfaction with place attributes on a 5-point Likert scale ranging from extremely dissatisfied to extremely satisfied. The items consisted of physical attributes (e.g., scenery of a recreation area, populations of wild life around it, its water quality, its level of development), social attributes (e.g., solitude and peacefulness around a re creation area, the number of people using it, other peoples recreation activities), and managerial attributes (e.g., user fee cost at a recreation area, its level of rules and regulations, its facili ties and services), which represent satisfaction with ROS setting indicators. The appendix provides a full list of items that were used in this study. Place-Related Behavior and Acceptabili ty of Potentia l Management Actions Warzecha and Lime (2001) asked visitors to rivers in Canyonlands National Park to rate items indicating levels of support for potential management actions such as prohibition of motorized rafts from the river, and campsite rese rvations on a 4-point Like rt scale ranging from strongly oppose to strongly support. Following the work of Heberlein and Vaske (1977), Kyle, et al. (2004a) used three items meas uring respondents sett ing density along the Appalachian Trail. The items include items measuring crowding on a 9-point scale, the acceptability of the number of people on a 9-point scale, and an appropriate number of people on the trail on a 5-point scale ranging from strongly di sagree to strongly agree. Noting that the Cronbachs alpha coefficient is .60 or above with the six or less items, Kyle, et al. (2004a) demonstrated acceptable internal re liability (Cronbachs alpha = .81). Examining the relationship between place atta chment and perceptions of environmental conditions along the Appalachian Trail, Kyle, et al. (2004b) employed multiple items that were

PAGE 74

74 grouped into six dimensions such as trail de velopment, use impact, depreciative behavior, perceived crowding, use conflict, and development encroachment after they discussed with trail managers and reviewed past research on respondents preferred environmental conditions. The items were rated on a 3-point Like rt scale ranging from not a problem, to a big problem. Confirmatory factor analysis demonstrated acc eptable construct reliabilities. However, their study did not measure explicit accep tability of environmental condi tions that is commonly used in social judgment theory. A list of multiple items measuring acceptabil ity of environmental impacts was developed by Floyd, Jang, and Noe (1997) after discussion with National Park Service managers. The items were measured by location and within situational contexts such as litter in picnic areas contrasted with litter on a hiking tr ail, instead of specific levels of im pacts such as an exact amount of litter on a trail because of the recall problem. Th e multiple items were rated on a 7-point Likert scale ranging from highly unacceptable to h ighly acceptable in two national parks. Seven dimensional factors held acceptable reliabi lity coefficients (Floyd, et al., 1997). Basically, recognizing that there is no sp ecific guideline on meas uring and including items representing potential mana gement actions, and that incl usion of items depends on an agencys needs and recreation rese arch context, the survey in this study used 18 items to measure the acceptability of potential management actions that are general and currently effective in the study area on a 7-point Li kert scale anchored from extremely unlikely to extremely likely. The items included three recreation setting attrib utes representing managerial (2 items on rules on recreation), environmental (2 items on envi ronmental quality) and social (2 items on increasing number of OHV riders; 2 items on increas ing number of general recreators) aspects of

PAGE 75

75 recreation planning. This study used a wider rang e of measurement scales to generate more variability in measures. The appendix provides a fu ll list of items that were used in this study. Pilot Study The primary purpose of a pilot test was to identify wording, interpretation and implementation problems before the actual survey was conducted. After the preliminary survey was developed, based on the lite rature review, the investigator distributed it to 15 graduate students and professors from the Department of Tourism, Recreation and Sport Management and 15 graduate students and professors from the School of Natural Resources at the University of Florida. They were asked to identify any potential problems in the instrument such as with the instructions, survey design, wording, definitio ns, response options or other difficulties. Additionally, the investigator distributed th e survey instrument to 150 students who took parks and recreation-related cl asses in the Department of Tourism, Recreation, and Sport Management at the University of Florida. The survey instrument wa s also distributed to users at Lake Wauburg, Gainesville, Florida. The reason for the selection of that site was that since they were at a water-based recreation area, they were a good source for checking whether the survey instrument items were comprehensible to them. In order to further increase the reliability of the instrument, the investigator followed the same procedures for a pilot test at the actual surv ey site. Fifty visitors at Ocala National Forest, Florida, were approached and were asked to an swer the preliminary onsite survey instrument. After their completion of the survey, they were asked a series of que stions to identify any problems in the instrument. Then the participants were asked to participate in and to return a follow-up mail survey after their trip back to th eir home residence. A new series of questions were asked to identify any problems in the mailback instrument. After the review of comments

PAGE 76

76 on the instrument from participan ts in the pilot study, changes were made to optimize the final version of the surveys. Data Analysis On-site and mail-back survey data for this study were coded and st ored in SPSS (version 15.). Data analysis including response and non-response bias checks and descriptive statistics for demographic variables was performed using SPSS (version 15). Other st atistical analysis including confirmatory factor an alysis (CFA), scale reliability and structural equation modeling (SEM) using Amos 7.0 were performed to examine the hypotheses. Structural equation modeling was implemented due to having a dvantages over separate regressi on analyses (Lavarie & Arnett, 2000). Spatial analysis of place-protective behavi or was performed using ArcGIS (version 9.2) In more detail, non-response bias checks were made, comparing onsite and mail-back surveys of visitors at Ocala NF to examine th e similarities and differences between respondents and non-respondents in onsite visitors demographi cs and recreation char acteristics. Respondents were those who filled out an onsite survey, prov ided their mailing information for a follow-up mail survey, and returned it to the researcher while non-respo ndents were those who did not reply to the follow-up mail-back surveys. Inde pendent samples t-tests were implemented to identify the differences in visitor group size, th e number of days spent at Ocala NF, and age. Chi-square tests were performed to examine the differences in categorical variables such as gender, income levels, and education levels. St atistical significance wa s accepted at the alpha level of .05. Procedures of Confirmatory Factor Anal ysis and Structural Equation Modeling Before testing hypothesized m odels, the following analytical processes were conducted: 1) confirmatory factor analyses were perfor med to examine the reliability and validity of recreation experience preferences, place meani ngs, place satisfaction, place attachment, and

PAGE 77

77 place protective behavior latent variables, 2) m ean scores of subdomains for each factor were achieved to create observed items for those latent variab les, 3) structural equation modeling were performed to test this hypothesi s. Criteria used for measurement and structural models follows. Internal consistency of factors has been commonly assessed using an acceptable value of Cronbachs alpha coefficients e qual to or larger than .70 (N unally & Bernstein, 1994; Nunally, 1978). However, if 6 items or less were used to measure internal consiste ncy, values equal to .60 and larger may be considered to be accepta ble (Cortina, 1983). Average variance extracted (AVE) for each latent construct was estimated to assess the amount of variance in the indicators explained by the latent construct. According to Hair et al. (1998), the value of AVE should be larger than .50. In addition, construct reliab ility (CR) was assessed using a recommended standardized loading of .70 (Nunally & Bernstein, 1994; Hair et al., 1998). The factor loadings were examined for converg ent validity of scales. Hair et al. (1998) indicated that while factor loadings equal to or higher than .707 i ndicate that the items reflect the construct well, indicator loadings equal to or larger than .50 al so are acceptable. Discriminant validity was assessed to be good if the correlations between the factors were less than .85 (Kline, 2005). However, if two or three factors were not equal to or hi gher than .90, the discriminant validity was also accepta ble (Hair et al., 1998), in dicating that the fact ors were distinctly measured from one another. The overall model fit of the measurement model referred to as CFA was examined by assessing the root-mean-square of approximation (RMSEA) (Steiger & Lind, 1980; Steiger, 1989). The RMSEA was the first index to determin e whether the measurem ent model was fitted to the observed data. According to Browne and Cudeck (1993), values of RMSEA less than .05 indicate a good fitting model and those as high as .08 indicate reasonable fit. According to

PAGE 78

78 MacCallum et al. (1996), values between .08 and .10 indicate a mediocre fitting model and values higher than .10 indicate un acceptable or poor f itting models. Similarly, Hu and Bentler (1999) accepted RMSEA values less than .06 for a close fitting model. Values ranging from .06 to .08 reflect reasonable fit, values from .081 to .100 are indicative of mediocre fit, and values larger than .100 indicate unacceptable fit. In addition, chi-square affected by a large sample size tends to reje ct the model despite minor differences between observed and estimate d covariances (Kline, 2005). Ratio of the chisquare to its degrees of freedom was examined due to the sensitivity of chi-square to a large sample size in this study. Values ranging from 2. 0 to 3.0 are generally acceptable. But there is no consensus on the values. Bollen (1989) accepted th e values as high as 5.0 for a fitting model. While values of the comparative fit index (CFI) close to .95 refl ect a well-fitting model to the data (Hu & Bentler, 1999), values higher than .9 0 were desirable for an acceptable fitting model (Bentler, 1992). After the measurement modeling was impl emented, the structural modeling was performed using Amos to determine the fit of the hypothesized models. The same goodness-offit indices used for the fit of the measurement mo del were also employed to assess the fit of the structural model. Standardized pa th coefficients were used to examine the direct and indirect effects of exogenous variables on endogenous variables in the hypothesized models. Procedures of Spatial Diagnostic Anal ysis of P lace-Protective Behavior This spatial analysis employed combined onsit e and mail-back survey data collected from visitors at Ocala NF to generate three separate sets of regressi on models. These models were to test the effects of major inde pendent variables (i.e., recreatio n experiences, place meanings, place satisfaction, place attachment) on place protect ive behavior by selecting a model-specific set of variables, to examine spatial patterns in residuals for each model, to determine which

PAGE 79

79 models better explain place-prot ective behavior of recreationists, and to determine any spatial pattern of place-protective behavior. As Schroeder (1983) mentioned, selection of variables used in this study was based on pr evious research findings and ove rall fit for regression models. The combined data include the following major information about visitors: trip characteristics (i.e., the number of people in group, length of stay), point of origin (five digit zip code, county, state) of participants, importance of recreation experience pr eferences, attainment of recreation experience prefer ences, place meanings, place satis faction, place attachment, and place-protective behavior. Of a ll returned 934 mail-b ack surveys, 87% (n=811) were instate visitors, 12% (n=115) were out-of-state visitors and 10% (n=8) were international visitors. A majority of respondents were from Florida and only three of six visitor groups were from two Georgia counties adjacent to two counties of the No rthern part of Florida. In addition, the county level boundary within Florida was mo re appropriate than the 3-dig it zip code for this study since in some cases two counties or more appear in one 3-digit zip code boundar y. In other words, the county boundary provides geographically smaller s cale than the zip code boundary, which means spatial analysis can be made more specifi c using the former boundary. Additionally, the reasonable number of boundaries for spatial analysis ranges from 30 to 50. Of the 67 Florida counties, mail-back respondents li ved in one of 43 counties. Thos e respondents lived in one of about 31 three-digit zip code bounda ries of Florida. After taking these into consideration, only Florida recreationists at Ocala NF were included for the spa tial analysis in this study. Composite place-protective behavi or variable was treated as a dependent variable across three models. Model 1 used the following indepe ndent variables: fact ors of importance of recreation experience preferences, place meani ngs, place satisfaction and place attachment. Model 2 replaced the variable importance of recr eation experience preferences in Model 1 with

PAGE 80

80 attainment of recreation experien ces and kept the remaining indepe ndent variables. Model 3 used a gap score variable (subtrac ting importance of recreation ex perience preferences from attainment of recreation experien ces), instead of importance of recreation experience preferences and kept the remaining variables used for Model 1. Factors of each of the variables were used as independent variables for three m odels, based on measurement model. Finally, this study ran stepwise regression procedures for th ree models to determine the significant variables where little is known about the significance of variables under consideration and to find accurate interpretati on of variables and appropriate models (Schroeder, 1983). This study also ran Morans I test fo r spatial autocorrelation to id entify spatial pa tterns of placeprotective behavior using ArcGIS (version 9.2). Autocorrelation sta tistics were used to determine whether each model was complete by identifying di fferences in regression residuals. Small errors and independent errors are two general statistical models. But only the latter is applied to spatial diagnostic tests in this study. That is, the errors should not be correlated over space. Otherwise, variables in the models are not accounted for and there are more variables to add in the models (Odland, 1988). Values of Morans I closer to zero indicate less clustering, while values closer to one indicate more clustering (Moran, 1950). After testing spatial diagnostic analysis of each model, spatial analysis of place-protective behavior was performed to determine its pattern. Statistical significance was accepted at the alpha level of .05. Over all, this spatial diagnostic analysis would help generate more appropriate model specification and lead to more accurate interpretation of regression coefficients. Autocorrelation Statistics Spatial autocorrelation statistics measure whet her the observed value of a variable at one location is independent of values of the variable at adjacent lo calities, assist in examining hypothesis about variables, and f acilitate developing statistical models that explain spatial

PAGE 81

81 patterns (Odland, 1988). The most co mmonly used global statistics for spatial autoco rrelation are Morans I and Gearys C (Cliff & Ord, 1973; Odland, 1988). Morans I is produced by standardizing the spatial autocovariance and measures the connectivity for the set of locations. Alternatively, Gearys C uses the sum of squared differences between pairs of data values as a measure of covariation. Morans I measures the correlation among adj acent values in a pattern. It is a productmoment correlation, similar to th e traditional correlation coeffi cient in the sense that the numerator is a cross-products term, while the denominator is a variance term (Odland, 1988; Sokal & Oden, 1978). The value of Morans I range s from -1.0 indicating perfect negative spatial autocorrelation, to 0 indicating a random patte rn to +1.0 indicating perfect positive spatial autocorrelation. Again, positive autocorrelation m eans similar (higher and higher) values are clustered together, while negative autocorrelatio n means dissimilar (higher and lower) values are nearby (Moran, 1950). Gearys C is similar to Mora ns I. The value of Gearys C ranges from 0 indicating positive spatial correlation, +1 repr esenting no correlation, to 2 indicating negative spatial autocorrelation. These statistics meas ure spatial autocorrel ation among continuous numerical variables such as ordinal, interval or ra tio data (Goodchild, 1987). Autocorrelation and Regression Models There are various ways to use spatial auto correlation as statistics. First, spatial autocorrelation can be used independently to exam ine the interdependence of values of a variable in a spatial pattern. Additionally it can be combined with other sta tistical models if they have a spatial data (Odland, 1988). The difference between the observed value and th e predicted value in statistical models is called an error or residual. The two common criteri a used for evaluation of statistical models are that the errors should be small and independent. Spatial autocorrelation is not related to whether

PAGE 82

82 the residuals are small but is applied to whether the residuals are independent. Spatial models or spatial diagnostic tests require th at the residuals should not be interdependent over space. Spatial interdependence between errors in a statistical model means th at the observed values are somewhat spatially arranged, which is not explained by the model (Odland, 1988). Regression models have been used in geography for years. From a geographical perspective, it is very important to implement spatial autocorrelation statistics in fitting a regression model to spatia l data. It is a condition that a re gression model must hold independence of the errors for reliable hypot hesis testing and interpretation of the model. If the spatial autocorrelation in the residuals is present in testing a regression model, then this means that the model does not provide a systematic organization in the data, the model is not correct and complete, and some source of variation from the model has been left out (Miron, 1984; Odland, 1988). Autocorrelation in the residuals is primaril y due to an unsuitable functional form of a model, rather than unsuitable data used in the model. Additionally, it may be due to left-out variable(s) that may interact with locations. The regression model with spatial autocorrelation in the residuals needs to be fixed before any statis tical inference is made. It may be sufficient to modify an incorrect functional form or add missi ng variables to remove autocorrelation in the residuals from regression models (Odland, 1988).

PAGE 83

83 Figure 3-1. Ocala National Forest, FL

PAGE 84

84 CHAPTER 4 RESULTS Response Rate A total of 2,144 Ocala NF vis itors agreed to participate in the follow-up surveys for this study. About 8 percent of the follow-up questionnair es (n=171) were return ed as non-deliverable envelopes due to invalid addresses. That is a total of 1,973 follow-up questionnaires reached recent visitors. 934 questionnaires were completed and returned, which yielded a response rate of 47%. Demographics of Respondents Table 4-1 shows the demographics of the res pondents. Half of the respondents were male (51%) and the rest were female (49%). The m ean age of the respondents was about 45 years and about half were between 31 and 50 years of age. More than half of the respondents (65%) had some college or 2 year degree or higher while the rest (35%) attended business or technical school or had a lower level of education. Appr oximately one third (33.6%) had their household incomes between $15,000 and 34,999. About two thirds of them (61.5%) had their household incomes of $65,000 and higher. The majority of the respondents (95.7%) were Caucasians. Non-Response Bias Test A non-response bias check was made to iden tify whether there were differences between respondents and non-respondents to mail-back ques tionnaires on their trip characteristics, demographics, group size, the number of days sp ent at Ocala NF, gender, age, income and education levels. Independent sample t-tests we re employed to identify the differences in group size, the number of days spent at Ocala NF at the time they were interviewed, and age. Chisquare tests were implemented to examine the differences in gender, income and education levels.

PAGE 85

85 As shown in Table 4-2, significant differences were found between the two groups in the number of people in their group [ t (1937) = -2.34 p = .019], the number of days spent at Ocala NF during the time they were interviewed [ t (1621.92) = 2.57 p = .010] and age [ t (1813.59) = 12.49, p < .001]. Respondents ( M = 4.29, SD =17.83) spent more days at Ocala NF than nonrespondents ( M = 2.40, SD =12.29) during the trip they were interviewed and respondents ( M = 45.12, SD =14.45) were more likely to be older than non-respondents ( M = 37.53, SD = 12.17). On the other hand, non-respondents ( M = 6.92, SD =7.28) on average visited with larger groups than respondents ( M = 6.17, SD = 12.29). No significant difference was found in gender [ 2 = .260 (1, N=1973), p = .610]. Significant differences were found in education levels (2 = 55.643 (6, N=1959), p < .001) and income levels. [ 2 = 17.040 (7, N=1845), p = .017]. Respondents were more likely to be more educated and to have higher income levels th an non-respondents. But the difference in income levels was minor. In sum, there were differe nces between respondents and non-respondents in trip characteristics and demographi cs of visitors at Ocala NF. This study considered the results to be more representative of those who spent more days at Ocala NF during a trip, had smaller group sizes, were more educate d, and had higher income levels. Differences in Behavioral Cons tructs among Different Users Item s used for each variable shown in Table 4-3 were the same items used for final model in Table 4-6 in order to sustain a more valid and reliable analysis in identifying differences in behavioral constructs among different recreati onists, OHV riders, swimmers, and non-motorized boaters (canoers and kayakers). As shown in Table 4-3, enjoyment of nature and escaping physical pressure were the variables that showed the greatest differen ces among different types of users in gap scores of recreation experiences. Visitors who had kayaking or canoeing

PAGE 86

86 experiences through Juniper wilderness canoe run in Ocala NF were more likely to report that they had higher enjoyment of nature (e.g., view the scenery, enjoy the smells and sounds of nature) than visitors who had swimming experi ences in major spring sites in Ocala NF ( F = 3.883, p <. 05). OHV riders were more likely to re port a higher gap score for escaping physical pressure (e.g., experience tranquility, experience solitude) than visitors at major swimming areas ( F = 7.599, p < .01). Juniper wilderness visitors and ma jor swimming areas visitors were not as likely to report escaping physical pressure experiences. Three significant differences were noted across the sense of place constructs. Firstly, canoers and kayakers were more likely to percei ve recreation settings to be pristine than OHV riders and swimmers ( F = 8.640, p < .001). Conversely, the latter users tended to consider recreation areas to be in less pristine conditi on. No significant difference was found across place attachment constructs (e.g., place attachment place dependence, place identity) among different users. However, place dependence scores appeared to be higher than scores of place attachment and place identity across different user types. That is, most users felt dependent upon specific recreation areas for their recreat ion activities. Secondly, OHV riders were not likely to be more satisfied with physical settings (e.g., scenery, level of land devel opment) than swimmers at major swimming areas and users at Junipe r wilderness canoe trail run ( F = 57.123, p < .001). Thirdly, OHV riders appeared to be less sa tisfied with managerial aspects (e.g., recreation fees cost, level of rules and regulations) than the other users (F = 3.073, p < .05). Overall, respondents did not show high levels of satisfaction with recreation settings, esp ecially social and managerial aspects. Additionally, OHV riders were less likely to re sist potential changes to recreation settings related to environmental, social actions except fo r managerial actions, compared to other users.

PAGE 87

87 OHV riders appeared to be less sensitive to potentially increasing numbers of OHV riders ( F = 38.599, p < .001) and general recreators ( F = 16.408, p < .001), and any potential deteriorating environments in recreation areas ( F = 8.425, p < .001), while they were more likely to show their intention to resist any potential rules and re gulations that would be more restrictive ( F = 32.764, p < .001). Overall, OHV riders were less likely to resist any increase in developed recreation settings except for rules and regulations, compared to other users. Other users tended to prefer environmentally protected, less crowded and less restrictive recreation se ttings. In particular, other users showed active inten tions to keep recreation areas as environmentally protected settings. Respondents Overall Rating of Each Construct As shown in Table 4-4, respondents at Ocala NF were m ore likely to seek enjoyment of nature (e.g., view scenery), escap ing social pressures (e.g., get aw ay from the usual demands of life) and escaping physical pressures (e.g., experien ce tranquility), while they were less likely to consider use of recreation equipment, achievement or taking risks to be important reasons for visiting major recreation areas in Ocala NF. Res pondents tended to attain recreation experiences such as enjoyment of nature, escaping social pr essure, and nostalgia (e.g., bringing back pleasant memories). However, they were less likely to report attaining recreati on experiences such as using recreation equipment, skill achievement, and risk taking. Respondents reported that they attained sligh tly higher levels of recreation experiences such as achievement, nostalgia, escaping social pressures, use of recreational equipment, and enjoyment of nature than they expected. But th ey were less likely to report attaining recreation experiences such as taking risks and escaping physical pressures (e.g., experience tranquility, solitude) higher than they expected.

PAGE 88

88 One of the three sense-of-place constructs is abstract or specific place meanings measured as perceived belief. Specific place meani ngs segregate into two factors, pristine and impacted, borrowed from Stedman (2002). Visito rs at Ocala NF were more often neutral regarding whether or not recreation areas we re impacted (e.g., extremely crowded, many recreationists, many developed f acilities). And, visitors were likely to consider recreation settings to be in pristine conditions (e .g., extremely scenic, forested, peaceful). Place attachment was employed as another aspe ct of the sense of place construct. The level of dependence (e.g., good place for recreation) that visitors at Oc ala NF ascribed to recreation areas was higher than level of attachment (e.g., relaxi ng) and level of identity (e.g., being oneself). That is, visitors were more likel y to be more dependent on recreation areas than they were attached to recreation areas and than they identified themselves with. However, those three domains of place attachment tended to be in slightly higher than neutral for their recreation experience in Ocala NF. Thus, this does not indi cate a high level of overall place attachment. Place satisfaction was regarded as a third sens e of place construct. Visitors satisfaction with physical settings was reported to be higher than neutral, while their satisfaction with both managerial and social settings tende d to be at a neutral level, on average. That is, respondents at Ocala NF were more likely to be satisfied with physical settings (e.g., scenery, populations of wildlife) than managerial aspect s (e.g., recreation fees cost, rules and regulations) and social settings (e.g., solitude, the number of people using recreation areas). Place-protective behavior was measured as an indication of behavioral intentions towards resisting potential future deve lopment of recreation settings consisting of four domains. Respondents at Ocala NF showed more active in tentions to resist future environmental deterioration of recreation setti ngs than resisting other changes in recreation settings (e.g., more

PAGE 89

89 restrictive rules and regulations and increasing number of recr eationists including OHV riders). In particular, visitors tended to be neutral towa rds resisting more restrict ive rules on the kinds of recreation allowed and increas ing the number of users. Measurement Model Analysis As a part of initial processes of checking da ta prior to data analysis, this study examined the values of skewness and kurtosis for each of the items used in the measurement model to determine whether each item has an absolute value of less than 2.0, m eeting the criteria of normality in the data. As shown in Table 4-4, all observed items met the criterion value of kurtosis coefficients at greater than 2.0. Overall, the items used in this study met the conditions of normality. Before performing the structural model an alysis, the study implemented confirmatory factor analyses for all scale va riables. Assessing the fit indices (e.g., RMSEA, CFI) for all latent constructs, this study found all hypo thesized models were acceptabl e for the latent variables, importance of recreation experien ce preference items, attainment of recreation experiences, gap scores of recreation experiences. The fit indices can be found in Table 4-5. The final measurement models examined data fit for specific place meaning, place attachment, and place satisfaction variables adopted from the items used by Stedman (2002) and were found to be different from latent factors extracted from exploratory factor analysis (EFA) by Stedman (2002). This study dropped items with poor factor loadings to adjust fit indices. For instance, specific place meanings ended up with one factor structure labeled Pristine with 4 items and removed another factor structure labeled Impacted with its poor factor loadings, poor reliability and weak fit indices. In addition, unlike the original number of items used for measuring the place attachment (sense of place) construct by Jorgensen a nd Stedman (2001), this study found that a place

PAGE 90

90 attachment domain with 3 items, place dependen ce domain with 2 items, and place identity domain with 2 items were more likely to be the most appropriate factors representing the place attachment construct. Lastl y, unlike the previous study c onducted by Stedman (2002) who converted place satisfacti on items into one composite variable, this study divided those adopted items and added one item (i.e., level of rules an d regulations) to form three latent factors, physical setting (3 items), social setting (2 items) and managerial setting (2 items). The analysis concluded that those three factors were more theoretically appropriate for ROS planning and found that they fit the data well. Overall, fit indices for all the scale variables revealed acceptable fitting models. Results from the final analysis (Table 4-6) showed that most scale variables held an appropriate factor structure, meeting reliability and validity criteria. For instance, Cronbachs alpha scores were all acceptable for each latent construct, exce pt for enjoyment of nature ( = .55) factor among the latent factors measuring ga p scores of recreation experience preferences. Internal reliability coefficients tended to be lower than values accepted by Cortina (1983) who stated that .60 or more is acceptable with six it ems or less for one factor structure. Construct reliability scores for most latent factors te nded to show the acceptable value of construct reliability (>.70) recommended by Nunally and Bernst erin (1994) except for factors for gap score of recreation experience preferen ce such as achievement (.59), enjoyment of nature (.55), and nostalgia (.62). AVE coefficients for all the s cale factors tended to s how an acceptable value of AVE (>.5) except for factors for gap score of recreation experience preference such as achievement (.42), enjoyment of nature (.38), and nostalgia (.45), and physical settings under place satisfaction (.39).

PAGE 91

91 Overall, most construct reliability and AVE coefficients for gap score of recreation experience preferences were likely to show marginally acceptable values. However, coefficients of internal reliability for all latent constructs were likely to show acceptable levels. Additionally, the same latent construct measured in a differe nt way for model 2 and model 3 were used for comparing the difference between the two models in structural fit. Taking these into consideration, therefore, this st udy included those latent factors in the structural model. Final items and latent factors were used to obtai n aggregate means of each latent factor for hypothesized measurement models. Results of Hypothesis Testing Structural Analys is of Model 1 After achieving a valid measurement model, this study examined structural model 1. After eliminating significant parameters, the final model in Figure 4-1 dem onstrated a reasonable model fit ( 2 = 442.006, df. = 95, p < .001, CMIN/df = 4.653, RMSEA = .063, CFI = .93, IFI = .93). H1: There is a positive association betw een the importance of REP and PM. As shown in Table 4-7, results identified that there was no significant relationship between level of importance of recreation experi ence preferences and perception of settings in the forest ( = -.04, t-value = -0.90). Th at is, visitors who had higher recreation experience preferences did not significantly consider forest recreation settings to be natural, although the effect is positive. H2: There is a positive association betw een the importance of REP and PS. Results identified that place satisfaction was significantly and negatively influenced by importance of recreation experience preferences ( = -.09, t-value = -2.26). That is, visitors that

PAGE 92

92 sought their recreation experiences in the forest were less likely to be satisfied with recreation settings in the forest. H3: There is a positive association betw een the importance of REP and PA. Levels of importance of various recreation experien ce preferences was strongly predictive of higher place attachment ( = .31, t-value = 8.73). The path coefficient indicates that individuals with higher recreat ion experience preferences were likely to have strong place attachment, explaining 41% of the variance in place attachment. H4: There is a positive association between PM and PS. Place satisfaction was also negatively affected by perceptions of pristine conditions in the forest ( = -.39, t-value = -8.29). That is, visitors who did not percei ve the forest to be less pristine were less likely to be satisfied with settings in the forest. Conversely, as visitors perceptions of recreation settings were more positive, their satisfaction with recreation settings tended to increase. H5: There is a positive association between PM and PA. Results revealed that percepti on of pristine conditions (setting s) in the forest also had a significant and negative influence on place attachment to the forest ( = -.14, t-value = -3.53). Visitors who perceived the recrea tion settings to be less pristine were less likely to show attachment to recreation areas. Inversely, vis itors who had positive perception of recreation settings tended to de velop attachment to t hose recreation areas. H6: There is a positive association between PS and PA. Results showed a significant and positive re lationship between place satisfaction and place attachment ( = .51, t-value = 10.19). As overall place sa tisfaction increase d, overall place

PAGE 93

93 attachment increased strongly. Th is indicated that visitors who were more satisfied with recreation settings were likely to devel op their attachment to recreation areas. H7: There is a positive association betw een the importance of REP and PB. Levels of importance of recr eation experience preferences di d not significantly predict behavioral intentions related to futu re changes in recreation settings ( = -.05, t-value = -1.21). Visitors who had higher motivation did not show their significant willingness to keep recreation settings more natural. H8: There is a positive association between PM and PB. Perception of pristine conditions ( = -.06, t-value = -1.38) did not have a significant direct impact on future behavioral intentions against potential future changes in recreation settings, respectively and their influences on th eir behavior were negative. This means that although visitors perceived recreation settings to be less pristine, they did not show their willingness to resist more developed recreation settings. H9: There is a positive association between PS and PB. Place satisfaction ( = -.04, t-value = -0.66) did not have a significant direct impact on future behavioral intentions ag ainst potential future changes in recreation settings, respectively and their influences on their behavior were ne gative. Rather, the rela tionship between the two tended to be negative although the re lationship was not significant. H10: There is a positive association between PA and PB. Results indicated that overal l place attachment was predic tive of greater intentions toward exhibiting place-protective behavior for the forest ( = .23, t-value = 4.58). In other words, visitors with strong place attachment were more inclined to partic ipate in place-protective behavior in the future, being sensitive to future changes in recreation settings.

PAGE 94

94 Additional Findings: Indirect Effects As shown in the results in Table 4-7, the relationship between im portance of recreation experience and place satisfaction was mediated by perception of settings in the forest ( = .01). Perception of settings, place satisfaction, and perception of settings and place satisfaction mediated the relationship between importance of recreation experience and place attachment to the forest ( = -.04). Although the mediating effect is ve ry modest, this finding indicates that visitors with higher motivation tended to have lower place attachment. This may suggest the importance of the mediating variables on place attachment in improving place attachment. In addition, the relationship between perception of settings and place attachment was mediated by place satisfaction ( = -.20). Regarding the mediating effects on place-pro tective behavior, the relationship between importance of recreation experiences and place-protective behavior was positively mediated by multiple mediating variables (e.g., perception of settings, place satisfaction, place attachment, perception of setting and place satisfaction, perc eption of settings and place attachment) ( = .07). The mediating variables, place satisfac tion, place attachment, a nd place satisfaction and place attachment, negatively mediated the relationship between perception of settings and placeprotective behavior ( = -.06). Place satisfaction had an i ndirect effect on place-protective behavior via place attachment ( = .12). The positive mediating e ffect indicates that visitors satisfied with settings in the forest were more likely to have intentions toward place-protective behavior. Structural Analys is of Model 2 After achieving another valid measurement mode l, this study examined structural model 2. After eliminating significant parameters, the final model in Figure 4-2 demonstrated a

PAGE 95

95 reasonable model fit ( 2 = 384.470, df. = 95, p < .001, CMIN/df = 4.047, RMSEA = .057, CFI = .939, IFI = .94). H11: There is a positive association betw een PM and the attainment of REP. As shown in Table 4-8, as perception of settings in the forest grew positive, attainment of recreation experience prefer ences tended to increase ( = -.34, t-value = -7.08). Visitors who perceived the forest recreation se ttings to be more natural were more likely to attain the recreation experiences they sought in the forest. These results indicate th e importance of natural settings in recreation areas th at could foster attainment of visitors preferred recreation experiences. H12: There is a positive association betw een the attainment of REP and PS. Place satisfaction was strongly influenced by attainment of recreation experience preferences ( = .38, t-value = 6.80): Visitors who ach ieved their pref erred recreation experiences were more likely to be satisfied with recreation settings in the forest. H13: There is a positive association betw een the attainment of REP and PA. Attainment of recreation expe rience preferences was strongly predictive of higher place attachment ( = .38, t-value = 7.53). Visitors who attained their preferred recr eation experiences were more likely to express attachment to those recreation areas. H14: There is a positive association betw een the attainment of REP and PB. Attainment of recreation expe rience preferences did not signi ficantly predict behavioral intentions toward future changes in recreati on settings although it s howed a weak positive influence on place-protective behavior against po tential future changes in recreation settings ( = .08, t-value = 1.69). These results in dicate that visitors who attain ed their preferred recreation settings did not tend to be signifi cantly sensitive to potential fu ture changes in those recreation

PAGE 96

96 settings. Maybe they might have thought that it would not be necessary to have more natural settings because they obtained re creation experiences they desired. Findings of Direct Effects Same as Revealed in Model 1 As shown in Table 4-8, place satisfaction was affected by negative perception of pristine conditions in the forest ( = -.26, t-value = -5.59), which means th at visitors who did perceive the forest to be less pristine were less likely to be sa tisfied with the settings in the forest. Perception of pristine conditions (settings) in the forest also significantly influenced place attachment to the forest ( = -.10, t-value = -2.52). In addition, as overal l place satisfaction increased, overall place attachment increased strongly ( = .33, t-value = 6.82). Perception of pristine conditions ( = -.05, t-value = -1.15) and place satisfaction ( = .04, t-value = -0.84) did not have si gnificant direct impact on future behavioral intentions toward potential future changes in recreation settings, respectively and their influences on the behavior were negative. Overall place attachment predicte d greater intentions toward place-protective behavior for the forest ( = .18, t-value = 3.55): visitors with strong place attachment were more likely to be sensitive to potential fu ture changes in recreation settings. Additional Findings: Indirect Effects As shown in the results in Table 4-8, the re lationship between per ception of settings in the forest and place satisfaction was mediated by attainment of recreation experience ( = -.13). This negative effect indicates that visitors perceiving the recreation setting to be less primitive tended to be less satisfied with overall recreation settings in the forest. Attainment of recreation experience, place satisfaction, a nd attainment of recreation experience and place satisfaction mediated the re lationship between perception of settings and place attachment to recreation areas ( = -.25). Although the mediati ng effect is modest, this finding indicates that visitors with negative perc eption of settings tended to have lower place

PAGE 97

97 attachment. This may suggest the importance of the mediating vari ables on place attachment in improving place attachment. In addition, the relationship between attainment of recreation experiences and place attachment wa s mediated by place satisfaction ( = .12). Regarding the mediating effects on place-pro tective behavior, the relationship between perception of settings and place-protective behavior was negatively mediated by multiple mediating variables (e.g., attainment of r ecreation experience, place satisfaction, place attachment, attainment of recreation experience and pl ace satisfaction, attainment of recreation experience and place attachment) ( = -.07). The mediating variab les, place satisfaction, place attachment, and place satisfacti on and place attachment, positively mediated the relationship between attainment of recreation experi ences and place-protective behavior ( = .07). Place satisfaction had an indirect effect on place -protective behavior via place attachment ( = .06). The positive mediating effect indicates that visitors satisfied with settings in the forest were more likely to have intentions toward place-protectiv e behavior, although the effect appeared to be very small. Structural Analys is of Model 3 After achieving another valid measurement mode l, this study examined structural model 3. After eliminating significant parameters, the final model in Figure 4-3 demonstrated a reasonable model fit ( 2 = 332.525, df. = 95, p < .001, CMIN/df = 3.500, RMSEA = .052, CFI = .944, IFI = .95). H15: There is a positive association between PM and GAP scores. As shown in Table 4-9, as perception of recr eation settings in the forest grew positive, gap scores of recreation experience pr eferences tended to increase positively ( = -.18, t-value = -4.17). These results indicate that the more nature-based settings in recreation areas were likely

PAGE 98

98 to have a positive influence on outcomes of recreation experiences. As mentioned earlier, natural settings tended to contribute to posi tive gap scores of recreation experiences. H16: There is a positive associatio n between GAP scores and PS. Place satisfaction was strongly influenced by recreation experience preferences gap scores ( = .24, t-value = 4.88): Visitors who achieved recreation experiences higher than their expectations were more likely to be satisfied with recreation settings in the forest. These results indicate the important role re source managers or the setting ca n play in managing visitors recreation experiences and satisfaction. H17: There is a positive associatio n between GAP scores and PA. Recreation experience preference s gap scores were not significantly predictive of place attachment ( = -.01, t-value = -0.17). This non-significan t relationship indicat es higher than expected outcome scores did not lead to an increa se in attachment to r ecreation areas. As shown previously in Table 4-9, however, attainment of preferred recreation experiences was positively associated with attachment to recreation areas. H18: There is a positive associatio n between GAP scores and PB. Recreation experience preference gap scores did not significantly predict behavioral intentions toward potential future changes in recreation settings although it showed a negatively weak influence on place-protecti ve behavior against future ch anges in recreation settings ( = .06, t-value = -1.58). This finding is similar to the finding that no-significant relationship between attainment of recreati on experiences and place-protective behavior was found, as shown previously in Table 4-8. Findings of Direct Effects Same as Revealed in Model 1 As shown in Table 4-9, place satisfaction was also affected by perception of pristine conditions in the forest ( = -.34, t-value = -7.35), which means that visitors who did not

PAGE 99

99 perceive the forest to be less pris tine were less likely to be satisfied with the settings in the forest. Perception of pristine conditions (settings) in the forest also significantly influenced place attachment to the forest ( = -.17, t-value = -4.09). Additionall y, as overall place satisfaction increased, overall place atta chment increased strongly ( = .48, t-value = 9.20). Perception of pristine conditions ( = -.06, t-value = -1.52) and place satisfaction ( = .01, t-value = -0.19) did not have a significant direct impact on future behavioral intentions against potential future change in recreation settings, respectively a nd their influences on behavior were negative. Place attachment was predictive of greater intentions toward placeprotective behavior for the forest ( = .21, t-value = 4.67): visitors w ith strong place attachment were more likely to be sensitive to future changes in recreation settings. Additional Findings: Indirect Effects As shown in the results in Table 4-9, the re lationship between per ception of settings in the forest and place satisfaction was mediat ed by recreation experience gap scores ( = -.04). This negative effect indicates that visitors perceiving the recreation setting to be less primitive tended to be less satisfied with overall recreation settings in the forest, although the mediating effect is minor. Recreation experience gap scores place satisfaction, and recr eation experience gap scores and place satisfaction mediated the relationshi p between perception of settings and place attachment to the forest ( = -.18). This finding indicates that visitors with negative setting perceptions tended to have lower place attachment. This ma y indicate the importance of mediating variables on place attachment in improvi ng place attachment att itudes. In addition, the relationship between recreation experience gap scores and place attachment was mediated by place satisfaction ( = .11).

PAGE 100

100 Regarding mediating effects on place-prot ective behavior, the relationship between perception of settings and place-protective behavior was negatively mediated by multiple mediating variables (e.g., recreation experience gap scores, place satisfaction, place attachment, recreation experience gap scores and place sati sfaction, recreation experience gap scores and place attachment) ( = -.06). The mediating variables, pl ace satisfaction, place attachment, and place satisfaction and place attachment, positively mediated the relationship between recreation experience gap scores and pl ace-protective behavior ( = .02). Place satisfaction had an indirect effect on place-protective beha vior via place attachment ( = .10). The positive mediating effect indicates that visitors satisfied with settings in the forest were more likely to have placeprotective behavioral inten tions, although the mediating effect was very small. Additional Results of Sp atial Diag nostic Tests H19: There is spatial autocorrelation in regression model residuals in predicting the effects of the importance of REP, PM, PS, and PA on PB. Table 4-10 shows spatial diagnostic test resu lts. Model 1 found signi ficant relationships between the predictors and place-protective behavior ( F = 8.561, p < .01, Adjusted R = .27). The coefficients on perceptions of a natural se tting in recreation areas (PM Pristine) ( = .276, p < .05) and place attachment to r ecreation areas of Ocala NF ( = .464, p < .01) were positive signs and both were significant. Standardized regression coefficient of place attachment was higher than that of perception of natural settings in recreation areas. Overall, twenty-seven percent of variance in place-protective behavior was accounte d for by those significant predictors. All other predictors in Model 1 were not significant.

PAGE 101

101 H20: There is spatial autocorrelation in regression model residuals in predicting the effects of the attainment of REP, PM, PS, and PA on PB. As shown in Table 4-11, Model 2 found signifi cant contributions of the explanatory variables on place-protective behavior ( F = 11.712, p < .001, Adjusted R = .56). Place-protective behavior was positively predicted by attainment of enjoying nature ( = .803, p < .001), perception of natural setting (PM Pristine) in Ocala NF ( = .222, p < .05), and place attachment ( = .696, p < .001). On the other hand, place-prot ective behavior was negatively influenced by attainment of nostalgia ( = -.571, p < .01) and satisfaction with physical settings in the forest ( = -.309, p < .05). Overall, the significant predicto rs explained fifty-six percent of variance in place-protective beha vior. All other predictors in Model 2 were not significant. H21: There is spatial autocorrelation in regression model residuals in predicting the effects of gap scores of REP, PM, PS, and PA on PB. As shown in Table 4-12, Model 3 also found si gnificant contributions of the explanatory variables on place-protective behavior ( F = 11.133, p < .001, Adjusted R = .49). Place-protective behavior was positively predicted by gap scores of enjoying nature ( = .526, p < .001) and place attachment ( = .474, p < .01). On the other hand, place-pro tective behavior was negatively influenced by gap scores of nostalgia ( = -.289, p < .05) and gap scores of escaping physical pressure ( = -.396, p < .01). Overall, those significant pr edictors accounted for forty-nine percent of variance in place-protective behavior. All other explanatory variables in Model 3 were not significant. Additionally, determining whether those three models were specified correctly or not, this study implemented spatial autocorrelation of Morans I. As shown in Table 4-13, this study found no significant spatial autoco rrelation in the residuals on Model 1 (Morans I = -.053, p <

PAGE 102

102 .001), Model 2 (Morans I = -.004, p < .001), and Model 3 (Morans I = -.019, p < .001), respectively. As Odland (1988) indicated, these results demonstrated that specification of variables in the models was correct. In other word s, those models were appropriate to interpret the coefficients of the significant predictors th at explain place-protectiv e behavior because so significant spatial autocorrelation in the regression residuals wa s found. Figures 4-4, 4-5, and 4-6 also support these findings in Model 1, Mode l 2, and Model 3, respectively. Additionally, Morans I test found a st atistically significant clustered spatial pattern on place-protective behavior (Morans I = -.067, p < .10). Figure 4-7 also support this finding. The implications and interpretations of these findings are explored in the di scussion chapter.

PAGE 103

103 Table 4-1. Demographi cs of respondents Characteristics N Percentagea Gender 934 Male 481 51.5 Female 453 48.5 Age 924 18-20 32 3.5 21-30 126 13.6 31-40 211 22.8 41-50 250 27.1 51-60 160 17.3 61-70 97 10.5 71 and more 48 5.2 Education 926 Less than high school 21 2.3 High school diploma 198 21.4 Attended business/technical school 104 11.2 Some college or 2 year degree 265 28.6 Completed 4 year college degree 167 18.0 Some graduate work 44 4.8 Completed graduate or advanced degree 127 13.7 Income 859 Less than $14,999 42 4.9 $15,000 to $34,999 135 15.7 $35,000 to $49,999 154 17.9 $50,000 to $64,999 152 17.7 $65,000 to $99,999 206 24.0 $100,000 to $149,999 116 13.5 $150,000 to $199,999 28 3.3 Over $200,000 26 3.0 Race/Ethnicity 925 Caucasian (White) 885 95.7 African American (Black) 5 .5 Latino (Hispanic) 27 2.9 Asian/Pacific Islander 2 .2 Native American/Indian 4 .4 Other 2 .2 *Mean = 45.12, SD=14.45 aPercentages may not equal 100 because of rounding.

PAGE 104

104 Table 4-2. Differences between respondents and non -respondents Variables Mean SD N t p R NR R NR R R NR Group size 6.17 6.92 12.29 7.28 914 1025 -2.34 .019 Days spent at Ocala NF for the trip 4.29 2.40 17.83 12.29 928 1034 2.57 .010 Age 45.12 37.53 14.45 12.17 924 1037 12.49 <.001 R denotes respondents. NR denotes non-respondents. Table 4-3. ANOVA to determine differences in behavioral constructs among different users Variables0 Overall mean OHV Swimming Kayaking/ Canoeing F -value Gap Score of Recreation Experience Preferences1 Achievement 0.29 0.29 0.25 0.43 0.880 Nostalgia 0.22 0.09 0.28 0.20 1.971 Escape of Social Pressure 0.19 0.16 0.15 0.45 2.967 Enjoyment of Nature 0.05 0.06 -0.01a 0.28b 3.883* Escape of Physical Pressure -0.22 0.08a -0.35b -0.23 7.599** Place Meanings Specific Meanings Pristine2 2.15 2.28a 2.16a 1.86b 8.640*** Place Attachment3 Place Dependence 5.04 5.22 4.99 4.91 2.128 Place Attachment 4.36 4.44 4.36 4.19 1.044 Place Identity 4.36 4.33 4.37 4.38 0.800 Place Satisfaction4 Physical Settings 4.15 3.74a 4.28b 4.39b 57.123*** Managerial Settings 3.50 3.36a 3.54b 3.61b 3.073* Social Settings 3.36 3.42 3.32 3.45 1.626 Place-Protective Behavior5 Environmental 4.87 4.46a 5.01b 5.03b 8.425*** Managerial 4.13 4.92a 3.90b 3.60b 32.764*** Social OHV Riders 4.12 3.12a 4.45b 4.60b 38.599*** Social General Recreators 3.62 3.05a 3.87b 3.60b 16.408*** 0Items for each variable were the same ite ms used for final model in Table 4-6. 1Values were obtained after subtracting importa nce of REP from attainment of REP. Res ponse code: 1= not at all important to 5= extremely important; Response code: 1= did not attain to 5= totally attained 2Response code: 1= more pristine to 5= less pristine 3Response code: 1= strongly disa gree to 7= strongly agree 4Response code: 1= extremely dissatisf ied to 5= extremely satisfied 5Response code: 1= extremely unlikely to 7= extremely likely a,bMeans with different superscripts differ significantly.

PAGE 105

105 Table 4-4. Descriptive statisti cs for the measurement model Variables Mean SD Skewness Kurtosis Importance of Recreation Experience Preferencesa Enjoy Nature (2 items) 4.16 View the scenery 4.21 0.91 -1.19 1.31 Enjoy the smells and sounds of nature 4.11 0.98 -1.09 0.79 Escape of Social Pressure (2 items) 3.83 Release or reduce built-up tensions 3.52 1.29 -0.52 -0.83 Get away from the usual demands of life 4.13 1.03 -1.13 0.74 Escape of Physical Pressure (2 items) 3.69 Experience tranquility 3.90 1.11 -0.86 0.05 Experience solitude 3.48 1.33 -0.48 -0.91 Nostalgia (2 items) 3.66 Think about good times I've had in the past 3.44 1.23 -0.38 -0.82 Bring back pleasant memories 3.88 1.09 -0.80 -0.06 Use Equipment (2 items) 2.73 Use my recreation equipment 3.28 1.46 -0.30 -1.26 Talk to others about our/my recreation equipment 2.17 1.33 0.84 -0.55 Achievement (2 items) 2.44 Show others I can do it 2.06 1.31 0.96 -0.34 Develop my skills and abilities 2.82 1.28 0.13 -0.97 Taking Risks (2 items) 2.01 Take risks 1.98 1.27 1.10 0.03 Chance dangerous situations 2.04 1.30 1.02 -0.17 Attainment of Recreation Experience Preferencesb Enjoy Nature (2 items) 4.19 View the scenery 4.29 0.87 -1.23 1.23 Enjoy the smells and sounds of nature 4.09 0.93 -0.86 0.27 Escape of Social Pressure (2 items) 4.02 Release or reduce built-up tensions 3.78 1.13 -0.77 -0.10 Get away from the usual demands of life 4.26 0.89 -1.26 1.51 Nostalgia (2 items) 3.88 Think about good times I've had in the past 3.79 1.10 -0.63 -0.36 Bring back pleasant memories 3.97 1.03 -0.87 0.15

PAGE 106

106 Table 4-4. Continued Escape of Physical Pressure (2 items) 3.46 Experience tranquility 3.69 1.14 -0.59 -0.50 Experience solitude 3.22 1.32 -0.24 -1.07 Use Equipment (2 items) 2.92 Use my recreation equipment 3.41 1.54 -0.49 -1.28 Talk to others about our/my recreation equipment 2.42 1.40 0.51 -1.05 Achievement (2 items) 2.76 Show others I can do it 2.55 1.38 0.37 -1.11 Develop my skills and abilities 2.97 1.27 -0.04 -0.98 Taking Risks (2 items) 1.96 Take risks 1.93 1.24 1.17 0.24 Chance dangerous situations 1.98 1.25 1.05 -0.06 Gap Score of Recreation Experience Preferencesc Achievement (2 items) 0.32 Show others I can do it 0.49 1.52 -0.04 0.57 Develop my skills and abilities 0.15 1.36 0.02 0.54 Nostalgia (2 items) 0.22 Think about good times I've had in the past 0.34 1.30 0.03 0.41 Bring back pleasant memories 0.10 1.25 0.21 0.75 Escape of Social Pressure (2 items) 0.20 Release or reduce built-up tensions 0.26 1.40 0.13 0.54 Get away from the usual demands of life 0.13 1.14 0.40 1.27 Use Equipment (2 items) 0.20 Use my recreation equipment 0.14 1.47 -0.12 0.72 Talk to others about our/my recreation equipment 0.25 1.52 -0.12 0.75 Enjoy Nature (2 items) 0.03 View the scenery 0.08 1.03 0.12 2.25 Enjoy the smells and sounds of nature -0.02 1.13 0.31 1.50 Taking Risks (2 items) -0.05 Take risks -0.04 1.28 -0.18 1.81 Chance dangerous situations -0.06 1.40 0.05 1.43 Escape of Physical Pressure (2 items) -0.23 Experience tranquility -0.21 1.35 -0.07 0.56 Experience solitude -0.25 1.54 -0.03 0.04

PAGE 107

107 Table 4-4. Continued Place Meanings Specific Meanings Impacted (6 items) d 3.12 This area has many developed recreation facilities and services 2.91 1.11 0.14 -0.57 This area is extremely crowded 3.08 0.99 -0.04 -0.33 This area has changed a lot over the years 2.94 0.96 -0.09 0.43 This area has many recreationists using it 2.25 1.02 0.62 -0.11 This area is extremely developed 3.44 0.95 -0.29 -0.11 The water in this area is extremely polluted 4.07 1.08 -1.19 0.91 Pristine (5 items) e 2.17 This area is extremely scenic 2.00 1.01 0.93 0.34 This area is extremely forested 2.22 1.02 0.61 -0.13 This area is extremely peaceful 2.35 1.10 0.58 -0.36 The water in this area is extremely clear 2.12 1.19 0.84 -0.20 This area has many species of wildlife and plants 2.15 1.00 0.74 0.15 Place Attachmentf Place Dependence (4 items) 4.75 This area is not a good place to do the things I most like to do* 5.44 1.60 -0.83 -0.32 This area is the best place for doing the things that I enjoy most 4.62 1.56 -0.38 -0.42 For doing the things that I enjo y most, no other place can compare to this area 4.43 1.51 -0.22 -0.29 As far as I am concerned, there are better places to be than at this area* 4.51 1.61 -0.11 -0.73 Place Attachment (4 items) 4.73 I feel relaxed when I'm at this area 5.88 1.15 -1.22 1.63 I feel happiest when I'm at this area 4.53 1.46 -0.37 -0.11 I really miss this area when I' m away from it for too long 4.22 1.72 -0.19 -0.75 This area is my favorite place to be 4.27 1.53 -0.21 -0.32 Place Identity (4 items) 4.68 Everything about this area is a reflection of me 4.25 1.43 -0.11 -0.14 I feel that I can really be myself at this area 4.95 1.45 -0.51 0.00 This area reflects the type of person I am 4.48 1.51 -0.34 -0.31 This area says very little about who I am* 5.04 1.53 -0.52 -0.32 Place Satisfactiong Physical Settings (4 items) 4.07 Its scenery 4.36 0.73 -1.15 1.65 Populations of wildlife around it 3.84 0.89 -0.49 -0.18 Its water quality 4.17 0.92 -0.98 0.48 Its level of land development 3.91 0.89 -0.47 -0.24

PAGE 108

108 Table 4-4. Continued Managerial Settings (3 items) 3.69 Recreation fee cost 3.60 1.11 -0.50 -0.45 Its level of rules and regulations 3.54 1.05 -0.33 -0.54 Its facilities and services 3.92 0.98 -0.82 0.14 Social Settings (3 items) 3.58 Its solitude and peacefulness 3.96 0.97 -0.93 0.52 The number of people using it 3.47 1.00 -0.32 -0.36 Other peoples recreation activities 3.31 0.84 0.15 0.54 Place-Protective Behaviorh Place-Protective Behavior: Environmental (2 items ) 4.91 If the natural environmental quality of the area I recreated in showed major deterioration, I would vote for or support rules that might keep it from environmentally deteriorating. 5.52 1.56 -1.16 0.98 If the natural environmental quality of the area I recreated in showed major deterioration, I would join or help form a group to fight for more environmental protection of the area. 4.30 1.93 -0.29 -0.91 Place-Protective Behavior: Soci al OHV Riders (2 items ) 4.20 If the number of Off-Highway Vehicle (OHV)s around here increased greatly, I would vote for regulations that would limit the number of OHVs here. 4.65 2.07 -0.46 -1.02 If the number of Off-Highway Vehicle (OHV)s around here increased greatly, I would join or help form a group to reduce the number of OHVs. 3.75 2.03 0.07 -1.10 Place-Protective Behavior: Managerial (2 items ) 4.08 If the rules on the kinds of recreation that can take place here were to become more restrictive, I would vote for or support rules that might keep it from becoming more restrictive. 4.51 1.75 -0.32 -0.60 If the rules on the kinds of recreation that can take place here were to become more restrictive, I would join or help form a group to fight such restrictions. 3.65 1.93 0.07 -0.96 Place-Protective Behavior: Social Recreators (2 items ) 3.69 If the number of recreators around here increased greatly, I would vote for regulations that would limit the number of recreators here. 4.02 1.76 -0.18 -0.76 If the number of recreators around here increased greatly, I would join or help form a group to reduce the number of recreators. 3.35 1.75 0.16 -0.76 *Negative items were reverse coded before calculating statistics. aResponse code: 1= not at all impor tant to 5= extremely important bResponse code: 1= did not attain to 5= totally attained cValues were obtained after s ubtracting importance of REP fr om attainment of REP. dResponse code: 1= more pristine to 5= less pristine eResponse code: 1= more impacted to 5= less impacted fResponse code: 1= strongly di sagree to 7="strongly agree gResponse code: 1= extremely dissatisf ied to 5= extremely satisfied hResponse code: 1= extremely unlik ely to 7= extremely likely

PAGE 109

109 Table 4-5. Final fit indices of measurem ent models before structural models Variables 2 df Chisquare/df RMSEA CFI IFI NFI GFI CN (.05/.01) REP 231.173 56 4.128 .058 .965 .965 .954 .966 301/338 REPAttained 221.070 56 3.948 .056 .968 .968 .958 .968 315/353 GAP 181.026 56 3.233 .049 .961 .961 .945 .974 384/431 PM 0.722 2 0.361 .000 1.000 1.000 .999 1.000 7740/11898 PA 15.196 11 1.381 .020 .999 .999 .996 .995 1209/1519 PS 35.786 11 3.253 .049 .984 .984 .977 .989 513/645 Table 4-6. Final factor load ings for the measurement mode l before structural models Variables CR AVE t-Value Importance of Recreat ion Experience Preferencesa Achievement (2 items) .69 .69 .52 Show others I can do it .71 Develop my skills and abilities .74 19.40 Nostalgia (2 items) .71 .74 .59 Think about good times I've had in the past .84 Bring back pleasant memories .66 15.59 Escape of Social Pressure (2 items) .76 .77 .63 Release or reduce built-up tensions .79 Get away from the usual demands of life .80 18.61 Use Equipment (2 items) .63 .64 .48 Use my recreation equipment .60 Talk to others about our/my recreation equipment .77 15.49 Taking Risks (2 items) .88 .88 .78 Take risks 0.86 Chance dangerous situations 0.91 25.83 Attainment of Recreation Experience Preferencesb Achievement (2 items) .74 .74 .59 Show others I can do it .73 Develop my skills and abilities .80 20.93 Enjoy Nature (2 items) .62 .62 .45 View the scenery .68 Enjoy the smells and sounds of nature .67 14.47

PAGE 110

110 Table 4-6. Continued Nostalgia (2 items) .76 .76 .61 Think about good times I've had in the past .80 Bring back pleasant memories .77 17.90 Escape of Social Pressure (2 items) .73 .74 .59 Release or reduce built-up tensions .76 Get away from the usual demands of life .77 18.69 Escape of Physical Pressure (2 items) .76 .78 .65 Experience tranquility .91 Experience solitude .68 17.43 Gap Score of Recreation Experience Preferencesc Achievement (2 items) .59 .59 .42 Show others I can do it .62 Develop my skills and abilities .68 14.41 Enjoy Nature (2 items) .55 .55 .38 View the scenery .63 Enjoy the smells and sounds of nature .60 11.72 Nostalgia (2 items) .62 .62 .45 Think about good times I've had in the past .73 Bring back pleasant memories .61 12.59 Escape of Social Pressure (2 items) .66 .67 .50 Release or reduce built-up tensions .67 Get away from the usual demands of life .74 14.36 Escape of Physical Pressure (2 items) .69 .71 .56 Experience tranquility .85 Experience solitude .62 13.74 Place Meanings Specific Meaningsd Pristine (5 items) .79 .79 .49 This area is extremely scenic .81 This area is extremely forested .69 18.45 This area is extremely peaceful .63 17.23 This area has many species of wildlife and plants .67 17.96

PAGE 111

111 Table 4-6. Continued Place Attachmente Place Attachment (3 items) .84 .84 .64 I feel happiest when I'm at this area .83 I really miss this area when I' m away from it for too long .75 25.05 This area is my favorite place to be .82 28.22 Place Dependence (2 items) .75 .78 .65 This area is not a good place to do the things I most like to do* .64 This area is the best place for doing the things that I enjoy most .95 27.53 Place Identity (2 items) .82 .82 .69 Everything about this area is a reflection of me .83 This area reflects the type of person I am .83 19.94 Place Satisfactionf Physical Settings (3 items) .69 .66 .39 Its scenery .55 Its water quality .58 12.06 Its level of land development .74 13.22 Social Settings (2 items) .73 .74 .59 The number of people using it .81 Other peoples recreation activities .72 13.70 Managerial Settings (2 items) .64 .65 .48 Recreation fee cost .61 Its level of rules and regulations .77 12.84 Place-Protective Behaviorg Place-Protective Behavior (8 items) (See Table 4-4) .79 na na na na *Negative items were reverse coded before calculating statistics. aResponse code: 1= not at all impor tant to 5= extremely important bResponse code: 1= did not attain to 5= totally attained cValues were obtained after s ubtracting importance of REP fr om attainment of REP. dResponse code: 1= more pristine to 5= less pristine eResponse code: 1= strongly di sagree to 7="strongly agree fResponse code: 1= extremely dissatisf ied to 5= extremely satisfied gResponse code: 1= extremely unlik ely to 7= extremely likely

PAGE 112

112 Table 4-7. Direct a nd indirect effects in structural model 1 Hypothesis CR SE H1: Importance of REP Place Meanings -.04 -0.90 .04 H2: Importance of REP Place Satisfaction -.09* -2.26 .02 H3: Importance of REP Place Attachment .31** 8.73 .04 H4: Place Meanings Pace Satisfaction -.39** -8.29 .03 H5: Place Meanings Pace Attachment -.14** -3.53 .06 H6: Place Satisfaction Place Attachment .51** 10.19 .11 H7: Importance of REP Place-Protective Behavior -.05 -1.21 .05 H8: Place Meanings Place-Protective Behavior -.06 -1.38 .06 H9: Place Satisfaction Place-Protective Behavior -.04 -0.66 .12 H10: Place Attachment Place-Protective Behavior .23** 4.58 .05 Indirect Effect Importance of REP Place Meanings Place Satisfaction .01 Importance of REP Place Meanings Place Attachment; -.04 Importance of REP Place Satisfaction Place Attachment; Importance of REP Place Meanings Place Satisfaction Place Attachment Place Meanings Place Satisfaction Place Attachment -.20** Importance of REP Place Meanings Place-Protective Behavior; .07 Importance of REP Place Satisfaction Place-Protective Behavior; Importance of REP Place Attachment Place-Protective Behavior; Importance of REP Place Satisfaction Place Attachment PlaceProtective Behavior; Importance of REP Place Meanings Place Satisfaction Place Attachment Place-Protective Behavior Place Meanings Place Satisfaction Place-Protective Behavior; -.06 Place Meanings Place Attachment Place-Protective Behavior; Place Meanings Place Satisfaction Place Attachment PlaceProtective Behavior Place Satisfaction Place Attachment Place-Prot ective Behavior .12** *Significant at p < .05 level. **Significant at p < .01 level.

PAGE 113

113 Figure 4-1. Final structural model 1 -.04 -.05 -.39** .51** -.06 .23** -.14** -.04 -.09* .31** Importance of REP Place Satisfaction Place Meanings Place Attachment PlaceProtective Behavior Fit Statistics *Beta values on solid lines significant at .05 level X2/df = 4.653 **Beta values on solid lines significant at .01 level CFI = .926 Dotted lines denote insignificant paths at the .05 level RMSEA = .063

PAGE 114

114 Table 4-8. Direct a nd indirect effects in structural model 2 Hypothesis CR SE H11: Place Meanings Attainment of REP -.34** -7.08 .03 H12: Attainment of REP Place Satisfaction .38** 6.80 .06 H13: Attainment of REP Place Attachment .38** 7.53 .11 H14: Attainment of REP Place-Protective Behavior .08 1.69 .11 Same Direct Effect as Model 1 H4a: Place Meanings Pace Satisfaction -.26** -5.59 .03 H5a: Place Meanings Pace Attachment -.10* -2.52 .05 H6a: Place Satisfaction Place Attachment .33** 6.82 .10 H8a: Place Meanings Place-Protective Behavior -.05 -1.15 .06 H9a: Place Satisfaction Place-Protective Behavior -.04 -0.84 .11 H10a: Place Attachment Place-Protective Behavior .18** 3.55 .05 Indirect Effect Place Meanings Attainment of REP Place Satisfaction -.13** Place Meanings Attainment of REP Place Attachment; -.25** Place Meanings Place Satisfaction Place Attachment; Place Meanings Attainment of REP Place Satisfaction Place Attachment Attainment of REP Place Satisfaction Place Attachment .12** Place Meanings Attainment of REP Place-Protective Behavior; -.07 Place Meanings Place Satisfaction Place-Protective Behavior; Place Meanings Place Attachment Place-Protective Behavior; Place Meanings Attainment of REP Place Satisfaction Place Attachment Place-Protective Behavior; Place Meanings Attainment of REP Place Attachment PlaceProtective Behavior Attainment of REP Place Satisfaction Place-Protective Behavior; .07 Attainment of REP Place Attachment Place-Protective Behavior; Attainment of REP Place Satisfaction Place Attachment PlaceProtective Behavior Place Satisfaction Place Attachment Place-Prot ective Behavior .06 *Significant at p < .05 level. **Significant at p < .01 level. aThe same hypotheses are shown in Model 1 and not examined in Model 2.

PAGE 115

115 Figure 4-2. Final structural model 2 -.04 -.05 .38** .33** .08 .18** .38** -.34** -.26** -.10* Place Meanings Place Satisfaction Attainment of REP Place Attachment PlaceProtective Behavior Fit Statistics *Beta values on solid lines significant at .05 level. X2/df = 4.047 **Beta values on solid lines significant at .01 level. CFI = .939 Dotted lines denote insignificant paths at the .05 level RMSEA = .057

PAGE 116

116 Table 4-9. Direct a nd indirect effects in structural model 3 Hypothesis CR SE H15: Place Meanings GAP Scores -.18** -4.17 .03 H16: GAP Scores Place Satisfaction .24** 4.88 .04 H17: GAP Scores Place Attachment -.01 -0.17 .08 H18: GAP Scores Place-Protective Behavior -.06 -1.58 .08 Same Direct Effect as Model 1 H4a: Place Meanings Pace Satisfaction -.34** -7.35 .03 H5a: Place Meanings Pace Attachment -.17** -4.09 .06 H6a: Place Satisfaction Place Attachment .48** 9.20 .11 H8a: Place Meanings Place-Protective Behavior -.06 -1.52 .06 H9a: Place Satisfaction Place-Protective Behavior -.01 -0.19 .11 H10a: Place Attachment Place-Protective Behavior .21** 4.67 .05 Indirect Effect Place Meanings GAP Scores Place Satisfaction -.04 Place Meanings GAP Scores Place Attachment; -.18** Place Meanings Place Satisfaction Place Attachment; Place Meanings GAP Scores Place Satisfaction Place Attachment GAP Scores Place Satisfaction Place Attachment .11** Place Meanings GAP Scores Place-Protective Behavior; -.06 Place Meanings Place Satisfaction Place-Protective Behavior; Place Meanings Place Attachment Place-Protective Behavior; Place Meanings GAP Scores Place Satisfaction Place Attachment Place-Protective Behavior; Place Meanings GAP Scores Place Attachment Place-Protective Behavior GAP Scores Place Satisfaction Place-Protective Behavior; .02 GAP Scores Place Attachment Place-Protective Behavior; GAP Scores Place Satisfaction Place Attachment PlaceProtective Behavior Place Satisfaction Place Attachment Place-Prot ective Behavior .10** *Significant at p < .05 level. **Significant at p < .01 level. aThe same hypotheses are shown in Model 1 and not examined in Model 3.

PAGE 117

117 Figure 4-3. Final structural model 3 -.01 -.06 .24** .48** -.06 .21** -.01 -.18** -.34** -.17** Place Meanings Place Satisfaction GAP Scores Place Attachment PlaceProtective Behavior Fit Statistics *Beta values on solid lines significant at .05 level X2/df = 3.500 **Beta values on solid lines significant at .01 level CFI = .944 Dotted lines denote insignificant paths at the .05 level RMSEA = .052

PAGE 118

118 Table 4-10. Results of regression m odel 1 for spatial diagnostic test Regression model 1 B t-value Intercept 1.951 3.500** Importance of REP Achievement ns ns ns Importance of REP Nostalgia ns ns ns Importance of REP Escaping social pressurens ns ns Importance of REP Using equipment ns ns ns Importance of REP Taking risk ns ns ns PM Pristine .381 .276 2.086* PA Place attachment .338 .464 3.506** PA Place dependence ns ns ns PA Place identity ns ns ns PS Physical settings ns ns ns PS Social settings ns ns ns PS Managerial settings ns ns ns R .30 Adjusted R .27 F2 42 8.561** Dependent variable: PB (n=43) *p < .05, ** p < .01, *** p < .001 ns=non-significant at p < .05 Table 4-11. Results of regression m odel 2 for spatial diagnostic test Regression model 2 B t-value Intercept 2.335 2.863** Attainment of REP Achievement ns ns ns Attainment of REP Enjoying nature .962 .803 5.431*** Attainment of REP Nostalgia -.710 -.571 -3.625** Attainment of REP Escaping social pressure ns ns ns Attainment of REP Escaping physical pressurens ns ns PM Pristine .307 .222 2.089* PA Place attachment .507 .696 5.461*** PA Place dependence ns ns ns PA Place identity ns ns ns PS Physical settings -.532 -.309 -2.615* PS Social settings ns ns ns PS Managerial settings ns ns ns R .61 Adjusted R .56 F5 42 11.712*** Dependent variable: PB (n=43) *p < .05, ** p < .01, *** p < .001 ns=non-significant at p < .05

PAGE 119

119 Table 4-12. Results of regression m odel 3 for spatial diagnostic test Regression model 3 B t-value Intercept 2.749 7.241*** GAP scores of REP Achievement ns ns ns GAP scores of REP Enjoying nature .772 .526 3.995*** GAP scores of REP Nostalgia -.312 -.289 -2.076* GAP scores of REP Escaping social pressure ns ns ns GAP scores of REP Escaping physical pressure-.373 -.396 -3.363** PM pri1235 ns ns ns PM Pristine .346 .474 3.703** PA Place attachment ns ns ns PA Place dependence ns ns ns PA Place identity ns ns ns PS Physical settings ns ns ns PS Social settings ns ns ns R .61 Adjusted R .56 F5 42 11.712*** Dependent variable: PB (n=43) *p < .05, ** p < .01, *** p < .001 ns=non-significant at p < .05 Table 4-13. Results of sp atial autocorrelation tests SA Test Std. residuals for model 1 Std. residuals for model 2 Std. residuals for model 3 Dependent variable: PB Moran's index -0.05367 -0.00481 -0.01978 -0.06731 Expected index -0.02381 -0.02381 -0.02381 -0.02381 Variance 0.00067 0.00068 0.000672 0.00067 Z score -1.14914 0.72467 0.155223 -1.67882* *p < .10

PAGE 120

120 Figure 4-4. Results of spatial autocorrelation test for model 1

PAGE 121

121 Figure 4-5. Results of spatial autocorrelation test for model 2

PAGE 122

122 Figure 4-6. Results of spatial autocorrelation test for model 3

PAGE 123

123 Figure 4-7. Results of spatial anal ysis of place-pro tective behavior

PAGE 124

124 CHAPTER 5 DISCUSSION Summary of Procedures, Sampling and Analyses The prim ary purpose of this dissertation sought to examine three major structural models designed to better understand visito r attitudes, which consisted of recreation experiences, sense of place constructs, and place-protective behaviors of visitors at major recreation areas in Ocala NF during the spring of 2008. Intercepted onsite surv eys of visitors at Ocala NF were conducted between March 1 and May 31, 2008 and three follow-up mail contacts were made, following a Dillmans modified method. The final sample for this study was composed of 934 Ocala NF visitors. Any visitors who did not participate in both the on-site and mail-back surveys were not represented in this study. Summary of Study Objectives and Major Findings Structural Models As stated in objective 1, the firs t structur al model included three major concepts: 1) importance of recreation expe rience preferences (REP), 2) place meanings (PM), place satisfaction (PS), and place attachment (PA) within perceptions of sense of place (SOP), and 3) place-protective behavior (PB). When it comes to comparing the results of identifying factors that influence placeprotective behavior in this st udy with those in the literature, no statistically significant relationship was found between im portance of recreation experi ence preferences and placeprotective behavior, between percep tion of settings in the forest and place-protective behavior, or between satisfaction with settings in the fo rest and place-protectiv e behavior, while the predictors were negatively rela ted to place-protective behavior However, attachment to the forest had a significantly positive effect on place-pr otective behavior. In more detail, this study

PAGE 125

125 found that place attachment is incl ined to foster place-protective behavior, which is consistent with the finding reported by Stedman (2002). Visitors who were more attached to the forest were more willing to participate in protesting against futu re changes in recreation settings in the forest. However, unlike Stedmans (2003) finding, visitors who were more satisfied with recreation settings in the forest had no signi ficant willingness to resist potent ial future changes in recreation settings in the forest. Rather, those visitors tended to express thei r unwillingness to resist although the relationship between the two was w eak and not significant. Additionally, specific perceptions of the recrea tion setting had no significant influenc e on behavioral intentions against potential future changes in forest recreation settings. It is interesting that visitors motivation was not strongly related to behaviors. That is, visitors sought recreati on experiences in the forest but they were not likely to participate in place-p rotective behaviors that might assure quality recreational settings in the future. Another objective was to identify factors that directly influence sense of place constructs (place meanings, place satisfaction, place atta chment). The results indicated that place attachment is significantly influenced by the level of importance for recreation experience preferences, specific perceptions of settings, and level of satisfac tion. More specifically, visitors whose recreational motivation was strong, had a si gnificantly more positiv e attachment to the forest. Additionally, higher leve ls of visitor satisfaction grea tly contributed to their place attachment to the forest. That is, if they were sa tisfied with recreation settings in the forest while seeking their recreation experiences they were more inclined to hold a strong attachment to the forest. This finding is consiste nt with findings reported by Me sh and Manor (1998). Importantly, those who perceived natural settings in the forest to be less pristine, were less inclined to have a positive attachment to the forest, which is in lin e with findings reported by Stedman (2002). That

PAGE 126

126 is, more pristine recreation settings tend to foster positive attachments to the forest. Additionally, visitors with higher motivation generally did not report negative perceptions of recreation settings. However, it is interes ting that those visitors with hi gher motivation expressed their dissatisfaction with recreati on settings in the forest. Regarding mediating effects, ratings of r ecreation experience im portance did not have significant indirect effects on satisfaction with re creation settings, attachme nt to the forest, and place-protective behavior. Perceptio ns of recreation settings in the forest had a mediating effect on attachment to the forest via sa tisfaction with the recreation sett ings. This mediating effect of setting perception on place attach ment through place satisfaction ( = -.20) is greater than the direct effect of setting percepti on on attachment to the forest ( = -.14). This finding demonstrates the full mediating effect of setti ng perception on place attachment and indicates the importance of place satisfaction in predicting pl ace attachment. In addition, place satisfaction had a significantly positive mediating effect on place protective behavior via place attachment, while the direct effect of place satisfaction on place-protective behavior was neither positive nor significant. In other words, as satisfaction w ith recreation settings in creased, place-protective behavior increased through attach ment to recreation areas. This mediating effect verifies the importance of place satisfaction and place attachment in terms of place-protective behavior. As stated in objective 2, the second structural model focu sed on two aspects: a) the impacts of PM on the attainment of REP, PS, an d PA, and b) the impacts of REP attainment on PS, PA, and PB. In summary, when it comes to comparing the re sults of identifying factors that influence place-protective behavior in this study with those in the literature, no significant relationship was found between attainment of recreation experien ce preferences and placeprotective behavior,

PAGE 127

127 between perceptions of recreation settings and place-protective behavior, or between satisfaction with recreation settings and pl ace-protective behavior. Howeve r, degree of attachment to recreation areas had a significantly positive effect on place-protective behavior. Specifically, this study found that place attachment likely fosters place-protective behavior, which is consistent with findings reported by Stedman (2002). Visitors who were more attached to recreation areas were more willing to participate in resisting against future changes in recreation settings. However, differing from findings reported by Stedman (2003), visitors who were more satisfied with recreation settings in the forest generally had no significant willingness to resist future development of recreation settings. Rather, forest visitors expressed their unwillingness to participate although the relations hip between the two was very weak and not significant. Additionally, specific perceptions of the recreation setting ha d no significant influence on behavioral intentions against pote ntial future changes in recreation settings. It is interesting that visitors attainment of quality recreation experiences was not strongly related to such behaviors. That is, visitors achieved their recreation experien ces in the forest but those experiences were not significantly related to their intention to participate in similar future behaviors. Another objective was to identify factors that influence sense of place constructs, place meanings, place satisfaction, place attachment. Th e results indicated that place attachment is significantly influenced by attainme nt of preferred recreation experi ences, specific perceptions of recreation settings, and higher sati sfaction with recreati on settings. Specifically, visitors that achieved their recreation experiences, reported more positive attachments to recreation areas in the forest and the relationship was statistically significant. Additionally, higher levels of visitor satisfaction greatly contributed to a positive attach ment to recreation areas. That is, visitors that were satisfied with recreation settings were more inclined to hold a stronger attachment. This

PAGE 128

128 finding is consistent with findings reported by Mesh and Manor (1998). Importantly, those who perceived the natural settings in recreation areas to be less pristine, were less favorable to a positive attachment, which is in line with findings reported by Stedman (2002). In other words, more pristine recreation settings tend to fost er positive attachments to recreation areas. Additional results found significant and negativ e relationships between setting perceptions and attainment of recreation experiences, and between perception of settings and place satisfaction. That is, visitors who perceived the recreation setting to be less pristine were not able to achieve their recr eation experience and were less sa tisfied with forest settings. However, visitors who achieved their recreation experiences were more likely to be satisfied with forest recreation settings. With regard to mediating effects, negative perceptions of recreati on settings did have significantly negative mediating eff ects on satisfaction with recreati on settings via attainment of recreation experiences. As a dire ct effect of recreation settin g perception on place satisfaction was demonstrated, this mediating e ffect indicates that visitors perc eiving the forest settings to be less pristine were not satisfied w ith recreation settings in the fore st because of less attainment of recreation experiences that result ed from negative perceptions of forest recreation settings. Negative perception of forest recreation settings did have significant negative mediating effects on attachment to recreation areas thr ough attainment of recreation experiences and satisfaction with recreation setti ngs. As a direct effect of recr eation setting perceptions on place attachment was demonstrated, the mediating eff ect indicated that visitors perceiving forest recreation settings to be less pris tine were not significantly attach ed to recreation areas because of less attainment of recreation experiences and negative satisfacti on that resulted from negative perception of recreation settings An additional mediating effect was found between attainment

PAGE 129

129 of recreation experiences and attachment to recreation areas. Visitors who achieved their recreation experience felt more attached to recreati on areas via the effect of their REP attainment on satisfaction with recreation set tings. This finding demonstrates th at as attainment of recreation experiences increased, attachment to recreation areas increased, as i ndicated from its direct effect on place attachment. Finally, no significant effect s of mediating variables on place-protective behavior were found. As stated in objective 3, the s econd structural model focused on two parts: a) the impacts of PM on the REP gap scores, PS, and PA, and b) the impacts of REP gap scores on PS, PA, and PB. A discussion of the findings follows. When it comes to comparing the identification of factors that influence place-protective behavior with those in the lite rature, no significant relationship was found between gap scores of recreation experience preferences and place-protective behaviors, between perceptions of forest recreation settings and place-protec tive behavior, or between satis faction with forest recreation settings and place-protective behavior. However, level of attachment to the forest had a significantly positive effect on place-protective beha vior. Specifically, this study found that place attachment fosters place-protective behavior, which is consistent with a finding reported by Stedman (2002). Visitors with more forest recreation attachment we re more willing to participate in protesting against future changes in recr eation settings. However, unlike a finding from Stedman (2003), visitors that were more satisfied with forest recreation se ttings were less willing to resist potential future changes in recreation settings. Rather, forest visitors expressed their unwillingness to engage in resisting behaviors although the relationship between the two was very weak and not significan t. Additionally, specific setting perceptions had no significant influence on place-protective behavi or against future changes in recreation settings. Interestingly,

PAGE 130

130 visitors gap scores of recreation experiences were not strongly re lated to their future behavior. That is, some visitors achieved higher than expected recreation experiences but that outcome was not significantly related to their intention to participate in su ch future behavior. Although this study sought to identify fact ors that influenced place attachment, no significant relationship was found between gap sc ores of recreation experiences and place attachment. However, place attachment was si gnificantly influenced by specific setting perceptions and higher satisfacti on, respectively. Specifically, visito rs who achieved higher than expected recreation experiences did not report a po sitive attachment to forest recreation settings. Additionally, higher levels of visitors satisf action greatly contributed to a more positive attachment to the forest. That is, visitors who were satisfied with recreation settings in the forest were more inclined to hold a strong attachment to the forest. This finding is consistent with the findings reported by Mesh and Manor (1998). Im portantly, those who perceived the natural forest setting to be less pristine were less attached to recreation areas. This finding is consistent with a finding reported by Stedman (2002). That is more natural forest settings tend to foster positive attachments. Additional results revealed significant and negative relationshi ps between setting perceptions and gap scores of recreation experiences, and between setting perceptions and place satisfaction. In other words, visitors who percei ved the recreation setting to be less pristine, thought that they were not able to attain their recreation experiences more than expected and that they were not satisfied with forest recreation settings. However, visito rs who achieved their recreation experience more than expect ed seemed to be more satisfied. Considering mediating effects, negative per ceptions of settings did not have significant mediating effects on satisfaction with settings vi a gap scores of recreatio n experiences while the

PAGE 131

131 direct effect of perceptions of settings on pl ace satisfaction was significantly negative. This indicates that visitors perceiving the forest to be less pristine we re not satisfied with recreation settings in the forest but the e ffect tends to be reduced via gap scores of recreation experiences. Negative setting perceptions did have si gnificantly negative me diating effects on attachment to recreation areas th rough gap scores of recreation experiences and satisfaction with recreation settings. As a direct effect of setting percepti ons on place attachment was demonstrated, the mediating effect indicates that visitors perceiving the forest to be less pristine were not attached to recreation areas because of poor gap scores of recreation experiences and negative satisfaction that resulted from negative setting perceptions. An additional mediating effect was found between gap scores of recrea tion experiences and att achment to recreation areas. Visitors who achieved recreation experiences higher than they expected felt more attached to recreation areas via the effect of their gap score attainment on satisfaction with recreation settings. This finding demonstrates that as ga p scores of recreation experiences increased, attachment to recreation areas increased, while its direct eff ect on place attachment was not significant. Finally this study al so found a significant mediating effect of place satisfaction on place-protective behavior via its e ffect on attachment to recreation areas, while its direct effect on place-protective behavior was not significant. Visitors who we re satisfied with recreation settings and were attached to the recreation areas tended to show si gnificant willingness to protect forest recreation settings from increased development. Spatial Diagnostic Tests As shown in spatial diagnostic test 1, th is study found a positive relationship between setting perceptions and place-prot ective behavior. That is, visitors who perceived natural forest recreation settings to be less pris tin e were likely to take actions to resist potential changes in forest recreation settings (e.g., more rest rictive rules and regulations on recreation,

PAGE 132

132 environmental deterioration of natural resources, and increasing numbers of OHV riders and other general recreationists). In addition, a pos itive relationship between place attachment and place-protective behavior was found. Visitors who had more positive affective or emotional bonds with recreation areas reported intentions to participate in place-protective behavior. All other predictors were not significant. As shown in spatial diagnostic test 2, atta inment of enjoying nature had a significant positive relation to place-protective behavior, while such behavior was negatively influenced by attainment of nostalgia. Those w ho viewed scenery and enjoyed th e smells and sounds of nature, had more propensities to resist potential fu ture changes in forest recreation settings. Additionally, a positive relati onship between setting perc eption and place-protective behavior emphasizes the importance of natural setti ngs because less pristine recreation settings in the future did not seem to be acceptable to visito rs who considered recreation area settings to be less natural. Among sense of place constructs, emotional attachment to recreation areas had a significant positive effect on place-protective behavior. On the other hand, satisfaction with physical settings is negatively rela ted to place-protective behavior. That is, as long as visitors are satisfied with physical settings (e.g., scenery, level of land devel opment) in the forest, they had no intention of participating in antidevelopment ac tions. All other predictors were not significant. As shown in spatial diagnostic test 3, gap sc ores for enjoyment of nature were positively related to place-protective behavior. Visitors who attained a higher than expected nature experience seemed to be more sensitive to unfa vorable settings in recreation areas and were willing to participate in activities that could protect recreation areas. However, gap scores of nostalgia and escape of physical pressure were negatively relate d. These findings indicate that visitors tended to be indifferent to resisting po tential future changes in recreation settings, as

PAGE 133

133 long as visitors achieved their recreation experi ences (e.g., nostalgia, es cape physical pressure) higher than expected. But visitors emotionally a ttached to recreation area s did express a stronger intention to protect natural-based settings in recreation areas. Discussion In this study sense of place was treated as an overarching umbrella concept that included three major constructs, place meanings, place satisfaction, and place a ttachment. This study sought to identify factors that in fluence a sense of place construc t and place-protective behavior. The importance of recreation experience in Mode l 1, attainment of recreation experience in Model 2, and gap scores of recrea tion experience in Model 3 were used as factors that would influence the sense of place construct as well as the place-protective behavior construct. Additionally, spatial diagnostic tests were im plemented. Some theoretical and empirical discussions were arrive d at, based on major fi ndings reported earlier. Structural Model 1 First of all, this study exam ined the rela tionship between overall recreation motivation and place attachment, and found that overall recr eation motivation was found to be a significant predictor of place attachment. Those who had higher levels of overall motivation reported a higher attachment to forest recreation areas than those who had lower levels of overall motivation. This result is deemed consistent with previous findings reported by Kyle, et al. (2004), although this study used an overall motiva tion construct comprised of diverse recreation experiences. For instance, Kyle, et al. used domains of recreation motivation to determine whether each of the motivation factors contribu ted to development of place attachment and found partial support for a significant relationshi p between the two. Overall, Lawlers (1973) theory of motivation for work organizations seem s to be supported for recreation motivation as well.

PAGE 134

134 There has been a debate about whether a significant relationshi p between expected recreation motivation and place meanings may i ndicate that meanings visitors assign to recreation areas tend to be partiall y developed for their recreation behavior prior to their visit to recreation areas (Echtner & R itchie, 1993; Gartner & Hunt, 1987; Hunt, 1975; Pearce, 1982; Young, 1995; Walmsley & Young, 1998, 1999a) or through direct experience in recreation areas (Tuan, 1974, 1977). The study conducted here hypothesi zed that visitors w ith higher levels of recreation motivation would be likely to perceive recreation areas to be significantly more or less pristine, and found that there wa s no significant relationship betw een them. This implies that place meanings tend to be developed through actu al recreation experien ce rather than through mass media before their visit. Th erefore, this study demonstrat es and supports Tuans position. Additional evidence of visitors development of place meanings through actual experience in recreation areas should be sought in future research. As mentioned earlier, the current study em ployed an overall dimension of recreation motivation to determine whether it affects plac e satisfaction, although it would not be necessary to use importance of recreation motivations in or der to measure satisfaction levels (Babakus & Boller, 1992; Carman, 1990). The an alysis found that recreation mo tivation had a significant and negative relation to place satisf action although the relationship was deemed to be weak. Visitors with higher levels of expected motivation in recreation areas tended to be less satisfied with overall forest recreation settings It does not seem clear why the causal relationship between the two was negative. As shown in structural model 1, recreation motivation did not have a significant indirect effect on pl ace satisfaction through recreation setting perceptions. In other words, since visitors perceivi ng recreation areas to be in le ss pristine conditions were less satisfied with forest recreation settings, visitors with higher levels of recreation motivation were

PAGE 135

135 also not satisfied with recreation settings. Howe ver, the indirect effect was positive and was not significant. Overall, place satisfaction was expl ained by a direct effect of overall recreation motivation. Future research should more closel y examine the relationship between recreation motivation and satisfaction. Place meanings significantly predicted place a ttachment. This finding supports the notion that a greater number of wilderness settings in a forest or park would foster visitors positive development of attachment to the area (Stedma n, 2003a). This study found a direct relationship between place meanings and place attachment and al so an indirect relationship between the two via the effect of place satisfacti on on place attachment. In other wo rds, visitors who considered recreation areas to have more pris tine settings or conditions, were directly and more attached to the recreation areas. In addition, visitors with positive perceptions were more satisfied with forest recreation settings and th eir levels of satisfaction tended to increase their attachment to recreation areas. As more positive setting per ceptions increased, posit ive place attachment increased. The mediating effect of setting perceptions on place attachment via place satisfaction was stronger than the direct effect of setti ng perceptions on place attachment. This finding implies that place satisfaction and place attachme nt are contingent upon how visitors consider recreation settings and it would be important for recreation and planning managers to be aware of this relationship to provide better quality recreation opportuniti es, which are discussed later in the management implications section. It is worth noting again that place satisfac tion predicted place attachment. The influence of place satisfaction on place attach ment seems to originate from assertions made by Mesch and Manor (1998) who stated that place satisfacti on is followed by place attachment. The authors found significant support for their position by exam ining whether local residents levels of

PAGE 136

136 satisfaction with environmental settings (e.g., no ise level, air pollution level) in a neighborhood and satisfaction with the open space in the neighborhood (e.g., parks, playgrounds, parking space) significantly foster their levels of place a ttachment. Consistent with their findings, this study also found that as visitors were satisfied with overall settings (physical, social, and managerial) in recreation areas, they developed more positive attachments to recreation areas. In this study, no direct relationship was found between place satisfaction and placeprotective behavior, which is consistent with previous findings re ported by Stedman (2003). Both studies found that while visitors held a positive satisfactory attitude toward recreation settings, they did not show a signi ficant intention to participate in place-protective behavior. This finding does not support the notion of the theory of reasoned action (Ajzen & Fishbein, 1980; Fishbein & Ajzen, 1975). However, this study found that there is a significant and positive mediating effect of place satisfa ction on place-protectiv e behavior via place attachment. In other words, visitors who were satisfied with recreation settings tended to have positive attachment to recreation areas and those attached to the recr eation areas expressed their positive intention of resisting potential future changes in recreation se ttings. That is, this mediating effect indicates that place satisfaction was signifi cantly predictive of place-protective behavior and supports the partial notion of attitudinal behavioral theory. Structural Model 2 The condition of recreation area physical elem en ts was measured as specific evaluative beliefs after Stedman (2000; 2002). The relationship between setting perceptions and place satisfaction was found to be significa nt. This direct effect is consis tent with the findings of other researchers (Floyd, 1997; Whisman & Hollenhorst, 1998). The additional relationship between setting pe rceptions of recreation areas and attainment of preferred recreation experiences was examined and was f ound to be significant. Visitors who

PAGE 137

137 perceived recreation settings to be more pristine reported their attainment of preferred recreation experiences. This finding suppor ts previous literature on vi sitor satisfaction (Floyd, 1997; Whisman & Hollenhorst, 1998). For instance, Whisman and Hollenhorst (1998) found that situational variables such as water flow leve ls were positively rela ted to achievement of recreation experiences. That is, physical conditions of recreati on settings affect visitors preferred recreation ex periences (Floyd, 1997). Also, visito rs who attained their desired recreation experiences were more satisfied with forest recreati on settings. This finding supports previous ROS framework literature adopting expectancy valence theory (Driver & Brown, 1978) and other researchers (Floyd, 1997; Whisman & Hollenhorst, 1998) fi ndings that attainment of recreational experiences is posi tively related to visitor satisf action. One additional important finding is that of a significant mediating effect of setting perceptions on place satisfaction via attainment of recreational experi ences. That is, visitors positive setting perceptions influenced attainment of recreational experiences, and as a result of the direct effect, attainment of recreational experiences were positively predictive of satisfaction with fore st recreation settings. Overall, direct linkages between setting per ception (specific evaluative beliefs) and place satisfaction and the indirect relationship between the two vi a attainment of recreation experiences were significantly related. These fi ndings indicate the great importance of setting perceptions (evaluative beliefs) because the degr ees to which visitors achieve their recreation experiences and are satisfied with recreati on settings depend upon their evaluative beliefs. By adopting the ROS framework and adopti ng expectancy valence theory (Driver & Brown, 1978), direct and indirect relationships between attainment of recreation experiences and place attachment was found to be significant. This finding indicates that respondents who attained their desired recreation experiences in recreation areas we re inclined to show a growing

PAGE 138

138 attachment to recreation areas because the areas were a good fit for th eir recreation outcomes, thereby increasing visitor satisf action. Attainment of recreation experiences contributed to place satisfaction, as discussed earlier In this sense, factors (e.g., setting perceptions) constraining attainment of recreation expe riences and place satisfaction w ould prevent visitors from developing attachment to recreat ion areas. This proposition suppor ts the direct and mediating effects of setting perceptions on place attachment, which was found in this study. The indirect effect was stronger than the direct effect. That is, attachment to r ecreation areas are more influenced by setting perception indirectly throug h attainment of preferred recreation experiences and place satisfaction. These overall findings de monstrate supporting evidence for expectancy valence theory within an ROS framework and the importance of recreation settings that may constrain attainment of recreation experience pur suits, satisfaction with recreation settings, and attachment to recreation areas from a managerial perspective discussed la ter in this section. No significant mediating relationship betw een causal endogenous variables and placeprotective behavior was found to be significant in model 2. In a ddition, visitors who attained their recreation experiences did not show their strong willingness to resist potential future changes in recreation settings. As long as they achieved their preferred recreation pursuits, they did not seem to be sensitive to such future chan ges. This finding is similar to the non-significant direct relationship between place satisfaction and place-protective behavior. However, those attached to recreation areas tended to favor wilderness-type settings more than developed settings in recreation areas. Structural Model 3 As shown in m odels 2 and 3, attainment of recreation experiences was replaced by gap scores of recreation experiences. The reason wa s to examine disconfirmation models applied by many recreation researchers (Absher, Howat, Crilley, & Milne, 1996; Backman & Veldkamp,

PAGE 139

139 1995; Burns et al., 2003; Hamilton, Crompton, & More, 1991; Holland, 1979; MacKay & Crompton, 1988, 1990). Most direct (e.g., setting pe rception and gap scores, gap scores and place satisfaction) and indirect relationships be tween gap score variable s and setting perception variables or other endogenous variables (e .g., place satisfaction, place attachment, placeprotective behavior) were found similar to the find ings in model 2 except for three differences. The first difference was a non-signi ficant direct relationship betw een gap scores of preferred recreation experiences and place attachment. The second difference was a non-significant indirect relationship between se tting perceptions and place satis faction through gap scores of preferred recreation experiences. The third was the mediating effect of place satisfaction and place-protective behavior tended to be acc ounted for through place a ttachment, which is consistent with the mediating effect reported in model 1. Overall, attainment of recreation experiences and gap score variable s predicted recreation satisfaction. This is consistent with the findings reported by Burns, et al. (2003). Additionally, both variables influenced place a ttachment via place satisfaction that served as a mediating variable, respectively. In testing expectancy valence theory, atta inment of preferred recreation experiences in model 2 seemed to result in a high er contribution to direct standardized regression coefficients in explaining and predicting attitudinal beliefs (e .g., place satisfaction, place attachment) than the performance-only variable in model 3. However, co mparing the overall fit values (e.g., RMSEA, CFI, CMIN/df.) that indicate whether or not the models fit the data in SEM models, the present study found th at both variables seem to c ontribute to the prediction of endogenous variables in model 2 a nd model 3 in a relatively equal manner. The fit indices values are a reasonably acceptable fit. This finding is slightly different from previous findings that

PAGE 140

140 reported performance-only variables predicted satisf action better than gap scores did, comparing R-squared values in regression m odels (e.g., Burns, et al., 2003). Spatial Diagnostic Tests Overall, the variab les common across separate se ts of the regression models indicate that visitors negative setting perceptions, positive atta chment to recreation areas, attainment of nature experience, and gap scores of nature experience tend to ha ve a significant positive relation with place-protective behavior, while nostalgic experiences were not positively related to placeprotective behavior. Those variables would be im portant for recreation resource managers that need to identify which set of fact ors might affect whether visitors would have future intentions to protect recreation areas if physical, social and managerial settings in recreation areas would be changed to more developed settings in the futu re. Those variables would be also important for recreation resource managers to improve manageme nt in recreation areas of the forest. These management implications are further discussed later in this section. In evaluating spatial diagnostic tests performe d in this study, Model 2 with attainment of recreation experiences and other variables explained place-protective behavior better than Model 3 with gap scores of recreation experiences and other variables. In other words, a regression model with only a performance variable provided slightly better predic tions of a dependent variable than with gap score variables. This finding seem to be consistent with that reported by Burns, et al. (2003), stating that overall satisfaction of visitors at recreation areas was predicted by performance-only variables better than by gap score variables in regr ession models. Overall, Morans I test examining spatial autocorrelation in regression residuals demonstrated that specification of variables in each of those three models was appropriate for estimating parameters that explain place-protective behavi or, thereby supporting the findings discussed in this study.

PAGE 141

141 Delimitations Data were collected only from recreationist s who rode off-highway vehicles at major OHV designated recreation areas an d from visitors at major recreation areas with swimming sites, camping sites, and canoein g/kayaking sites available within the Ocala NF. Only visitors who were 18 years old or older were included as study participants. Only visitors who responded to both the on-site and follow-up mail survey were included in the analysis leading to the conclusions previously mentioned. Limitations The selected study areas were designated recr eation areas within Ocala N F, so results of the present study can be only generalized to Oc ala NF and other recr eation areas which are similar to those within Ocala NF in terms of ROS recreation settings. The recreation areas in Ocala NF where data collection was implemented we re considered to be classified as roaded natural to primitive settings. Similarly, all onsite data were collected between March 1, 2008 and May 31, 2008. Any OHV riders who did not recreat e at the designated Ocala NF areas and visitors at major recreation areas with sw imming areas, camping areas and canoeing/kayaking areas during this period were not represented. Therefore, the resu lts of the study are generalized to spring season visitors at the selected study sites. The completed and returned questionnaires (n = 934) yielded a 47% response rate. There are different opinions about the need for a non -response bias check. For instance, acceptable response rates for mail-back surveys can range from 30 to 70% (Goyder, 1985). According to Babbie (1990), non-response bias ch ecks should be utilized if the response rate is under 50%, while all studies should consider it (Linde r, Murphy, & Briers, 2001). Despite differing suggestions, this study performed a bias chec k between mail-back survey respondents and nonrespondents. Respondents tended to spend more days at Ocala NF during the time of the

PAGE 142

142 interview and have smaller group sizes. They were also likely to be older, have higher income levels and have higher education levels th an non-respondents. Overall, the results are representative of those who res ponded to both onsite and mail-back surveys and less so of those who did not respond to both surveys. Management Implications Descriptive Behavioral Constr ucts among Different Users Visitors had diverse recreational m otivations to visit Ocala NF and sought to achieve their desired recreation experien ces. It is important for recrea tion managers to be aware of whether different types of vis itors attained the desired recr eation experiences they sought, judging from gap scores of recreation experien ces. For instance, visitors at major swimming areas did not tend to achieve outcomes to both en joy nature (e.g., view the scenery, enjoy the smells and sounds of nature) and escape physic al pressure (e.g., experience tranquility and solitude). Kayakers and canoers were not likely to achieve recreati onal experiences of tranquility and solitude. On other hand, OHV riders tended to achieve their desired recreation experiences. Managers should make efforts to provide t hose unattained recreation experiences to visitors. Otherwise, visitors are more likely to report being dissatisfi ed with recreational experiences and lose attachment to recreation areas. For instance, after not attaining expected tranquility and solitude, wilderness canoers and ka yakers were likely to report lower levels of satisfaction with social settings (e.g., solitude, the number of people) and managerial settings (e.g., rules and regulations), which deterred de veloping attachments to those recreation areas. This also applies to management of swim ming areas. That is, managers should focus on providing those kinds of quality recreation experiences to visi tors, especially swimmer and wilderness canoe trail users, in order to improve visitors satisfaction with social and managerial settings and foster attachment to recreation areas. Additionally, although OHV riders tended to

PAGE 143

143 attain recreational experiences, th ey were likely to be least satis fied with physical, social, and managerial aspects, which negati vely affected attachment to recreation areas. Managers should consider providing quality pristine settings, social settings and preferred managerial aspects to OHV riders. For instance, managers ought to prov ide more natural recr eation settings to OHV riders. Also, managers may need to consider adju sting current levels of recreation fee costs, and existing level of rules and regula tions that are applied to OHV ride rs. Managers need to control OHV riders recreational activities (e.g., speed limit for safety) through more formal information or education, thereby increasing appreciative behavior. These kinds of overall management actions would foster visitors attachment to recreation areas. Respondents Overall Rating of Each Construct Managers should be aware of what types of recreational e xperiences visitors prim arily seek at recreation areas and do what they can to assure that those desired experiences are met. For instance, the recreati onal experiences that a majority of vi sitors considered to be important include enjoyment of nature, as well as escaping social and physical pressures. However, they did not tend to attain recreational experiences of tranquility and solitude. Managers may need to control factors that deter visitors experiences of tranquility and solitu de, for instance, the number of users in recreation areas, noise from motorized or non-motorized users at recreation areas, and other depreciative behavi ors. Otherwise, visitors satis faction with social recreation settings and attachment to recreation areas w ould most likely decrease. Additionally, managers should improve visitors opportunities to achieve satisfaction with managerial settings by adjusting and controlling recreation fee costs, le vel of rules and regulations, and facilities and services if adjustment of those factors can co mply with management objectives. Such actions would contribute to visitors a ttachment to recreation areas.

PAGE 144

144 Structural Models Recreation resource m anagers should be awar e of which factors affect place meanings, place satisfaction and place attachment if resource ma nagers are interested in integrating sense of place constructs into recreation management objectives and practices to facilitate quality recreation opportunities as well as protection of natura l resources. Results showed the importance of both recreation experiences and re creation setting perceptions. More specifically, the importance of recreation experiences tends to have a significant and direct impact on place satisfaction and place attachment. Forest recreationists who vis it recreation areas with higher expectations of recreatio n experiences, tend to ex press higher satisfaction with recreation settings as well as higher attachment to recreation areas. Additionally, place satisfaction and place attachment are directly dependent upon visitor perceptions of natural re creation settings that are provided by forest managers. As abundant literature on visitor satisfaction has indicated, it is important for recreation managers to maintain and improve visitor satisfaction as a major goal of recreation pr oviders. Further results from model 1 indicate that setting perceptions had a si gnificant mediating effect on place attachment via place satisfaction. That is, this finding reveals the importa nce of setting perceptions and place satisfaction that lead to how attached to recrea tion settings visitors are. Overall, recreation resource managers need to be aware of visitors preferred motivations as well as how visitors perceive recreation settings. That is, it is important for recreation resource managers to provide natural recreation settings and monitor any futu re changes in recreati on motivation. If forest managers fail to provide quality recreation areas in more pristine conditions, visitors would be less likely to be satisfied a nd attached to recreation areas. Further implications include th e importance of both setting perceptions and attainment of desired recreation experiences if managers integrated sense of pl ace concepts into recreation and

PAGE 145

145 planning management. As discus sed earlier, setting perceptions significantly affect place satisfaction and place attachment. Findings showed that setting perceptions had a significant direct influence on attainment of desired recreation e xperiences, thereby in directly affecting place satisfaction and place attachment. Visitors perceiving recreation settings to be less pristine did not tend to achieve their desi red recreation pursuits. This failu re additionally affected place satisfaction and place attachment negatively. Overall, these findings suggest that recreation resource managers should provide and maintain more pristine recreati on settings that would foster attainment of visitors desired recreation experiences, satisfaction and attachment to recreation areas. In this study, willingness to resist potentia l future changes in recreation settings was directly dependent upon place attachment wh ile it was not directly predicted by place satisfaction. Those who are attach ed to recreation areas tend to have their recreation areas in more natural settings and resist more develope d recreation settings. However, place satisfaction had a mediating effect on intention to engage in place-protective behavior through place attachment. This indicates the im portance of place satisfaction a nd place attachment in predicting place-protective behavior. Overall, managers sh ould keep recreation areas in more pristine settings because natural settings are part of what visitors repo rted satisfaction with, and those attached to recreation ar eas are more inclined to resist pote ntial development related changes in recreation settings. Additionally, managers shou ld keep in mind which factors influence attainment of desired recreation experiences, place satisfaction and place attachment, as recommended previously for future management. Spatial Diagnostic Tests As discussed and recom mended previously, se tting perceptions, attainment of desired recreation experiences in nature, and attachment to recreation areas tend to have a strong positive

PAGE 146

146 influence on place-protective beha vior. Based on the results, it would be important for recreation resource managers to focus on providing natura l recreation settings. Fo r instance, respondents who perceived recreation settings to be less prim itive expressed their strong willingness to resist future potential changes in recreation settings towards less pristine environments. Similarly, those who achieved their desired recreation experience in enjoying nature and were attached to recreation areas tended to show the same inclinations. Future Research Structural Models The presen t study used multiple sets of constr ucts such as recreational motivation (e.g., importance of recreation experiences attainment of desired recrea tion experiences, gap scores of recreation experiences), sense of place constructs (e.g., place meanings, place satisfaction, place attachment), and place-protective behavior to investigate the cau sal relationships between them by employing structural sets of multiple-equation regression models, instead of multiple sets of ordinary regression models. In this study, se nse of place concepts (Stedman, 2002, 2003a) were the primary constructs used, recr eation motivation was integrated into sense of place constructs, and place-protective behavior was employed as indicators of future intended behaviors. The relationships between multiple sets of the constructs were examined, based on expectancy valence models and theory of reasoned acti on. Future research s hould investigate the relationships between those constructs, which would contribute to recreation and planning management as well as exte nd the results reported here. Future research should center on identifying causal relationships be tween two or three major constructs employed in this study by using first-order factors of each construct. For instance, it is important to investigate the relationship between domains for attainment of recreation experiences and domains for place attachment by adopting expectancy valence

PAGE 147

147 models. Kyle, Mowen and Tarrant (2004) examin ed the relationship between importance of recreation experiences and place attachment by comparing su b-dimensions of recreation motivation with those of place attachment, wh ich was based on Lawlers (1973) expectancy value model of motivation. Invest igation of the first-order fact or relationships would help recreation resource managers spec ifically determine which domains within each construct are considered to be more important to satiate visitors attainment of recr eation experiences and to improve forest recreation settings. Reliability and validity scores of all the scale variables were examined in this study. Each latent construct used in this study met criteria of reliability, ex cept for a few latent factors (e.g., achievement, enjoyment of nature, equipment use) with Cronbachs alpha of less than .6. Those latent factors were used to measure gap scores of recreation experien ce preferences and were included in model 3 because they held acceptable le vels of construct reli ability and AVE and the same latent factors measured in a different way were used for model 3 to examine the differences in structural fit between the two models. Future st udies should make an effort to apply them in different settings or use other si milar additional items representing the same latent factors in a survey instrument to further examine the re liability of the latent scale variables. A sense of place construct that consisted of place meaning, place satisfaction and place attachment was adapted from Stedman (2002). Specific place meanings used as specific evaluative beliefs about recreati on settings were composed of tw o extracted factors, pristine meaning and impacted meaning. Items for each f actor were included for confirmatory factor analysis. Items of impacted meanings did not pr oduce reasonable factor loadings and were not included in the measurement and structural models The probable reason may be related to the fact that semantic differential scales (e.g., extr emely crowded not at all crowded) used by

PAGE 148

148 Stedman (2002) were used to measure specific place meanings. Respondents might have been confused with that scale format. Therefore, fu ture studies may wish to test for a possible instrumentation effect by utilizing a more tr aditional Likert scale (e.g., ranging from strongly disagree to strongly agree) that was successfully used to measur e other constructs (e.g., place attachment) in this study. For future studies, it would also be better to include more place meaning items that represent ROS settings (e.g., physical, social, managerial). That is, perceptions of ROS settings woul d be evaluated as specific belie fs about recreation areas. For instance, this present study a dded items representing managerial settings to items used by Stedman (2002) to measure place satisfaction constructs that consisted of physical, social, and managerial setting attributes. It is believed that this type of an ongoing effo rt would contribute to further research into re source recreation planning. Spatial Diagnostic Tests Data used for spatial diagnostic analyses in this study were collected from visitors at Ocala NF by conducting onsite surveys from Ma rch to May, 2008 and follow-up mail surveys from March to July, 2008. The spatial analysis of place protective behavior was implemented by using Florida county-wide boundaries. In general, the results show th at there is a small sample of visitors in some counties living far away from Ocala NF while there is a large sample of visitors in the more proximate counties close to Ocala NF. There were no visitor responses from most counties in the Florida Panhandle. It is quite likely that very few if any FL residents from Panhandle go to Ocala NF, since they have the option of visiting a fairly comparable site at the Apalachicola NF, so there may be nothing wrong with onsite and mail-back survey methodologies finding few Panhandle visitors. Howeve r, future research may need to consider implementing a longer period of onsite surveys to obtain a larger sample of visitors living farther

PAGE 149

149 from Ocala NF. Another alternativ e way to obtain a larger sample of visitors from each county would be to conduct a telephone survey of resi dents who have previous ly visited Ocala NF. Using a stratified random sampling method, a telephone survey can make proportionate contacts with residents living in each county of Florida. Adjusted onsite and mail-back questions can be used for a telephone survey. Similarly, resident s who have visited other national forests in Florida, Apalachicola NF and Osceola NF can be also included in a teleph one survey to examine differences in sense of place perceptions among re spondents who have visited at least one of the national forests. The sense of place concept that consists of place meanings, place satisfaction, place attachment can be mapped to detect spatial patt erns of visitors beliefs about and attitudes towards recreation areas of Ocala NF. For instance, mapping how strongly visitors are dependent upon major recreation areas of Ocala NF for their recreation activities and experiences by county would help recreation resource managers determin e the spatial value of forest recreation areas. Other additional research could integrate sense of place constructs with socio-demographics, distance from each county to Ocala NF, residence types (e.g., small town or major city), and conventional intervening opportuniti es (e.g., the number of recreation resources available within each county) (Fik, 1988; Fik & Mulligan, 1990; K il, et al., 2007; Wilson & Bennett, 1985) using spatial analysis to achieve a better un derstanding of sense of place perceptions.

PAGE 150

150 APPENDIX SURVEY INSTRUMENT On-Site Survey Questionnaire Interviewer: __________ Date: ___________ Site Location: __________ Interview Time: __________ Ocala National Forest Onsite Visitor Survey Spring 2008 Section A: Use Profile 1. Is this your first trip to this forest? No [ Go to Question 1a] Yes [ Go to Question 2] [If no] 1a. In what year did you make your first trip to this area? ____________year 1b. In a typical year, about how many days do you spend recreating at this area? ___________days 1c. In a typical year, about how many days do you spend recreating at other areas besides this area? _________days 2. On this trip did you recreate only at Ocala National Forest or did you go to other National Forests, parks, or recreation areas? Only Ocala National Forest Also other places Is Ocala NF your primary destination for this trip? Yes No What other areas besides Ocala NF have you or will you visit during this trip? _______________________________________________________ _______________________________________________________ 3. In total, about how many days (or hours) long will this trip at this area be? _______days _______hours (If day trip) 3a. During this trip, where ar e you staying? (Check one) Camping in this forest Staying in motel/hotel Staying at home Camping in another area outside this forest Staying with friends/relatives Other (Specify______) 4. About how many miles did you drive from your home to visit this area? _______miles 5. Which of the following best describes the composition of your group? (Check one) Alone Family Friends Family & Friends Other (Specify) __________ 6. How many people are in your group today? ________ 6a. How many ____ # of children under 18 ____ # of adults 18 & older 7. Are you a member of a recreation (i.e., Bicycle Club) or conservation-related (i.e., Sierra Club) organization? (Check all that apply) Recreation Conservation Neither

PAGE 151

151Section B: Your Recreation Activities 1. What kinds of activities have you and/or your group done or plan to do during this visit? (Check ALL that apply) Day Hiking/Walking Camping Swimming Viewing Scenery Picnicking Hunting Fishing Photography Nature Study Mountain Biking Jogging/Trail Running Horseback Riding Canoeing/Kayaking Driving Off-Road Vehicles Other (Specify) ________________ 2. Which of the activities above that you do at Ocala National Forest is MOST IMPORTANT to you? (Specify one) _____________ 2a. How many times a year do you participate in the most important activity in the area? __________times a year Section C: Importance of Recreation Experiences 1. Different people may expect different recreation experiences when they come to visit this area. Please look at the list of recreation experiences below and tell us how importa nt each experience is to you when you visit this area (Circle ONE number per line.) Recreation Experiences Not at all Important Somewhat Important Moderately Important Very Important Extremely Important Enjoy the smells and sounds of nature 1 2 3 4 5 Test my endurance 1 2 3 4 5 Bring back pleasant memories 1 2 3 4 5 Get exercise 1 2 3 4 5 Get to know the lay of the land 1 2 3 4 5 Learn more about nature 1 2 3 4 5 Experience solitude 1 2 3 4 5 Use my recreation equipment 1 2 3 4 5 Think about good times I've had in the past 1 2 3 4 5 Show others I can do it 1 2 3 4 5 Release or reduce built-up tensions 1 2 3 4 5 Get away from the usual demands of life 1 2 3 4 5 Develop my skills and abilities 1 2 3 4 5 Experience tranquility 1 2 3 4 5 Keep physically fit 1 2 3 4 5 Talk to others about our/my recreation equipment 1 2 3 4 5 View the scenery 1 2 3 4 5 Develop personal spiritual values 1 2 3 4 5 Be with friends and family 1 2 3 4 5 Experience open spaces 1 2 3 4 5 Relax physically 1 2 3 4 5 Chance dangerous situations 1 2 3 4 5 Be with members of my group 1 2 3 4 5 Take risks 1 2 3 4 5

PAGE 152

152Section D: Your Opinions of Current and Po tential Off-Highway Ve hicle (OHV) Management 1. Are you an OHV rider at this area? Yes No [ If yes, go to Question 2; if no, go to Section E ] 2. Are you a member of an organized OHV club? No Yes Which one?_______________________ 3. Which riding fee option did (will) you choose to pay today? (Check all that apply) 1-3 Day Pass $10 per Operator Yearly Pass $75 per Operator Other (Specify____________) 7-Day Pass $25 per Operator None 4. Please indicate your level of acceptance for each of the following management conditions at this OHV area (Circle ONE per line) Current Management Conditions Highly Unacceptable Neutral Highly Acceptable The specific fee you paid to ride (not camping fee) today. 1 2 3 4 5 6 7 Current number of parking spots for trailers at staging area. 1 2 3 4 5 6 7 Current number of parking spots for vehicles at staging area. 1 2 3 4 5 6 7 5. Please indicate the extent to which you would either support or oppose these potential management actions (Circle ONE per line.) Potential Management Action Strongly Oppose Neutral Strongly Support Removing some of the current trail markers. 1 2 3 4 5 6 7 Adding more trail markers. 1 2 3 4 5 6 7 Providing drinking water along trails (not at the trailhead). 1 2 3 4 5 6 7 Providing bathrooms along trail (not at the trailhead). 1 2 3 4 5 6 7 Providing picnic areas along trail (not at the trailhead). 1 2 3 4 5 6 7 Providing designated regrouping points along trails. 1 2 3 4 5 6 7 Providing designated children riding areas. 1 2 3 4 5 6 7 Providing destination trails. 1 2 3 4 5 6 7 6. Would you be interested in volunteer opportunities if volunteers were needed for trail maintenance and patrol in order to keep riding areas open? Yes No Section E: Demographic Profile 1. What is your gender? Male Female 2. What year were you born? 19________year born 3. Are you? Single Married/partnered Divorced/separated Widowed 4. How would you describe yourself? (Check one) Caucasian (White) African American (Black) Latino (Hispanic) Asian/Pacific Islander Other (specify_________________)

PAGE 153

1535. What is the highest educational level you have attained? (Check one) Less than high school Completed 4 year college degree High school diploma Some graduate work Attended business/technical school Completed graduate or advanced degree Some college or 2 year degree 6. Which of the following categories best describe your annual household income in 2007, before taxes? (Check one) Less than $14,999 $50,000 to $64,999 $150,000 to $199,999 $15,000 to $34,999 $65,000 to $99,999 Over $200,000 $35,000 to $49,999 $100,000 to $149,999 7. What is your permanent residence Zip Code? _____________ 7a. How long have you lived at this location? ________years ________months 7b. What is your permanent residence COUNTY? _________________ 8. In which of the following places do you currently live? (Check only one) A farm or ranch Small city (5,000-50,000 population) Rural or small town (Under 1,000 population) Medium city (50,000-1 million population) Town (1,000-5,000 population) In a major city or metropolitan area (Over one million) 9. In which of the following places di d you spend the most time while growing up (to age 18)? (Check only one) On a farm or ranch Small city (5,000-50,000 population) Rural or small town (Under 1,000 population) Medium city (50,000-1 million population) Town (1,000-5,000 population) In a major city or metropolitan area (Over one million)

PAGE 154

154Do you have any suggestions or comments? Thank you for completing this survey. We would also like to ask you some additional questions after you have returned home from your visit. Would you be willing to give us your name and address so we can mail you a followup survey to complete at your convenience? Again this information will be kept conf idential. Your name will not appear in any publication or be linked with the results you provide. NAME: ______________________________________ ADDRESS: _____________________________________________________________________________________________ CITY: ____________________________ STATE: __________________ ZIP CODE: ______________________________ E-MAIL ADDRESS: ___________________________

PAGE 155

155 Mail-Back Survey Questionnaire Ocala National Forest Visitor Study Mail-back Survey Spring 2008

PAGE 156

156Ocala National Forest Visitor Study: Mail-back Survey Spring 2008 We would like to start by asking questions about your recreation experiences riding an off highway vehicle ( For the purpose of the survey, Off-Highway Vehicle (OHV) stands for both All Terrain Vehicle (ATV) and OffHighway Motorcycle. ) Section A: Use Profile of OHV Riders 1. Are you an OHV rider? Yes No [ If no, go to Section B ] [ If yes ] 1a. How many years have you ridden an OHV for recreational purposes? _____number of years 1b. How many years have you ri dden at the Ocala National Fo rest? _______number of years 1c. Did you travel to any OHV recreation areas ot her than the Ocala NF within the last year? Yes No [If yes] 1d. Approximately how many trips did you take? ________total # of trips taken 1e. Please indicate three other areas you went to most often. 1) ___________________ 2) ____________________3) ______________________ 2. Please indicate the type of off-highway vehicle you were riding during your visit to Ocala NF when we contacted you. ATV only 4 WD only ATV/4WD Motorcycle/combo 3. How do you rate your skill level in OHV riding? Beginner Novice Intermediate Advanced Expert Section A: Importance of Site Attributes 1. Different people may expect different things when they come to visit this area. Please tell us how important each of the following site attributes are to you when you visit the area where we contacted you (Please circle ONE number per line.) Site Attributes Not at all Important Somewhat Important Moderately Important Very Important Extremely Important Easy access to the area 1 2 3 4 5 Minimum impacts on plants and animals 1 2 3 4 5 Encountering other forest visitors 1 2 3 4 5 Enforcement of rules and regulations 1 2 3 4 5 Seeing many different types of recreation activities 1 2 3 4 5 Having no developed facilities available 1 2 3 4 5 Natural features (wildlife, trees, clean water, etc.) 1 2 3 4 5 Facilities for comfort and conveniences 1 2 3 4 5 Lack of presence of other visitors 1 2 3 4 5 Natural area with few signs of human development 1 2 3 4 5

PAGE 157

157Section B: Attainment of Recreation Experiences 1. Below is a list of possible experiences that you may ha ve had during your recent visit at Ocala National Forest. Please look over the list of possible expe riences and indicate how much you were able to attain each experience in the area where we contacted you (Please circle ONE number per line.) Recreation Experiences Did Not Attain Somewhat Attained Moderately Attained Mostly Attained Totally Attained Enjoy the smells and sounds of nature 1 2 3 4 5 Test my endurance 1 2 3 4 5 Bring back pleasant memories 1 2 3 4 5 Get exercise 1 2 3 4 5 Get to know the lay of the land 1 2 3 4 5 Learn more about nature 1 2 3 4 5 Experience solitude 1 2 3 4 5 Use my recreation equipment 1 2 3 4 5 Think about good times I've had in the past 1 2 3 4 5 Show others I can do it 1 2 3 4 5 Release or reduce built-up tensions 1 2 3 4 5 Get away from the usual demands of life 1 2 3 4 5 Develop my skills and abilities 1 2 3 4 5 Experience tranquility 1 2 3 4 5 Keep physically fit 1 2 3 4 5 Talk to others about our/my recreation equipment 1 2 3 4 5 View the scenery 1 2 3 4 5 Develop personal spiritual values 1 2 3 4 5 Be with friends and family 1 2 3 4 5 Experience open spaces 1 2 3 4 5 Relax physically 1 2 3 4 5 Chance dangerous situations 1 2 3 4 5 Be with members of my group 1 2 3 4 5 Take risks 1 2 3 4 5 Section C: Importance of Site Attributes 1. Different people may expect different things when they come to visit this area. Please tell us how important each of the following site attributes are to you when you visit the area where we contacted you (Please circle ONE number per line.) Site Attributes Not at all Important Somewhat Important Moderately Important Very Important Extremely Important Easy access to the area 1 2 3 4 5 Minimum impacts on plants and animals 1 2 3 4 5 Encountering other forest visitors 1 2 3 4 5 Enforcement of rules and regulations 1 2 3 4 5 Seeing many different types of recreation activities 1 2 3 4 5 Having no developed facilities available 1 2 3 4 5 Natural features (wildlife, trees, clean water, etc.) 1 2 3 4 5 Facilities for comfort and conveniences 1 2 3 4 5 Lack of presence of other visitors 1 2 3 4 5 Natural area with few signs of human development 1 2 3 4 5

PAGE 158

158Section D: Your Attitudes Towards the Recreation Area Where We Contacted You 1. What kind of place is the recreation area where we contacted you? (Please circle ONE number per line.) This area is Strongly Disagree Disagree Neutral Agree Strongly Agree A family place 1 2 3 4 5 A special place for me 1 2 3 4 5 A pristine wilderness 1 2 3 4 5 A place mostly for vacationers 1 2 3 4 5 A place of high environmental quality 1 2 3 4 5 A community of fellow recreators 1 2 3 4 5 A place to escape from civilization 1 2 3 4 5 A place for local Floridians 1 2 3 4 5 The real original Florida 1 2 3 4 5 2. Wed like to learn more about your opinions about the recreation area where we contacted you ( Please read both left and right sides of the table and circle ONE number per line.) Neutral This area is extremely scenic 1 2 3 4 5 This area is not at all scenic This area has many developed recreation facilities and services 1 2 3 4 5 This area has few developed recreation facilities and services This area is extremely forested 1 2 3 4 5 This area is not at all forested This area has too many rules and regulations 1 2 3 4 5 This area has very few rules and regulations Recreation fee on this area is extremely expensive 1 2 3 4 5 Recreation fee on this area is not at all expensive This area is extremely peaceful 1 2 3 4 5 This area is not at all peaceful This area is extremely crowded 1 2 3 4 5 This area is not at all crowded The water in this area is extremely clear 1 2 3 4 5 The water in this area is not at all clear This area has changed a lot over the years 1 2 3 4 5 This area has not changed at all over the years This area provides a lot of historical and cultural information 1 2 3 4 5 This area provides no historical and cultural information This area is extremely likely to have more development in the future 1 2 3 4 5 This area is not at all likely to have more development in the future This area has many species of wildlife and plants 1 2 3 4 5 This area has few species of wildlife and plants This area has many recreationists using it 1 2 3 4 5 This area has few recreationists using it This area has a lot of public access 1 2 3 4 5 This area has very little public access This area is extremely developed 1 2 3 4 5 This area is not at all developed The water in this area is extremely polluted 1 2 3 4 5 The water in this area is not at all polluted 3. We would like to know how satisfied you were with your overall experience in the area where we contacted you at in Ocala National Forest. On a scal e of 1-10, with 10 being most satisfied, how satisfied were you with this trip?_________

PAGE 159

1594. For the following items, please respond with the specific recreation area we contacted you at in mind. Overall, how important is the recreation area to you? (Please circle ONE number per line.) Strongly Disagree Neutral Strongly Agree I feel relaxed when I'm at this area. 1 2 3 4 5 6 7 For doing the things that I enjoy most, no other place can compare to this area. 1 2 3 4 5 6 7 Everything about this area is a reflection of me. 1 2 3 4 5 6 7 Many of my friends/family prefer this area over other site. 1 2 3 4 5 6 7 If I were to stop visiting this area, I would lose contact with a number of friends. 1 2 3 4 5 6 7 This area reflects the type of person I am. 1 2 3 4 5 6 7 This area is the best place for doing the things that I enjoy most. 1 2 3 4 5 6 7 This area is not a good place to do the things I most like to do. 1 2 3 4 5 6 7 I really miss this area when I'm away from it for too long. 1 2 3 4 5 6 7 My friends/family would be disappointed if I were to start visiting other settings and facilities. 1 2 3 4 5 6 7 I feel that I can really be myself at this area. 1 2 3 4 5 6 7 I feel happiest when I'm at this area. 1 2 3 4 5 6 7 As far as I am concerned, there are better places to be than at this area. 1 2 3 4 5 6 7 This area says very little about who I am. 1 2 3 4 5 6 7 This area is my favorite place to be. 1 2 3 4 5 6 7 5. How satisfied are you with different aspects of the r ecreation area where we contacted you: (Circle ONE number per line.) Extremely Dissatisfied Somewhat Dissatisfied Neutral Somewhat Satisfied Extremely Satisfied Its solitude and peacefulness 1 2 3 4 5 Its scenery 1 2 3 4 5 Recreation fees cost 1 2 3 4 5 Populations of wildlife around it 1 2 3 4 5 Its level of rules and regulations 1 2 3 4 5 Its water quality 1 2 3 4 5 Its level of land development 1 2 3 4 5 Its facilities and services 1 2 3 4 5 The number of people using it 1 2 3 4 5 Other peoples recreation activities 1 2 3 4 5 6. How would you rate the recreation ar ea where we contacted you OVERALL? 1 (poor) 2 3 4 5 6 7 8 9 10 (perfect) 7. Considering everything, how do you feel about the c ondition of the recreation area where we contacted you? Strongly Dislike Dislike Neutral Like Strongly Like 8. Overall, how do you feel about changes that have happened to th e recreation area where we contacted you? Extremely Unhappy Somewhat Unhappy Neutral Somewhat Happy Very Happy

PAGE 160

1609. How likely is it that you will visit Ocala NF again based on your overall recreation experience? Very Unlikely UnlikelyNeutral Likely Very Likely 10. How many times do you think you may recreate at Ocala National Forest in the next 12 months? ___times a year Section E: Your Reactions to Possible Changes in th e Recreation Area Where We Contacted You 1. If the rules on the kinds of recreation that can take place here were to become more restrictive, I would: (Circle One per line.) Extremely Unlikely Uncertain Extremely Likely Spend less time here 1 2 3 4 5 6 7 Vote for or support rules that might keep it from becoming more restrictive 1 2 3 4 5 6 7 Join or help form a group to fight such restrictions 1 2 3 4 5 6 7 2. If the natural environmental quality of the area I recreated in showed major deterioration, I would: (Circle One per line.) Extremely Unlikely Uncertain Extremely Likely Spend less time here 1 2 3 4 5 6 7 Vote for or support rules that might keep it from environmentally deteriorating 1 2 3 4 5 6 7 Join or help form a group to fight for more environmental protection of the area 1 2 3 4 5 6 7 3. If the number of Off-Highway Vehi cle (OHV)s around here increased gr eatly, I would: (Circle One per line.) Extremely Unlikely Uncertain Extremely Likely Spend less time here 1 2 3 4 5 6 7 Vote for regulations that wo uld limit the number of OHVs here 1 2 3 4 5 6 7 Join or help form a group to reduce the number of OHVs 1 2 3 4 5 6 7 4. If the number of recreator s around here increased greatly, I would: (Circle One per line. ) Extremely Unlikely Uncertain Extremely Likely Spend less time here 1 2 3 4 5 6 7 Vote for regulations that would limit the number of recreators here 1 2 3 4 5 6 7 Join or help form a group to reduce the number of recreators 1 2 3 4 5 6 7

PAGE 161

161Do you have any suggestions or comments? Thank you for your participating in this survey! Please place this completed questionnaire in the po stage-paid business return envelope provided. If you have any questions about this survey, please contact: Namyun Kil, Principal Investigator Dept. of Tourism, Recrea tion & Sport Management, University of Florida, Gainesville, FL 32611 Phone: (352)392-4042 Email: ecopark@ufl.edu

PAGE 162

162 LIST OF REFERENCES Absher, J. D. (1998). Custom er service measures for national forest recreation. Journal of Park and Recreation Administration, 16, 31. Absher, J. D., Howat, G., Crilley, G., & Milne, I. (1996). Toward customer service: Market segment differences for sports and leisure centres. Australian Leisure, 7 (1), 25. Ajzen, I. (1991a). Benefits of leisure: A social psychological perspective. In B. L. Driver, P. J. Brown, & G. H. Peterson (Eds.), Benefits of Leisure (pp. 411-418). State College, PA: Venture. Ajzen, I. (1991b). The theory of planned behavior. Organizational Behavior and Human Decision Processes 50, 179-211. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs. NJ: Prentice Hall. Altman, I., & Low, S. M. (1992). Place attachment New York: Plenum Press. Babakus, E., & Boller, G. W. (1992). An empi rical assessment of the SERVQUAL scale. Journal of Business Research, 24 253. Babbie, E. (1990). Survey research method (2nd ed.). Belmont, CA: Wadsworth Publishing Company. Babbie, E. (2004). The practice of social research (10th ed.). Belmont, CA: Wadsworth. Backman, S. J., & Veldkamp, C. (1995). Examina tion of the relationship between service quality and user loyalty. Journal of Park and R ecreation Administration, 13 29. Beard, J. B., & Ragheb, M. G. (1983). Measuring leisure motivation, Journal of Leisure Research, 3, 219-228. Bell, F. W., & Leeworthy, V. R. (1990). Recrea tional demand by tourists for saltwater beach days. Journal of Environmental Economics and Management, 18, 189-205. Bentler, P. M. (1992). On the fit of models to covariance and methodol ogy to the bulletin, Psychological Bulletin, 112 400. Bollen, K. A. (1989). Structural equations with latent variables New York: Wiley. Brandenburg, A. M., & Carroll, M. S. (1995). Your place or mine?: The e ffect of place creation on environmental values and landscape meanings. Society and Natural Resources 8, 381398.

PAGE 163

163 Bricker, K, & Kerstetter, D. (2002). An interp retation of special place meanings whitewater recreationists attach to the Sout h Fork of the American River. Tourism Geographies 4, 396-425. Bricker, K. S. (1998). Place and preference: A study of wh itewater recreationists on the South Fork of the American River Ph.D. Dissertation, The Pennsyl vania State Univ., University Park. Brown, G., Reed, P., & Harris, C. C. (2002). Te sting a place-based theory for environmental evaluation: An Alaska case study. Applied Geography, 22 49-76. Browne, M. W. & Cudeck, R. (1993). Alternative ways of assessing model fit. In: Bollen, K. A. & J. S. Long (Eds.), Testing structural equation models (pp.136-162). Sage, Newbury Park, CA. Bruns, D., Driver, B. L., Lee, M. E., Anderson, D. H., & Brown, P. J. (1994). Pilot test for implementing benefits-based management. Paper presented at the Fifth International Symposium on Society and Resource Management (pp. 243-256). Fort Collins: Color Burns, R. C., Graefe, A. R., & Absher, J. D. (2003). Alternate meas urement approaches to recreational customer satisfaction: Sa tisfaction-only versus gap scores, Leisure Sciences, 25, 363-380. Buttimer, A., & Seamon, D. (1980). The human experience of space and place. London: Croom Helm. Carman, J. (1990). Consumer perceptions of se rvice quality: An assessm ent of the SERVQUAL dimensions. Journal of Retailing, 66 33. Chawla, L. (1992). Childhood place attachments. In I. Altman, & S. M. Low (Eds.), Place Attachment (pp. 63-86). NY: Plenum Press. Cheng, T., Kruger, L., & Daniels, S. (2000). Place as a mediating soci al variable in natural resource politics: A propositional analysis. Unpublished manuscript. Churchill, G. A., & Suprenant, C. (1982). An i nvestigation into the dete rminants of customer satisfaction. Journal of Marketing Research, 16 491. Clark, R. N., & Stankey, G. H. (1979). The Recr eation Opportunity Spec trum: A Framework for planning, management and research. U.S. Department of Agriculture Forest Service, General Tech. Report PNW-98. Cliff, A. D., & Ord, J. K. (1973). Spatial autocorrelation Brondesbury, London: Pion Limited. Cole, D. N. (1996). Wilderness recreation us e trends, 1965 through 1994. (Gen. Tech. Rep. INTRP-488). Ogden, UT: USDA Forest Servic e, lntermountain Research Station.

PAGE 164

164 Connelly, N. A. (1987). Critical factors and th eir threshold for camper satisfaction at two campgrounds. Journal of Leisure Research, 19, 159. Crompton, J. L., & Love, L. L. (1995). The pred ictive validity of alternative approaches to evaluation quality of a festival. Journal of Travel Research, 34 (1), 11. Crompton, J. L., & MacKay, K. J. (1989). Users perceptions of the relative importance of service quality dimensions in sel ected public recreation programs. Leisure Sciences, 11 367. Davenport, M. A. (2005). Getting from sens e of place to place-based management: An interpretive investigation of place meanings and perceptions of landscape change. Society and Natural Resources, 18, 625-641. Dillman, D. A. (2000). Mail and Internet surveys: The tailored design method (2nd ed.). New York, NY: John Wiley & Sons, Inc. Dorfman, P. W. (1979). Measurement and meaning of recreation satisfaction: a case study in camping. Environment and Behavior, 11 483. Driver, B. L. & Tocher, S. R. (1970). Toward a behavioral interpre tation of recreational engagements, with implications for planning. Driver, B. L. (Ed.), Elements of outdoor recreation planning (pp. 9-13). University of Mich igan Press, Ann Arbor, Michigan Driver, B. L. (1996). Benefits-drive n management of natural areas. Natural Area Journal, 16 (2), 94-99. Driver, B. L., & Brown, P. J. (1978). The opportunity spectrum concept and behavioral information in outdoor recreation re source supply inventories: A rationale In: Integrated inventories of renewable natural resources USDA Forest Service General Technical Report. Driver, B. L., & Brown, P. J. (1983). Contributions of behavioral scientists to recreation resource management. In: Altman, I. & J. F. Wohlwill (Eds.). Behavior and the natural environment (pp. 307-339). New York: Plenum. Driver, B. L., Brown, P. J., & Peterson, G. L. (eds.) (1991). Benefits of Leisure. Venture Publishing, State College, Pennsylvania. Driver, B. L., Brown, P. J., Stankey, G. H., & Gregoire, T. H. (1987). The ROS planning system: Evolution, basic concepts, and research needed. Leisure Sciences, 9 201.

PAGE 165

165 Driver, B. L., Tinsley, H. E. A., & Manfredo, M. J. (1991). The paragraphs about leisure and recreation experience preference scales: Results from two inventories designed to assess the breadth of the perceived psychological benef its of leisure. In: B. L. Driver, P. Brown, & G. L. Peterson (Eds.), Benefits of leisure (pp. 263). State Co llege, PA: Venture Publishing Inc. Driver, B. L. (1983). Master list of items for Recreati on Experience Preference scales and domains Unpublished document. USDA Forest Service, Fort Collins, CO: Rocky Mountain Forest and Range Experiment Station. Eagly, A., & Chaiken, S. (1993). The psychology of attitudes. New York: Harcourt Brace Jovanovich. Echtner, C. M., & Ritchie, J. R. (1993). The me asurement of destinati on image: An empirical assessment. Journal of Travel Research, 31, 3-13. Eisenhauer, B. W., Krannich, R. S., & Blahna, D. J. (2000). Attachments to special places on public lands: An analysis of activities, reasons for attachments, and community connections. Society & Natural Resources, 13 421-441. Fik, T. J. (1988). Hierarchical interaction: The modeling of a competing central placed system. The Annals of Regional Science, 22 48-69. Fik T. J., & Mulligan, G. F. (1990). Spatial flows and competing central places: Towards a general theory of hier archical interaction. Environment and Planning A, 527-549. Fishbein, M. (1967). A behavior th eory approach to th e relations between be liefs about an object and the attitude towards that object. In M. Fishbein (Ed.), Readings in attitude theory and measurement (pp. 389-400). New York: John Wiley. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research Reading, MA: Addison-Wesley. Fletcher, D., & Fletcher, H. (2003). Manageable predictors of park visitor satisfaction: Maintenance and personnel. Journal of Park and Recr eation Administration, 21 21-37. Flood, J. P. (2002). Influence of benchmarking on wilderness visitor and manager perceptions of campsite conditions. Proceedings of the 2002 Northeastern Recreation Research Symposium GTR-NE-302, 165-172. Floyd, M. F. (1997). Pleasure, arousal and domin ance: Exploring affective determinants of recreation satisfaction. Leisure Sciences, 19, 83-96. Floyd, M. F., & Gramann, J. H. (1997). Experiencebased setting management: Implications for market segmentation of hunters. Leisure Sciences, 19 113-127.

PAGE 166

166 Floyd, M. F., Jang, H., & Noe, F. P. (1997). Th e relationship between environmental concern and acceptability of environmental impacts am ong visitors to two U.S. national park settings. Journal of Environmental Management, 51 391-412. Fried, M. (1982). Residential attachment: Sources of residen tial and community satisfaction. Journal of Social Issues 38, 107-120. Gartner, W. C., & Hunt, J. D. (1987). An analys is of state image change over a twelve-year period (1971-1983). Journal of Travel Research, 27 15-19. Goodchild, M. F. (1987). Spatial autocorrelation CATMOG 47. Goyder, J. (1985). Face-to-face interviews and mailed questionnaires: The net difference in response rates. Public Opinion Quarterly, 77, 644-662. Graefe, A. R. Thapa, B., Confer, J. J., & Absher, J. D. (2000). Relationships between trip motivations and selected vari ables among Allegheny National forest visitors. In: D. N. Cole, S. F. McCool, W. T. Borrie, & J. OLoughlin (Eds.). Proceedings of the 1999 Wilderness Science in a Ti me of Change Conference RMRS-P-15, Vol. 4. Ogden, UT: USDA Forest Service, Rocky M ountain Research Station. p. 107-112. Graefe, A. R., & Fedler, A. J. (1986). Situational and subjective determinants of satisfaction in marine recreational fishing. Leisure Sciences, 8, 275. Greene, T. (1999). Cognition and the management of place. In B. L. Driver, D. Dustin, T. Baltic, G. Elsner, & G. Peterson (Eds.), Nature and the human spirit: Toward an expanded land management ethic. State College, PA: Venture Publishing, Inc. Greider, T., & Garkovich, L. (1994). Landscapes: The social construction of nature and the environment. Rural Sociology, 59 1. Gruen, T. W., Sommers, J. O., & Acito, F. (2000). Relationship marketing activities, commitment, and membership behavi ors in professional associations, Journal of Marketing, 64, 34. Haas, G., Aukerman, R., Lovejoy, V., & Welch, D. (2004). Water Recreation Opportunity Spectrum (WROS) User's Guidebook United States Department of Interior, Bureau of Reclamation. Office of Program and Policy Services, Denver Federal Center, Lakewood, Colorado. Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Englewood Cliffs, NJ: Prentice Hall. Hall, T., & Shelby, B. (1993). Wilderness monitoring. Mount Jefferson, Mount Washington, and Three Sisters Wildernesses. (Unpublished Tech. Rep. on file, pp. 93). Eugene, OR: USDA, Forest Service, Willamette National Forest.

PAGE 167

167 Hamilton, J. A., Crompton, J. J., & More, T. A. (1991). Identifying the dimensions of service quality in a park context. Journal of Environmental Management, 32 211. Hammitt, W. E., & Patterson, M. (1991) Coping behavior to avoid visual encounters: Its relationship to wildland privacy. Journal of Leisure Research, 23 225. Heberlein, T. A., & Vaske, J. J. (1977). Crowding and visitor conflict on the Bois Brule River. Technical report WIS WRC 77-04. Madison, WI: University of Wisconsin, Water Resources Center. Herrick, T. A., & McDonald, C. D. (1992). Factor s affecting overall satisfaction with a river recreation experience. Environmental Management, 16 243. Herting, J. R., & Guest, A. M. (1983). Compone nts of satisfaction with local areas in the metropolis. Sociological Quarterly 26, 99-115. Heywood, J. L., Christensen, J. E., & Stankey, G. H. (1991). The relationship between biophysical and social setting factors in the recreation opportunity spectrum. Leisure Sciences, 13 239-246. Holland, S. M. (1979). The relationship of satisfaction to visitor activities in a park setting Unpublished masters thesis, Texas A & M University. Howat, G., Absher, J. D., Crilley, G., & Milne, I. (1996). Measuring customer service quality in sports and leisure centers. Managing Leisure, 1, 77-89. Hu, L., & Bentler, P. M. (1999). Cutoff criteria fo r fit indexes in covari ance structure analysis: Conventional criteria vers us new alternatives. Structural Equation Modeling, 6 1-55. Hummon, D. M. (1992). Community attachment: Lo cal sentiment and sense of place. In I. Altman, & S. M. Low (Eds.), Place Attachment (pp. 253-278). New York: Plenum Press. Hunt, J. D. (1975). Image as a factor in tourism development. Journal of Travel Research, 13 17. Iso-Ahola, S. E. (1980). The social psychology of leisure and recreation. Dubuque, Iowa: W. C. Brown Company Publishers. Jacob, G. R., & Schreyer, R. (1980). Conflict in outdoor recreation: A theoretical perspective. Journal of Leisure Research, 12, 368-380. Jorgensen, B. S., & Stedman, R. C. (2001). Sense of place as an attit ude: Lakeshore owners attitudes toward their properties. Journal of Environmental Psychology, 21 233-248. Kaltenborn, B. P. (1998). Effects of sense of pl ace on responses to environmental impacts. Applied Geography, 18, 169-189.

PAGE 168

168 Kaltenborn, B. P. (1999). Setting preferences of Arc tic tourists: A study of some assumptions in the recreation opportunity spectrum framework from the Svalbard Archipelago, Norsk Geografisk Tidsskrift Norweg ian Journal of Geography, 53 145-55. Kaltenborn, B. P., & Williams, D. R. (2002). The meaning of place: attachments to Femundsmarka National Park, Norway, among tourists and locals, Norsk Geografisk Tidsskrift Norwegian Journal of Geography, 56, 189-198. Kaplan, S., & Talbot, J. F. ( 1983). Psychological benefits of a wilderness experience. In I. Altman & J. F. Wohlwill (Eds.), Behavior and the Natural Environment (pp. 163-203). New York: Plenum. Kil, N., Fik, T., Holland, S., & Glover, J. (2007, September). Modeling the effects of distance and intervening opportunities on recreati on demand: GIS spatial analysis approach A paper presented at the Nati onal Recreation and Park Asso ciation (NRPA) Congress & Exposition, Indianapolis, IN, September 25-29, 2007. Kline, R. B. (2005). Principles and practice of st ructural equation modeling (2nd ed.). New York, NY: Guildford Press. Kyle, G. T., Graefe, A. R., & Manning, R. E. (2003/2004). Satisfaction derived through leisure involvement and setting attachment. Leisure/Loisir, 28 (3-4), 277-306. Kyle, G. T., Graefe, A. R., & Manning, R. E. (2005). Testing the dimensionality of place attachment in recreational settings. Environment and Behavior, 37 153-177. Kyle, G. T., Graefe, A. R., Manning, R. E., & Bacon, J. (2003). An examination of the relationship between leisure activity involvement and place attachment among hikers along the Appalachian Trail. Journal of Leisure Research, 35 249. Kyle, G. T., Graefe, A. R., Manning, R. E., & B acon, J. (2004a). Effect of activity involvement and place attachment on recreationist s perceptions of setting density, Journal of Leisure Research, 36, 209-231. Kyle, G. T., Graefe, A. R., Manning, R. E., & B acon, J. (2004b). Effects of place attachment on users perceptions of social and environmental conditions in a natural setting. Journal of Environmental Psychology, 24 213. Kyle, G. T., Mowen, A. J., & Tarrant, M. (2004) Linking place preferences with place meaning: An examination of the relationship between place motivation and place attachment. Journal of Environmental Psychology, 24 439-454. Lawler, E. E. (1973). Motivations in work organization Monterey, CA: Brooks/Cole. Linder, J. R., Murphy, T. H., & Briers, G. E. (2001). Handling nonresponse in social science research. Journal of Agricultural Education, 42, 43-53.

PAGE 169

169 Loomis, J. B. (2000). Counting on recreation use data: A call for l ong-term monitoring. Journal of Leisure Research, 32, 93-96. Lynch, K. (1960). The image of the city Cambridge, MA: MIT Press. MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size fo r covariance structure modeling. Psychological Methods, 1, 130-149. MacKay, K. J., & Crompton, J. L. (1988). A conceptual model of consumer evaluation of recreation service quality. Leisure Studies, 7, 41. MacKay, K. J., & Crompton, J. L. (1990). Meas uring the quality of recreation services. Journal of Park and Recreation Administration, 8 47. Manfredo, M. J., & Driver, B. L. (1996). Measuri ng leisure motivation: A meta-analysis of the recreation experience preference scales, Journal of Leisure Research, 28 188-213. Manning, R. (1999). Studies in outdoor recreation: Sear ch and research for satisfaction, (2nd ed.) Corvallis OR: Oregon State University Press. Marcin, T. C. (1993). Demographic change : Implications for forest management, Journal of Forestry, 91 (11), 39-45. McGuire, F., OLeary, J. T., & Dottavio, F. D. (1989) The importance of selected facilities, programs and services to older visitors to national parks. Journal of Park and Recreation Administration, 7 1. Meinig, D. W. (1979). The interpretation of ordinary landscapes New York: Oxford University Press. Mesch, G. S., & Manor, O. (1998). Social ties, environmental perception, and local attachment, Environment and Behavior, 30 504-519. Miron, J. (1984). Spatial autocorrelation in regr ession analysis: A beginners guide. In: G. L. Gaile & C. J. Willmott (Eds.), Spatial statistics and models (pp. 201). Reidel, Dordrecht. Mitchell, M. Y., Force, J. E., Carroll, M. S., & McLaughlin, W. J. (1993). Forest places of the heart: Incorporating special spaces into public management. Journal of Forestry 4, 3237. Moore, R. L., & Scott, D. (2003). Place attachment and context: Comparing a park and a trail within. Forest Science, 49 877-884.

PAGE 170

170 Moore, R., & Graefe, A. (1994). Attachments to recreation settings: The case of rail-trail users. Leisure Sciences, 16, 17-31. Moran, P. A. P. (1950). Notes on continuous stochastic phenomena. Biometrika, 31 17-23. Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. Nunnally, J. C. & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGrawHill. Odland, J. (1988). Spatial autocorrelation Beverly Hills, CA: Sage Publications. Oliver, R. L., & DeSarbo, W. S. (1988). Re sponse determinants in satisfaction judgments. Journal of Consumer Research, 14, 495. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A conceptual model of service quality and its implications for further research. Journal of Marketing, 49 41. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple-item scale for measuring consumer per ceptions of service quality. Journal of Retailing, 64 12. Patterson, M. E., & Williams, D. R. (1991). A transactional approach to characterize relationship to resource. Paper presented at the NRPA Sy mposium, October, Baltimore, MD. Payton, M. A., Fulton, D. C., & Anderson, D. H. ( 2005). Influence of place attachment and trust on civic action: A study at Sherbur ne National Wildlife Refuge. Society and Natural Resources, 18, 511. Pearce, P. L. (1982). Perceived ch anges in holiday destinations. Annals of Tourism Research, 9 145-64. Peterson, G. L. (1974). Evaluating the qua lity of the wilderness environment. Environment and Behavior, 6 169. Prentice, R. C., Witt, S. F., & Hamer, C. (1998) Tourism as experience: The case of Heritage Park. Annals of Tourism Research, 22 1-15. Proshansky, H. M. (1978). The city and self-identity. Environment and Behavior, 10 147-169. Proshansky, H. M., Fabian, A. K., & Kaminoff, R. (1983). Place identity: Physical world and socialization of the self. Journal of Environmental Psychology, 3 57. Relph, E. (1976). Place and placelessness. London: Pion Limited.

PAGE 171

171 Richards, M. T., King, D. A., Daniel, T. C., & Brown, T. C. (1990). The lack of an expected relationship between travel cost and contingent value estimates of forest recreation value. Leisure Sciences, 12, 303-319. Rossman, B. B., & Ulehla, Z J. (1977). Psychological reward va lues associated with wilderness use: A functional reinforcement approach. Environment and Behavior, 9 41-65. Ryden, K. C. (1993). Mapping the invisible landscape: Folklore, writing, and the sense of place IA: University of Iowa Press. Sack, R. D. (1992). Place, modernity, and the consumers world: A relational framework for geographical analysis Baltimore: John Hopkins University Press. Schreyer, R. (1989). Recreation habitat: An emergi ng concept. In: Towards Serving Visitors and Managing Our Resources Proceedings of a North American Workshop on Visitor Management in Parks and Protected Areas, pp. 93-105. Tourism Research and Education Schreyer, R., & Roggenbuck, J. (1978). The infl uence of experience expectations on crowding perceptions of social-psyc hological carrying capacities. Leisure Sciences, 1 373-394. Shafer, C., & Hammitt, W. (1995). Congruenc y among experience dimensions, cognitive indicators, and coping be haviors in wilderness. Leisure Sciences, 17 263-279. Shamai, S. (1991). Sense of place: An empirical measurement. Geoforum, 22 347-358. Shelby, B. (1980). Crowding models for backcountry recreation. Land Economics, 56, 43-55. Shin, W. S., Jaakson, R., & Kim, E. I. (2001). Benefits-based analysis of visitor use of Sorak-San National Park in Korea, Environmental Management, 28 413-419. Shumaker, S. A., & Taylor, R. B. (1983). Toward a clarification of peopl e-place relationships: A model of attachment to place. In N. R. Feimer & E. S. Geller (Eds.), Environmental psychology: Directions and perspectives (p. 219-251). New York: Praeger. Smaldone, D., Harris, C., Sanyal, N., & Lind, D. (2005). Place attachment and management of critical park issues in Grand Teton National Park. Journal of Park and Recreation Administration, 23 90-114. Sokal, R. R., & Oden, N. L. (1978). Spa tial autocorrelation in biology 1: Methodology. Biological Journal of the Linnean Society, 10 199-228. Steiger, J. H. (1989). EZPATH: A supplementary modul e for SYSTAT and SYGRAPH Evanston, IL: SYSTAT. Steiger, J. H. & Lind, J. C. (1980, June). Statistically based tests for the number of common factors Paper presented at the Psychometric Society Annual Meeti ng, Iowa City, IA.

PAGE 172

172 Stedman, R. C. (2000). Up North: A social psychology of place Unpublished doctoral dissertation, University of Wisconsin-Madison. Stedman, R. C. (2002). Toward a social psyc hology of place: Predicting behavior from place based cognitions, attitude, and identity. Environment and Behavior, 34 561. Stedman, R. C. (2003a). Is it rea lly just a social construction?: The contribution of the physical environment to sense of place. Society and Natural Resources, 16 671-685. Stedman, R. C. (2003b). Sense of place and forest science: Toward a program of quantitative research. Forest Science, 49 822-829. Steele, F. (1981). The sense of place. Boston, MA: CBI. Stein, T. V., & Lee, M. E. (1995). Managing re creation resources for positive outcomes: An application of benefits -based management. Journal of Park and Recreation Administration, 13, 52-70. Stokols, D., & Shumaker, S. A. (1981). People in places: A transactional view of settings. In D. Harvey (Ed.), Cognition, social behavi or, and the environment (pp. 441-488). Hillsdale, NJ: Erlbaum. Tinsley, H. E. A., Driver, B. L., & Kass, R. A. (1982). Reliability and concurrent validity of the recreation experience preference scales. Journal of Educatio nal and Psychological Measurement, 41, 897. Tobler, W. R. (1970). A computer movie simu lating urban growth in the Detroit region. Economic Geography, 46 234-240. Tuan, Y. F. (1974). Topophilia: A study of environmental perception, attitudes and values Englewood Cliffs, NJ: Prentice-Hall. Tuan, Y. F. (1975). Place: An experiential perspective. Geographical Review, 65 151-165. Tuan, Y. F. (1977). Space and place: The perspective of experience Minneapolis, MN: University of Minnesota Press. Tuan,Y. F. (1980). Rootedness verses sense of place. Landscape, 24 (1), 3-8. Urry, J. (1990b). The tourist gaze. London: Routledge. Urry, J. (1995). Consuming places London: Routledge. Vaske, J. J., & Kobrin, K. C. (2001). Place attachment and environmentally responsible behavior, Journal of Environmental Education, 32 (4), 16-21.

PAGE 173

173 Vorkinn, M., & Riese, H. (2001). Environmental concern in a local context: The significance of place attachment. Environment and Behavior 33 249-263. Walker, G. J., & Chapman, R. (2003). Thinking lik e a park: The effects of sense of place, perspective-taking, and empathy on pro-environmental intentions. Journal of Park and Recreation Administration, 21 71-86. Walmsley, D. J., & Young, M. (1998). Evaluative images and tourism: The use of personal constructs to describe the st ructure of destination images. Journal of Travel Research, 36, 65-69. Warzecha, C., & Lime, D. (2001). Place attachment in Canyonlands National Park: Visitors assessments of setting attributes on the Colorado and Green Rivers. Journal of Park and Recreation Administration 19, 59-78. Wellman, J. D. (1987). Wildland recreation policy New York: John Wiley. Wellman, J. D., Roggenbuck, J. W., & Smith, A. C. (1982). Recr eation specialization and norms of depreciative behavior among canoeists. Journal of Leisure Research, 14, 323-340. Whisman, S. A., & Hollenhorst, S. J. (1998). A Pa th model of whitewater boating satisfaction on the Cheat River of West Virginia. Environmental Management, 22 109. Wickham, T. D., & Kerstetter, D. L. (2000). The relationship between place attachment and crowding in an event setting. Event Management, 6 167-174. Williams, D. R. (1989). Great expectations and the limits to satisfaction: A review of recreation and consumer satisfaction re search. In: A. Watson (Ed.), Outdoor recreation benchmark 1988: Proceedings of the nati onal outdoor recreation forum USDA Forest Service General Technical Report SE-52, 422-438. Williams, D. R., & Patterson, M. E. (1994). Mapping the meaning of landscape: A framework for research on human dimensions of natural resources. Paper presented at the Fifth International Symposium on So ciety and Natural Resource Ma nagement, Fort Collins, Williams, D. R., Patterson, M. E., Roggenbuck, J. W., & Watson, A. E. (1992). Beyond the commodity metaphor: Examining emotional and symbolic attachment to place. Leisure Sciences, 14, 29-46. Williams, D. R., & Roggenbuck, J. W. (1989). Measuring place attachment: some preliminary results. In L. H. McAvoy, & D. Howard (Eds.), Abstracts-1989 leisure research symposium (pp. 32). Arlington, VA: National Re creation and Park Association. Williams, D. R., & Stewart, S. I. (1998). Sense of place: An elusive concept that is finding a home in ecosystem management. Journal of Forestry, 96 18-23.

PAGE 174

174 Williams, D. R., & Vaske, J. J. (2003). The measurement of place attachment: Validity and generalizability of a psychometric approach. Forest Science, 49 830-840. Wilson, A. G., & Bennett, R. J. (1985). Mathematical methods in human geography and planning Wiley, Chichester. Young, M. (1995). Evaluative constructions of domestic tourist places. Australian Geographical Studies, 33, 272-286. Young, M. (1999). The relationship between tourist motivations and the interpretation of place meanings. Tourism Geographies, 1 387-405. Yuan, M. S., & McEwen, D. (1989). Test for campers' experience preference differences among three ROS setting classes. Leisure Sciences, 11, 177-185. Zeithaml, V. A., Berry, L. L., & Parasuraman, A. (1988). Communication and control processes in the delivery of service quality. Journal of Marketing, 52, 35. Zeithaml, V. A., Parasuraman, A., & Berry, A. A. (1985). Problems and strategies in services marketing. Journal of Marketing, 49 33.

PAGE 175

BIOGRAPHICAL SKETCH Na myun Kil earned a Bachelor of Arts in tour ism and interpretation at Kwangju University and a Master of Science in recreation at S outhern Illinois Universi ty-Carbondale. Before attending the University of Florid a, he served as a recreation th erapist working with people with disabilities in Southern Illinois. He earned hi s Ph.D. in natural resource recreation in December 2008. He has focused his research on recreation a nd planning management in natural resource recreation areas and community-bas ed recreation areas including pa rks, forests, and protected areas.