The Context of Place: Issues Measuring Place Attachment across Urban Forest Contexts

The Context of Place: Issues Measuring Place Attachment across Urban Forest Contexts Abstract Since the late 1990s, forest managers and researchers have been encouraged to incorporate place attachment into their projects and decision-making. Place attachment measures indicate the strength of the human-place bond and provide insight as to why people are attached, thus informing management decisions and approaches to visitor and community relations. Place-attachment scales have been modified and translated across the globe with an assumed measurement equivalence. This assumption may present issues when comparing data collected across different contexts and implementing results based on those findings. This project assessed place attachment measurement among visitors to urban forested areas in Austria and Minnesota. Confirmatory factor analyses assessments of equivalent structure, factor loading pattern, and intercepts between samples revealed that place-attachment scales, as currently used, may not measure identical conceptualizations across contexts. As such, their use without critical examination could lead to inappropriate interpretations of and actions taken related to place attachment. cross-cultural comparison, scale psychometric property assessment, urban forest visitors Management and Policy Implications For over two decades, forest managers have been encouraged to incorporate noneconomic values into their decision-making frameworks. Place attachment has often been promoted as one of the constructs that may accomplish this goal. In part, this is because of the relative ease of including established place-attachment scales in stakeholder surveys and the scales’ quantitative results. These results should not preclude using place attachment, when appropriate, in decision-making frameworks. Indeed, measures of the human-place bond have been shown to be useful in understanding stakeholder values, pro-environmental attitudes, and behaviors, and envisaging potential reactions to management decisions. However, more care is needed when developing and interpreting the data from projects that use place-attachment scales. Specifically, these results and those of previous investigations suggest that the entire set of place attachment items should be used in every survey (rather than a subset of half the items or sometimes even less, as has occurred in many projects). Furthermore, projects should include survey pretests to identify data collection or measurement problems early. Doing so will allow managers to confidently use place-attachment scale data to inform decisions about environmental concern, stakeholder group conflict, recreational visitor motivations, and other management issues. Place attachment, or the human-place bond (Altman and Low 1992), has been present in academic discourse since the 1990s, and its use has been advocated in forest management for a similar duration (e.g., Williams and Stewart 1998, Morgan and Messenger 2009). As understanding of the construct and its applicability has grown, so has its use across contexts and cultures. However, assessments of the scale’s psychometric properties have not kept pace with their widespread use. Some authors (e.g., Williams and Vaske 2003, Budruk 2010, R. Cottrell and S.P. Cottrell 2015) note that as context is important to understand and subsequently measure individuals’ place attachment, more work needs to assess the conceptual equivalence and measurement properties of commonplace-attachment scales across sites and contexts. While calls for scale assessments have increased, few actual studies have emerged. One recent example investigated the measurement equivalence of a place-attachment scale used in multiple-forest sites and noted significant measurement differences across forest areas in Germany, Colorado, and Minnesota (Wynveen et al. 2017). This investigation built on these findings by assessing the measurement of place attachment among visitors to urban forests in Austria and the United States. Place Attachment Conceptualization While place attachment has been operationalized as the degree to which an individual values or identifies with a setting (Williams and Roggenbuck 1989), Wynveen et al. (2011) observed that place attachment measures also abstractly provide insight into the reasons why people value the setting via the theoretical and operational definitions of the multiple dimensions measured by place-attachment scales. In forest recreation research, the most frequently utilized place-attachment scale has been Williams and Roggenbuck’s (1989) two-dimensional (i.e., place identity and place dependence) measure. Place identity refers to the connections people have between a setting and their personal identity (Proshansky 1978). Moreover, place identity provides “the source of meaning for a given setting by virtue of relevant cognitive clusters that indicate what should happen in it, what the setting is supposed to be like, and how the individual and others are supposed to behave in it” (Proshansky et al. 1983, 67). Twigger-Ross and Uzell (1996) suggested place identity is a cognitive component of place attachment that can be observed as individuals use places to affirm their identity and express it to others. The other dimension in Williams and Roggenbuck’s scale is place dependence. Place dependence was first defined by Stokols and Shumaker (1981) as the “perceived strength of association between [an individual] and a specific place” (547). However, it has since been operationalized as the conative component of place attachment concerning “how well a setting serves goal achievement given an existing range of alternatives (‘how does this setting compare to others for what I like to do?’)” (Jorgensen and Stedman 2001, 234). Beyond place dependence and place identity, some researchers (e.g., Kyle et al. 2004) have suggested and found empirical support for other dimensions of place attachment. For example, Mesch and Manor (1998) suggested the dimension of social bonding (i.e., social ties to a setting that are develop though shared experiences the setting). Others (e.g., Jorgensen and Stedman 2001) suggested an affective attachment dimension (defined as an emotional bond to a place). We chose to limit this investigation to the original bidimensional conceptualization of place dependence and place identity because these have been the most commonly used and empirically verified dimensions. Implications for Forest Management In their report, Farnum et al. (2005) review several projects where place-related concepts have been used to inform several types of forest management decisions. For example, they indicated that place concepts have been used to understand stakeholders’ responses to environmental impacts, stakeholders’ support or opposition to management policies and initiatives (e.g., user fees), and conflict between stakeholder groups. Others, such as Kil et al. (2012) have observed that ideas of place could be used in conjunction with management frameworks (e.g., the recreation opportunity spectrum) to inform management of the wildland-urban interface. Furthermore, research indicates that place attachment may be able to be incorporated into more projects where understanding stakeholder perceptions are important. For example, Rickard and Stedman (2015) observed the correlation between place and communication in shaping stakeholder perceptions of protected areas and the activities they participate in at those areas. Similarly, Wynveen, Kyle, and Sutton (2010) and Wynveen, Wynveen, and Sutton (2015) have observed or argued that understanding the human-place bond can be used to help prime environmental communication and education messages designed to encourage pro-environmental behavior. Besides communication, understanding the reasons for stakeholders’ place attachment can be used in planning and managing protected areas because place “provides the context in which the problems can be recognized and articulated, and within which different values can be understood, conflicts resolved and choices made” (Potschin and Haines-Young 2013, 1054). For instance, Kil et al. (2012) suggested that place provides insight into understanding stakeholders desired benefits from protected areas, and Wynveen and Sutton (2015) and Wynveen , Kyle, and Theodori (2010) observed the role that fostering place attachment has in building trust between stakeholders and management agencies. Clearly, these few examples suggest that place attachment has a role to play in forest management. Measurement Concerns Place attachment survey items have been modified and translated for use in contexts across the globe. However, item equivalency has often been assumed in the literature, which Budruk (2010) indicated may be problematic when comparing data collected among samples from differing social and physical contexts. In her review of cross-cultural and cross-language research, Budruk identified four areas of equivalence that should be met when using a scale to measure a construct and make comparisons across contexts: semantic, functional, normative, and conceptual. Semantic equivalence involves ensuring nearly identical definitions of words or phrases when translating a scale (Behling and Law 2000). Beyond semantic equivalence is functional equivalence. Functional equivalence occurs when the meaning and intent of the items are virtually identical across samples (Kozak et al. 2004). In other words, people from different groups similarly interpret both the denotation and the connotation of survey questions. Normative equivalence (i.e., similarity in which social rules influence communication) can be difficult to achieve when social conventions discourage directly addressing a concept or cause differing operational interpretations or expressions of a concept (Behling and Law 2000). For example, some groups consider directly providing constructive criticism, even when asked, inappropriate. Lastly, conceptual equivalence is achieved when a concept can be operationalized in the same manner between populations (Groth-Marnat 2009). For instance, each group understanding that the concept of “identity” refers to how an individual conceives of or expresses themselves when responding to the question, “Who am I?” Failure to achieve these equivalences may lead to faulty interpretations of data when comparing samples from varying social (and sometimes physical) contexts. For example, reliability issues may result from translating items between languages (Harkness et al. 2010), or cultural differences may lead to different interactions with the setting resulting in varying conceptualizations of place (Kyle and Johnson 2008; Trentelman 2009). Similarly, previous research has indicated that activity type (Moore and Scott 2003) and involvement (Bricker and Kerstetter 2000) may influence individuals’ place conceptualizations. As Wynveen et al. (2017) posit, it is possible that the different experiences (i.e., variance in activity type and activity involvement) moderated by their context may influence how people interpret and respond to the place-attachment scale items. (For a further review of issues associated with the use of place-attachment scales across contexts, see Budruk [2010] and Wynveen et al. [2017]). Failure to achieve these equivalences can manifest in metric inequivalence—differing psychometric properties of a scale across populations (Groth-Marnat 2009). Hence, this project explored the measurement equivalence of place attachment between two urban forest contexts: Vienna, Austria (Lobau and Prater parks) and Minneapolis, Minnesota (Fort Snelling State Park and the adjacent Fort Snelling Minnesota Historical Site). Methods To assess the measurement equivalence of a place-attachment scale across the Minnesotan and Austrian contexts, we used identical on-site instruments and performed analysis similar to that of developing new instruments (e.g., tests of reliability and validity suggested by DeVellis [2003]). Specifically, we interpreted confirmatory factor analysis results, assessed scale reliability and validity, and conducted scale-measurement-equivalence assessments between the subsamples. The instruments and methods were designed and implemented with the approval of the University of Minnesota Institutional Review Board and fulfilled Austrian national ethical requirements. Instrument Design Data were collected via on-site instruments developed by a research team that included native English- and German-speakers familiar with the study settings. Four items, chosen based on their successful application in previous studies, measured place dependence, and three items measured place identity (see Table 1), adapted from Williams and Vaske’s (2003) two-dimensional place-attachment scale. Table 2 includes the items originally developed by Williams and Vaske. Table 1. Factor loading and descriptive results. Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — *p ≤ 0.05, ***p ≤ 0.001 a,b Means with different superscritps are significantly different c PD4 removed due to low-factor loading View Large Table 1. Factor loading and descriptive results. Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — *p ≤ 0.05, ***p ≤ 0.001 a,b Means with different superscritps are significantly different c PD4 removed due to low-factor loading View Large Table 2. Williams and Vaske’s (2003) place-attachment scale. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. View Large Table 2. Williams and Vaske’s (2003) place-attachment scale. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. View Large All items were measured on a five-point agreement scale (i.e., 1 = “strongly disagree,” 2 = “disagree,” 3 = “neither,” 4 = “agree,” and 5 = “strongly agree”). Respondents also answered a series of demographic questions. The researchers checked normative question equivalence between languages and the questionnaire was translated into German and back-translated to check semantic equivalence. A team member fluent in German and English also provided guidance regarding functional and conceptual equivalence. Since both sites in Austria and both sites in Minnesota were adjacent to one another, data of the two Austrian and the two Minnesota sites were pooled together to have a more robust database. Study Sites and Sampling On-site data collection occurred during the summer of 2015 (June to September) through an intercept survey at each of the parks. In Minnesota, Fort Snelling State Park is a 1,500-hectare day-use area managed by the Minnesota Department of Natural Resources located on the Mississippi River between Minneapolis and St. Paul. Over 400,000 visitors per year hike, bike, cross-county ski, and fish in its floodplain forests, meadows, marshes, and lakes. Adjacent to the state park and also bordering the Mississippi River, historic Fort Snelling, a 14-hectare state-owned parcel abuts a significant urban regional park managed by the Minneapolis Park Board. Trails in both the state park and the historical site connect nearby urban regional trail systems. Visitors can also arrive by personal vehicle and public transportation. Similarly, Vienna’s Lobau (1,037 hectares) and Prater (600 hectares) parks are located in a forested floodplain near the city center. Located in close proximity, they are almost adjacent to one another on either side of the Danube River. Both parks have extensive trails for hiking and biking. Prater also has more developed areas similar to those found in large city parks (e.g., playgounds, sport courts, and dog park). Lobau is managed by Municipal Department 49: Forestry Office and Urban Agriculture of the city of Vienna, and Prater is managed by Municipal Department 42: Parks and Gardens. As with the Minnesota sites, the Austrian parks are accessible via a network of regional trails and public transportation. Visitors were selected through a stratified-cluster sample and systematically selected sampling period that varied in time of day and day of the week to capture a diverse visitor segment. One adult per travel group was asked to complete the questionnaire. Respondents took between 10 and 20 minutes to complete the questionnaire. The length of the survey limited sample sizes to 471 in Minnesota (response rate = 39 percent) and 310 in Austria (response rate = 34 percent). Based on our observations during data collection, we believe that reasons for nonresponse had primarily to do with time constraints: visitors did not want to disrupt their activity. This was especially true of bicyclists. Nonresponse information collected in Austria, via observation of those who declined participation, indicated that most nonrespondents were bicyclists (48.9 percent), followed by hikers (25.5 percent), dog walkers (14.1 percent), joggers (10.7 percent) and Nordic walkers and horse riders (0.8 percent). Hence, compared to previous counts observed in Lobau (Arnberger 2006), bicyclists may have been under represented and dog walkers overrepresented. However, as the purpose of this investigation was not to generalize the findings to these populations of visitors but rather to identify measurement concerns with the place-attachment scale, we chose to acknowledge the limitation and proceed with the analysis. Data Analysis After calculating the descriptive statistics to document the makeup of the sample, we began the scale assessment by conducting a confirmatory factor analysis (CFA) for each subsample (Minnesotan, Austrian) using LISREL 9.2 to assess the hypothesized two-dimensional place-attachment model. Within the measurement model, correlation was allowed between the place-dependence and place-identity dimensions. The individual CFAs allowed for identification of items that should be removed due to low-factor loading onto the intended dimension or cross-loading. After determining a final item set, goodness-of-fit statistics for each sample were calculated to evaluate the place-attachment scale measurement model across both subsamples independently. Several measures of Goodness-of-fit assessed root mean square error of approximation (RMSEA), comparative fit index (CFI), and non-normed fit index (NNFI) (Byrne 1998). For the RMSEA, values ≤ 0.08 indicate acceptable fit (Steiger and Lind 1980); for the CFI (Hu and Bentler 1995), values ≥ 0.95 indicate acceptable fit; and Bentler and Bonnett (1980) designated values ≥ 0.95 as acceptable for the NNFI. Place attachment measurement was cross-validated via LISREL 9.2’s multiple-group procedure. That is, testing ensued to see if the models measured place attachment equivalently via tests of equality of factor structure (equal form) and metric invariance (equal pattern of factor loadings). The equal form test indicates whether “the number of factors and pattern of indicator-factor loadings is identical across groups,” and metric invariance indicates whether factor loadings are equivalent across samples (Brown 2006, 268). We also tested for scalar invariance—the equality of intercepts (Marsh et al. 2017). Modification indices provided by LISREL allowed hypothesizing about the origin of any differences identified. Finally, several calculations assessed the scale’s reliability and construct validity. Cronbach’s alpha calculations for both the place-dependence and place-identity dimensions for each subsample assessed reliability (i.e., internal consistency). The CFA results provided insight into discriminant validity. Comparing sample demographics and t-tests between the place identity and place-dependence dimensions’ composite scores for each subsample assessed convergent validity. Results Overall, the sample was slightly more male (53 percent, n = 398) than female, and respondents had a mean age of 43 years (sd = 14.72); however, the Austrian sample was slightly younger (mean difference = 2 years; t = 2.22; p ≤ 0.03). A majority of the Minnesotan sample (78 percent, n = 355) had earned at least a college degree, while the Austrian sample was more evenly split, with 57 percent (n = 175) having earned a college degree or higher. Differences existed among the subsamples concerning the number of previous visits and the first year they visited their respective park. The Minnesota sample had visited significantly less in the 12 months prior to the survey than the Austrian visitors (~16 days, sd = 48.15 and ~100 days, sd = 105.87, respectively; t = 12.62, p < 0.001). On average, the Austrian sample first visited their park in 1988; the Minnesotan sample first visited in 2005 (t = 14.17, p < 0.001). While each subsample cited a variety of activities as their “main activity” on the day they were surveyed, the most frequently cited activity was hiking/walking (Minnesota: 59 percent, n = 279; Austria 34 percent, n = 103). After analyzing the demographic data, the overall measurement fit of the data to the place-attachment model was assessed. Four items hypothesized to measure place dependence and three items hypothesized to measure place identity were included in the instrument distributed to visitors at each site. Initial results for pooled data indicated that one of the place dependence items, PD4—“Doing my activities here is more important to me than doing them in any other place” —should be removed due to low-factor loading (Table 1). The fit indices for the final set of items were all within an acceptable range (Austria: χ2df=7 = 15.26, RMSEA = 0.06, NNFI = 0.97, CFI = 0.99, GFI = 0.98; Minnesota: χ2df=7 = 11.78, RMSEA = 0.04, NNFI = 0.99, CFI = 0.99, GFI = 0.99), indicating that the measurement model was a good fit for the data from each sample. Scale Reliability and Convergent Validity After determining the adequacy of the measurement model via the initial CFAs, reliability and validity were assessed. The place-attachment dimensions had an acceptable internal consistency statistic (α ≥ 0.68) for place identity and place dependence across the samples. Following a suggestion by Williams and Vaske (2003), the subsamples’ descriptive statistics and a pair of t-tests designed to identify differences in mean place-dependence and place-identity scores between the subsamples (see Table 1) assessed convergent validity. As the subsamples had varying visitation frequencies and tenure lengths, past research (e.g., Hammitt et al. 2004; Williams and Vaske 2003) suggested the place-dependence and place-identity scores should also significantly differ from one another. Differences were observed for both the place-dependence and place-identity mean scores between the samples (place dependence: Austria, m = 3.45, sd = 0.88; Minnesota, m = 3.18, sd = 0.97, t = 3.93, p ≤ 0.001; place identity: Austria, m = 3.94, sd = 0.90; Minnesota, m = 3.80, sd = 0.96, t = 2.09, p ≤ 0.05). Combined, these analyses suggest that the place-attachment scale meets minimum reliability and convergent validity expectations. Scale Cross-Validation The next analysis step was to cross-validate the place-attachment scale between the two subsamples. Acceptable goodness-of-fit indices (χ2df=14 = 29.28, RMSEA = 0.05, NNFI = 0.99, CFI = 0.99, GFI = 0.99; Table 3) for the test of equal form indicated the number of factors and pattern of indicator-factor loadings were identical among the samples. However, a chi-square difference test indicated a significant difference in the pattern of factor loadings between the datasets (Δχ2df=20 = 27.80, p ≤ 0.001). To identify where the difference occurred in the factor-loading pattern, post hoc analyses utilizing the modification indices (i.e., substantive change in chi-squared values) calculated via LISREL 9.2 were used. These results suggested that while the place dependence pattern of factor loadings was equivalent between the two samples, the factor loading pattern for place identity was not. Lastly, we tested the place dependence measurement for scalar invariance between the groups. Fit indices worsened (χ2df=3 = 32.01, RMSEA = 0.15, NNFI = 0.94, CFI = 0.97, GFI = 0.93) when the intercepts were constrained equally across the groups. Post hoc analysis indicated that the intercept corresponding to item PD2, “I wouldn’t substitute any other recreation area for the type of recreation I can do here,” was the likely source of the variance. However, it is important to note that scalar invariance is an ideal not often achieved in practice by psychometric scales (Marsh et al. 2017). Table 3. CFA goodness of fit for the measurement model. Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 ***≤0.001 View Large Table 3. CFA goodness of fit for the measurement model. Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 ***≤0.001 View Large Discussion We continue to encourage the use of place attachment to aid in understanding forest stakeholders and inform decision-making process, especially when used for monitoring and making within site comparisons. However, in light of the current observations and similar results from previous studies (Budruk 2010; Wynveen et al. 2017), these results suggest place-attachment measurement varies among samples from differing contexts. One explanation may be linguistic differences, even though the current and past investigations have used standard methods of forward- and back-translation. Alternatively, several researchers have observed differences in place conceptualization among people with different experience and cultural backgrounds (e.g., Kyle and Johnson 2008, Trentelman 2009). Hence, it may be that existing place-identity items are sensitive to these differences and therefore result in the measurement variance. Future research should attempt to identify the source of the variance within each item. Possible areas to focus on may involve linguistic or behavioral experience issues. For example, concerning linguistics, there may be differences in how different groups of people interpret words and phrases when they are not clear about their exact meaning (conceptual meaning), differences in how groups have used specific words in the past (linguistic semantics), or differences in how context affects word meaning (pragmatics) (Akmajian et al. 2001). Alternatively, there could be slight variation in the way different groups experience places that have not been captured by common experience use history or activity preference variables. Future research should focus on an exhaustive list of activity preference, experience use history, mode of experience, and other variables that may influence how people experience a place, which in turn, may influence how they respond to particular place-identity measurement items. Given the current results, managers and researchers should take steps to mitigate measurement issues with the place-attachment scales, especially when making comparisons across contexts. Two recommendations can easily be implemented. First, the entire set of items hypothesized to measure place attachment should be used rather than just three or four. Using the breadth of items will allow investigators to remove items that do not statistically behave as expected. Second, as with all research, investigations should include measurement instrument pretests to identify measurement problems before data are collected. Hopefully, by following these guidelines and others regarding data collection, analysis, and interpretation, forest managers can feel confident in using data elicited from place-attachment scales to make decisions affecting recreational visitors and other stakeholders. Specifically, there are several examples in the literature where natural resource managers have used place attachment to help understand environmental concern, awareness of resource impacts, pro-environmental behavior adoption, attitudes toward fees, inter- and intra-stakeholder group conflict, and recreational visitor motivations (Eder and Arnberger 2012; Farnum et al. 2005). To use place attachment data to inform decisions about these issues and others, Williams and Stewart (1998) provide a four-step framework for managers to follow. First, identify and use place names that stakeholders use to identify spaces within the forest. Next, use these names and other place-specific language when communicating management plans to stakeholders. Third, recognize that the reasons for attachment may vary between, or even within, stakeholder groups. Lastly, plan for greater management attention among contested areas. In sum, this investigation indicated a need to assess the psychometric properties of the place-attachment scale and take appropriate action to address any abnormalities during every investigation to ensure that the data provide valid information to inform an understanding of people’s connection to place. Acknowledgements The authors would like to thank the USDA Forest Service Northern Research Station for providing funding that contributed to this project. They would also like to thank Dr. Gerard Kyle for his advice on the manuscript. Literature Cited Akmajian , A. , R.A. Demer , A.K. Farmer , and R.M. Harnish . 2001 . Linguistics: An introduction to language and communication . MIT Press , Cambridge, MA . Altman , I. and S. Low . 1992 . Place attachment, human behavior, and environment: Advances in theory and research Vol. 12 . Plenum Press , New York, NY . Arnberger , A . 2006 . Recreation use of urban forests: an inter-area comparison . Urban For. Urban Green . 4 ( 3–4 ): 135 – 144 Google Scholar CrossRef Search ADS Behling , O. , and K.S. Law . 2000 . Translating questionnaires and other research instruments: Problems and solutions . Sage , Thousand Oaks, CA . Google Scholar CrossRef Search ADS Bentler , P. , and D. Bonett . 1980 . Significance tests and goodness of fit in the analysis of covariance structures . Psychol. Bull . 88 : 588 – 606 . Google Scholar CrossRef Search ADS Bricker , K.S. , and D. Kerstetter . 2000 . Level of specialization and place attachment: an exploratory study of whitewater recreationists . Leis. Sci . 11 : 233 – 257 . Brown , T.A . 2006 . Confirmatory factor analysis for applied research . Guilford Press , New York, NY . Budruk , M . 2010 . Cross-cultural equivalency of the place attachment scale: a multigroup confirmatory factor analysis approach . J. Leisure Res . 42 ( 1 ): 25 – 42 . Google Scholar CrossRef Search ADS DeVellis , R.F . 2003 . Scale development: Theory and applications . Sage , Thousand Oaks, CA . Eder , R. , and A. Arnberger . 2012 . The influence of place attachment and experience use history on perceived depreciative visitor behavior and crowding in an urban national park . Environ. Manage . 50 ( 4 ): 566 – 580 . Google Scholar CrossRef Search ADS PubMed Farnum , J. , T. Hall , and L.E. Kruger . 2005 . Sense of place in natural resource recreation and tourism: an evaluation and assessment of research findings . General technical report PNW-GTR-660: US Department of Agriculture, Forest Service, Pacific Northwest Research Station . Groth-Marnat , G . 2009 . Handbook of psychological assessment . Wiley , Hoboken, NJ . Hammitt , W.E. , E.A. Backlund , and R.D. Bixler . 2004 . Experience use history, place bonding, and resource substitution of trout anglers during recreation engagements . J. Leisure Res . 36 : 356 – 378 . Google Scholar CrossRef Search ADS Harkness , J.A. , A. Villar , and B. Edwards . 2010 . Translation, adaptation, and design . P. 115 – 140 , in Survey methods in multinational, multiregional, and multicultural contexts , Harkness , J.A. , M. Braun , B. Edwards , T.P. Johnson , L. Lyberg , P. Mohler , B. Pennell , and T.W. Smith (eds). Wiley , Hoboken, NJ . Google Scholar CrossRef Search ADS Hu , L. , and P.M. Bentler . 1995 . Evaluating model fit . P. 76 – 99 , in Structural equation modeling: concepts issues and applications , R.H. Hoyle (ed.). Sage , Thousand Oaks, CA . Jorgensen , B.S. , and R. Stedman . 2001 . Sense of place as an attitude: lakeshore property owners’ attitudes toward their properties . J. Environ. Psychol . 21 : 233 – 248 . Google Scholar CrossRef Search ADS Kil , N. , T.V. Stein , S.M. Holland , and D.H. Anderson . 2012 . Understanding place meanings in planning and managing the wildland–urban interface: the case of Florida trail hikers . Landscape Urban Plan . 107 ( 4 ): 370 – 379 . Google Scholar CrossRef Search ADS Kozak , M. , E. Bigné , and L. Andreu . 2004 . Limitations of cross-cultural customer satisfaction research and recommending alternative methods . J. Qual. Assur. Hosp. Tour . 4 ( 3–4 ): 37 – 59 . Google Scholar CrossRef Search ADS Kyle , G.T. , and C.Y. Johnson . 2008 . Understanding cultural variation in place meaning . P. 109 – 134 , in Understanding concepts of place in recreation research and management , Kruger , L.E. , T.E. Hall , and M.C. Stiefel (eds). USDA For. Serv. Gen. Tech. Rep, Portland, OR . PNW-GTR-744. Kyle , G.T. , A.J. Mowen , and M. Tarrant . 2004 . Linking place preferences with place meaning: an examination of the relationship between place motivation and place attachment . J. Environ. Psychol . 24 : 439 – 454 . Google Scholar CrossRef Search ADS Marsh , H.W. , J. Guo , P.D. Parker , B. Nagengast , T. Asparouhov , B. Muthén , and T. Dicke . 2017 . What to do when scalar invariance fails: the extended alignment method for multi-group factor analysis comparison of latent means across many groups . Psychol. Methods . doi: 10.1037/met0000113 Mesch , G. , and O. Manor . 1998 . Social ties, environmental perception, and local attachment . Environ. Behav . 30 : 504 – 519 . Google Scholar CrossRef Search ADS Moore , R.L. , and D. Scott . 2003 . Place attachment and context: comparing a park and a trail within . For. Sci . 49 ( 6 ): 877 – 884 . Morgan , M. and B. Messenger . 2009 . Using an activity and place-based typology to explain visitor motivations . J. For . 107 ( 1 ): 44 – 49 . Potschin , M. , and R. Haines-Young . 2013 . Landscapes, sustainability and the place-based analysis of ecosystem services . Landsc. Ecol . 28 ( 6 ): 1053 – 1065 . Google Scholar CrossRef Search ADS Proshansky , H.M . 1978 . The city and self-identity . Environ. Behav . 10 ( 2 ): 147 – 169 . Google Scholar CrossRef Search ADS Proshansky , H.M. , A.K. Fabian , and R. Kaminoff . 1983 . Place identity: physical world socialization of the self . J. Environ. Psychol . 3 ( 1 ): 57 – 83 . Google Scholar CrossRef Search ADS Rickard , L.N. , and R.C. Stedman . 2015 . From ranger talks to radio stations: the role of communication in sense of place . J. Leisure Res . 47 ( 1 ): 15 – 33 . Google Scholar CrossRef Search ADS Raadik Cottrell , J. and S.P. Cottrell . 2015 . Sense-of-place influences on perceived environmental change effects on future holiday experiences to Saaremaa, Estonia . Scand. J. Hosp. Tour . 15 ( 4 ): 425 – 446 . Google Scholar CrossRef Search ADS Steiger , J.H. , and J.C. Lind . 1980 . Statistically-based tests for the number of common factors . Paper presented at the Annual Spring Meeting of the Psychometric Society , May, Iowa City, IA . Stewart , W.P. , D.R. Williams , and L. Kruger , eds. 2013 . Place-based conservation: Perspectives from the social sciences . Springer , Dordrecht, NL . Google Scholar CrossRef Search ADS Stokols , D. , and S.A. Shumaker . 1981 . People and places: a transactional view of settings . P. 441 – 488 , in Cognition, social behavior and the environment , J. Harvey (ed.). Erlbaum , Mahwah, NJ . Trentelman , C.K . 2009 . Place attachment and community attachment: a primer grounded in the lived experience of a community sociologist . Soc. Nat. Resour . 22 ( 3 ): 191 – 210 . Google Scholar CrossRef Search ADS Twigger-Ross , C.L. , and D.L. Uzzell . 1996 . Place and identity processes . J. Environ. Psychol . 16 : 205 – 220 . Google Scholar CrossRef Search ADS Williams , D.R. , and J.W. Roggenbuck . 1989 . Measuring place attachment: some preliminary results . P. 32 , in Abstracts: 1989 Leisure Research Symposium , eds. McAvoy , L.H. and D. Howard , National Recreation and Park Association , Arlington, VA . Williams , D.R. , and S.I. Stewart . 1998 . Sense of place: an elusive concept that is finding a home in ecosystem management . J. For . 96 ( 5 ): 18 – 23 . Williams , D.R. , and J.J. Vaske . 2003 . The measurement of place attachment: validity and generalizability of a psychometric approach . For. Sci . 49 ( 6 ): 830 – 840 . Wynveen , C.J. , G.T. Kyle , J.D. Absher , and G.L. Theodori . 2011 . The meanings associated with varying degrees of attachment to a natural landscape . J. Leisure Res . 43 ( 2 ): 290 – 311 . Google Scholar CrossRef Search ADS Wynveen , C.J. , G.T. Kyle , and S.G. Sutton . 2010 . Place meanings ascribed to marine settings: the case of the Great Barrier Reef Marine Park . Leisure Sci . 32 ( 3 ): 270 – 287 . Google Scholar CrossRef Search ADS Wynveen , C.J. , G.T. Kyle , and G.L. Theodori . 2010 . Place bonding and trust: the case of feral hog management surrounding Big Thicket National Preserve . J. Rural Soc. Sci . 25 ( 2 ): 58 – 80 . Wynveen , C.J. , I.E. Schneider , S. Cottrell , A. Arnberger , A.C. Schlueter , and E. Von Ruschkowski . 2017 . Comparing the validity and reliability of place attachment across cultures . Soc. Nat. Resour . doi: 10.1080/08941920.2017.1295499 Wynveen , C.J. , and S.G. Sutton . 2015 . Engaging the public in climate change-related pro-environmental behaviors to protect coral reefs: the role of public trust in the management agency . Mar. Policy . 53 : 131 – 140 . Google Scholar CrossRef Search ADS Wynveen , C.J. , B.J. Wynveen , and S.G. sutton . 2015 . Applying the value-belief-norm theory to marine contexts: implications for encouraging pro-environmental behavior . Coastal Manage . 43 ( 1 ): 84 – 103 . Google Scholar CrossRef Search ADS © 2018 Society of American Foresters This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Forestry Oxford University Press

The Context of Place: Issues Measuring Place Attachment across Urban Forest Contexts

Journal of Forestry , Volume Advance Article (4) – May 24, 2018

Loading next page...
 
/lp/ou_press/the-context-of-place-issues-measuring-place-attachment-across-urban-LAOQKMaCY7
Copyright
© 2018 Society of American Foresters
ISSN
0022-1201
eISSN
1938-3746
D.O.I.
10.1093/jofore/fvy001
Publisher site
See Article on Publisher Site

Abstract

Abstract Since the late 1990s, forest managers and researchers have been encouraged to incorporate place attachment into their projects and decision-making. Place attachment measures indicate the strength of the human-place bond and provide insight as to why people are attached, thus informing management decisions and approaches to visitor and community relations. Place-attachment scales have been modified and translated across the globe with an assumed measurement equivalence. This assumption may present issues when comparing data collected across different contexts and implementing results based on those findings. This project assessed place attachment measurement among visitors to urban forested areas in Austria and Minnesota. Confirmatory factor analyses assessments of equivalent structure, factor loading pattern, and intercepts between samples revealed that place-attachment scales, as currently used, may not measure identical conceptualizations across contexts. As such, their use without critical examination could lead to inappropriate interpretations of and actions taken related to place attachment. cross-cultural comparison, scale psychometric property assessment, urban forest visitors Management and Policy Implications For over two decades, forest managers have been encouraged to incorporate noneconomic values into their decision-making frameworks. Place attachment has often been promoted as one of the constructs that may accomplish this goal. In part, this is because of the relative ease of including established place-attachment scales in stakeholder surveys and the scales’ quantitative results. These results should not preclude using place attachment, when appropriate, in decision-making frameworks. Indeed, measures of the human-place bond have been shown to be useful in understanding stakeholder values, pro-environmental attitudes, and behaviors, and envisaging potential reactions to management decisions. However, more care is needed when developing and interpreting the data from projects that use place-attachment scales. Specifically, these results and those of previous investigations suggest that the entire set of place attachment items should be used in every survey (rather than a subset of half the items or sometimes even less, as has occurred in many projects). Furthermore, projects should include survey pretests to identify data collection or measurement problems early. Doing so will allow managers to confidently use place-attachment scale data to inform decisions about environmental concern, stakeholder group conflict, recreational visitor motivations, and other management issues. Place attachment, or the human-place bond (Altman and Low 1992), has been present in academic discourse since the 1990s, and its use has been advocated in forest management for a similar duration (e.g., Williams and Stewart 1998, Morgan and Messenger 2009). As understanding of the construct and its applicability has grown, so has its use across contexts and cultures. However, assessments of the scale’s psychometric properties have not kept pace with their widespread use. Some authors (e.g., Williams and Vaske 2003, Budruk 2010, R. Cottrell and S.P. Cottrell 2015) note that as context is important to understand and subsequently measure individuals’ place attachment, more work needs to assess the conceptual equivalence and measurement properties of commonplace-attachment scales across sites and contexts. While calls for scale assessments have increased, few actual studies have emerged. One recent example investigated the measurement equivalence of a place-attachment scale used in multiple-forest sites and noted significant measurement differences across forest areas in Germany, Colorado, and Minnesota (Wynveen et al. 2017). This investigation built on these findings by assessing the measurement of place attachment among visitors to urban forests in Austria and the United States. Place Attachment Conceptualization While place attachment has been operationalized as the degree to which an individual values or identifies with a setting (Williams and Roggenbuck 1989), Wynveen et al. (2011) observed that place attachment measures also abstractly provide insight into the reasons why people value the setting via the theoretical and operational definitions of the multiple dimensions measured by place-attachment scales. In forest recreation research, the most frequently utilized place-attachment scale has been Williams and Roggenbuck’s (1989) two-dimensional (i.e., place identity and place dependence) measure. Place identity refers to the connections people have between a setting and their personal identity (Proshansky 1978). Moreover, place identity provides “the source of meaning for a given setting by virtue of relevant cognitive clusters that indicate what should happen in it, what the setting is supposed to be like, and how the individual and others are supposed to behave in it” (Proshansky et al. 1983, 67). Twigger-Ross and Uzell (1996) suggested place identity is a cognitive component of place attachment that can be observed as individuals use places to affirm their identity and express it to others. The other dimension in Williams and Roggenbuck’s scale is place dependence. Place dependence was first defined by Stokols and Shumaker (1981) as the “perceived strength of association between [an individual] and a specific place” (547). However, it has since been operationalized as the conative component of place attachment concerning “how well a setting serves goal achievement given an existing range of alternatives (‘how does this setting compare to others for what I like to do?’)” (Jorgensen and Stedman 2001, 234). Beyond place dependence and place identity, some researchers (e.g., Kyle et al. 2004) have suggested and found empirical support for other dimensions of place attachment. For example, Mesch and Manor (1998) suggested the dimension of social bonding (i.e., social ties to a setting that are develop though shared experiences the setting). Others (e.g., Jorgensen and Stedman 2001) suggested an affective attachment dimension (defined as an emotional bond to a place). We chose to limit this investigation to the original bidimensional conceptualization of place dependence and place identity because these have been the most commonly used and empirically verified dimensions. Implications for Forest Management In their report, Farnum et al. (2005) review several projects where place-related concepts have been used to inform several types of forest management decisions. For example, they indicated that place concepts have been used to understand stakeholders’ responses to environmental impacts, stakeholders’ support or opposition to management policies and initiatives (e.g., user fees), and conflict between stakeholder groups. Others, such as Kil et al. (2012) have observed that ideas of place could be used in conjunction with management frameworks (e.g., the recreation opportunity spectrum) to inform management of the wildland-urban interface. Furthermore, research indicates that place attachment may be able to be incorporated into more projects where understanding stakeholder perceptions are important. For example, Rickard and Stedman (2015) observed the correlation between place and communication in shaping stakeholder perceptions of protected areas and the activities they participate in at those areas. Similarly, Wynveen, Kyle, and Sutton (2010) and Wynveen, Wynveen, and Sutton (2015) have observed or argued that understanding the human-place bond can be used to help prime environmental communication and education messages designed to encourage pro-environmental behavior. Besides communication, understanding the reasons for stakeholders’ place attachment can be used in planning and managing protected areas because place “provides the context in which the problems can be recognized and articulated, and within which different values can be understood, conflicts resolved and choices made” (Potschin and Haines-Young 2013, 1054). For instance, Kil et al. (2012) suggested that place provides insight into understanding stakeholders desired benefits from protected areas, and Wynveen and Sutton (2015) and Wynveen , Kyle, and Theodori (2010) observed the role that fostering place attachment has in building trust between stakeholders and management agencies. Clearly, these few examples suggest that place attachment has a role to play in forest management. Measurement Concerns Place attachment survey items have been modified and translated for use in contexts across the globe. However, item equivalency has often been assumed in the literature, which Budruk (2010) indicated may be problematic when comparing data collected among samples from differing social and physical contexts. In her review of cross-cultural and cross-language research, Budruk identified four areas of equivalence that should be met when using a scale to measure a construct and make comparisons across contexts: semantic, functional, normative, and conceptual. Semantic equivalence involves ensuring nearly identical definitions of words or phrases when translating a scale (Behling and Law 2000). Beyond semantic equivalence is functional equivalence. Functional equivalence occurs when the meaning and intent of the items are virtually identical across samples (Kozak et al. 2004). In other words, people from different groups similarly interpret both the denotation and the connotation of survey questions. Normative equivalence (i.e., similarity in which social rules influence communication) can be difficult to achieve when social conventions discourage directly addressing a concept or cause differing operational interpretations or expressions of a concept (Behling and Law 2000). For example, some groups consider directly providing constructive criticism, even when asked, inappropriate. Lastly, conceptual equivalence is achieved when a concept can be operationalized in the same manner between populations (Groth-Marnat 2009). For instance, each group understanding that the concept of “identity” refers to how an individual conceives of or expresses themselves when responding to the question, “Who am I?” Failure to achieve these equivalences may lead to faulty interpretations of data when comparing samples from varying social (and sometimes physical) contexts. For example, reliability issues may result from translating items between languages (Harkness et al. 2010), or cultural differences may lead to different interactions with the setting resulting in varying conceptualizations of place (Kyle and Johnson 2008; Trentelman 2009). Similarly, previous research has indicated that activity type (Moore and Scott 2003) and involvement (Bricker and Kerstetter 2000) may influence individuals’ place conceptualizations. As Wynveen et al. (2017) posit, it is possible that the different experiences (i.e., variance in activity type and activity involvement) moderated by their context may influence how people interpret and respond to the place-attachment scale items. (For a further review of issues associated with the use of place-attachment scales across contexts, see Budruk [2010] and Wynveen et al. [2017]). Failure to achieve these equivalences can manifest in metric inequivalence—differing psychometric properties of a scale across populations (Groth-Marnat 2009). Hence, this project explored the measurement equivalence of place attachment between two urban forest contexts: Vienna, Austria (Lobau and Prater parks) and Minneapolis, Minnesota (Fort Snelling State Park and the adjacent Fort Snelling Minnesota Historical Site). Methods To assess the measurement equivalence of a place-attachment scale across the Minnesotan and Austrian contexts, we used identical on-site instruments and performed analysis similar to that of developing new instruments (e.g., tests of reliability and validity suggested by DeVellis [2003]). Specifically, we interpreted confirmatory factor analysis results, assessed scale reliability and validity, and conducted scale-measurement-equivalence assessments between the subsamples. The instruments and methods were designed and implemented with the approval of the University of Minnesota Institutional Review Board and fulfilled Austrian national ethical requirements. Instrument Design Data were collected via on-site instruments developed by a research team that included native English- and German-speakers familiar with the study settings. Four items, chosen based on their successful application in previous studies, measured place dependence, and three items measured place identity (see Table 1), adapted from Williams and Vaske’s (2003) two-dimensional place-attachment scale. Table 2 includes the items originally developed by Williams and Vaske. Table 1. Factor loading and descriptive results. Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — *p ≤ 0.05, ***p ≤ 0.001 a,b Means with different superscritps are significantly different c PD4 removed due to low-factor loading View Large Table 1. Factor loading and descriptive results. Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — Austria (n = 309) Minnesota (n = 422) Factor/Item Mean SD λ SE Chronbach’s Alpha Mean SD λ SE Chronbach’s Alpha t Place Dependence 3.45a 0.88 0.68 3.18b 0.97 0.89 3.93*** PD1 Doing what I do at this recration area is more important to me than doing it in any other place. 3.63 1.06 0.86 0.07 — 3.36 1.00 0.93 0.04 — — PD2 I wouldn’t substitute any other recreation area for the type of recreation I do here. 2.91 1.22 0.53 0.07 — 3.08 1.08 0.83 0.05 — — PD3 No other place can compare to this recreation area. 3.82 1.11 0.71 0.08 — 3.12 1.13 0.89 0.05 — — PD4 Doing my activities here is more important to me than doing them in any other places.c — — — — — — — — — — — Place Identity 3.94a 0.90 — — 0.80 3.80b 0.96 — — 0.94 2.09* PI1 This recreation area means a lot to me. 4.20 1.03 0.80 0.05 — 3.95 0.99 0.89 0.04 — — PI2 I identify strongly with this recraetion area. 3.53 1.12 0.63 0.06 — 3.74 1.01 0.94 0.04 — — PI3 I am very attached to this recreation area. 4.08 1.06 0.87 0.05 — 3.70 1.05 0.95 0.04 — — *p ≤ 0.05, ***p ≤ 0.001 a,b Means with different superscritps are significantly different c PD4 removed due to low-factor loading View Large Table 2. Williams and Vaske’s (2003) place-attachment scale. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. View Large Table 2. Williams and Vaske’s (2003) place-attachment scale. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. Dimension/Item Place identity I feel “X” is a part of me. “X” is very special to me. I identify strongly with “X”. I am very attached to “X”. Visiting “X” says a lot about who I am. “X” means a lot to me. Place dependence “X” is the best place for what I like to do. No other place can compare to “X”. I get more satisfaction out of visiting “X” than any other. Doing what I do at “X” is more important to me than doing it in any other place. I wouldn’t substitute any other area for doing the types of things I do at “X”. The things I do at “X” I would enjoy doing just as much at a similar site. View Large All items were measured on a five-point agreement scale (i.e., 1 = “strongly disagree,” 2 = “disagree,” 3 = “neither,” 4 = “agree,” and 5 = “strongly agree”). Respondents also answered a series of demographic questions. The researchers checked normative question equivalence between languages and the questionnaire was translated into German and back-translated to check semantic equivalence. A team member fluent in German and English also provided guidance regarding functional and conceptual equivalence. Since both sites in Austria and both sites in Minnesota were adjacent to one another, data of the two Austrian and the two Minnesota sites were pooled together to have a more robust database. Study Sites and Sampling On-site data collection occurred during the summer of 2015 (June to September) through an intercept survey at each of the parks. In Minnesota, Fort Snelling State Park is a 1,500-hectare day-use area managed by the Minnesota Department of Natural Resources located on the Mississippi River between Minneapolis and St. Paul. Over 400,000 visitors per year hike, bike, cross-county ski, and fish in its floodplain forests, meadows, marshes, and lakes. Adjacent to the state park and also bordering the Mississippi River, historic Fort Snelling, a 14-hectare state-owned parcel abuts a significant urban regional park managed by the Minneapolis Park Board. Trails in both the state park and the historical site connect nearby urban regional trail systems. Visitors can also arrive by personal vehicle and public transportation. Similarly, Vienna’s Lobau (1,037 hectares) and Prater (600 hectares) parks are located in a forested floodplain near the city center. Located in close proximity, they are almost adjacent to one another on either side of the Danube River. Both parks have extensive trails for hiking and biking. Prater also has more developed areas similar to those found in large city parks (e.g., playgounds, sport courts, and dog park). Lobau is managed by Municipal Department 49: Forestry Office and Urban Agriculture of the city of Vienna, and Prater is managed by Municipal Department 42: Parks and Gardens. As with the Minnesota sites, the Austrian parks are accessible via a network of regional trails and public transportation. Visitors were selected through a stratified-cluster sample and systematically selected sampling period that varied in time of day and day of the week to capture a diverse visitor segment. One adult per travel group was asked to complete the questionnaire. Respondents took between 10 and 20 minutes to complete the questionnaire. The length of the survey limited sample sizes to 471 in Minnesota (response rate = 39 percent) and 310 in Austria (response rate = 34 percent). Based on our observations during data collection, we believe that reasons for nonresponse had primarily to do with time constraints: visitors did not want to disrupt their activity. This was especially true of bicyclists. Nonresponse information collected in Austria, via observation of those who declined participation, indicated that most nonrespondents were bicyclists (48.9 percent), followed by hikers (25.5 percent), dog walkers (14.1 percent), joggers (10.7 percent) and Nordic walkers and horse riders (0.8 percent). Hence, compared to previous counts observed in Lobau (Arnberger 2006), bicyclists may have been under represented and dog walkers overrepresented. However, as the purpose of this investigation was not to generalize the findings to these populations of visitors but rather to identify measurement concerns with the place-attachment scale, we chose to acknowledge the limitation and proceed with the analysis. Data Analysis After calculating the descriptive statistics to document the makeup of the sample, we began the scale assessment by conducting a confirmatory factor analysis (CFA) for each subsample (Minnesotan, Austrian) using LISREL 9.2 to assess the hypothesized two-dimensional place-attachment model. Within the measurement model, correlation was allowed between the place-dependence and place-identity dimensions. The individual CFAs allowed for identification of items that should be removed due to low-factor loading onto the intended dimension or cross-loading. After determining a final item set, goodness-of-fit statistics for each sample were calculated to evaluate the place-attachment scale measurement model across both subsamples independently. Several measures of Goodness-of-fit assessed root mean square error of approximation (RMSEA), comparative fit index (CFI), and non-normed fit index (NNFI) (Byrne 1998). For the RMSEA, values ≤ 0.08 indicate acceptable fit (Steiger and Lind 1980); for the CFI (Hu and Bentler 1995), values ≥ 0.95 indicate acceptable fit; and Bentler and Bonnett (1980) designated values ≥ 0.95 as acceptable for the NNFI. Place attachment measurement was cross-validated via LISREL 9.2’s multiple-group procedure. That is, testing ensued to see if the models measured place attachment equivalently via tests of equality of factor structure (equal form) and metric invariance (equal pattern of factor loadings). The equal form test indicates whether “the number of factors and pattern of indicator-factor loadings is identical across groups,” and metric invariance indicates whether factor loadings are equivalent across samples (Brown 2006, 268). We also tested for scalar invariance—the equality of intercepts (Marsh et al. 2017). Modification indices provided by LISREL allowed hypothesizing about the origin of any differences identified. Finally, several calculations assessed the scale’s reliability and construct validity. Cronbach’s alpha calculations for both the place-dependence and place-identity dimensions for each subsample assessed reliability (i.e., internal consistency). The CFA results provided insight into discriminant validity. Comparing sample demographics and t-tests between the place identity and place-dependence dimensions’ composite scores for each subsample assessed convergent validity. Results Overall, the sample was slightly more male (53 percent, n = 398) than female, and respondents had a mean age of 43 years (sd = 14.72); however, the Austrian sample was slightly younger (mean difference = 2 years; t = 2.22; p ≤ 0.03). A majority of the Minnesotan sample (78 percent, n = 355) had earned at least a college degree, while the Austrian sample was more evenly split, with 57 percent (n = 175) having earned a college degree or higher. Differences existed among the subsamples concerning the number of previous visits and the first year they visited their respective park. The Minnesota sample had visited significantly less in the 12 months prior to the survey than the Austrian visitors (~16 days, sd = 48.15 and ~100 days, sd = 105.87, respectively; t = 12.62, p < 0.001). On average, the Austrian sample first visited their park in 1988; the Minnesotan sample first visited in 2005 (t = 14.17, p < 0.001). While each subsample cited a variety of activities as their “main activity” on the day they were surveyed, the most frequently cited activity was hiking/walking (Minnesota: 59 percent, n = 279; Austria 34 percent, n = 103). After analyzing the demographic data, the overall measurement fit of the data to the place-attachment model was assessed. Four items hypothesized to measure place dependence and three items hypothesized to measure place identity were included in the instrument distributed to visitors at each site. Initial results for pooled data indicated that one of the place dependence items, PD4—“Doing my activities here is more important to me than doing them in any other place” —should be removed due to low-factor loading (Table 1). The fit indices for the final set of items were all within an acceptable range (Austria: χ2df=7 = 15.26, RMSEA = 0.06, NNFI = 0.97, CFI = 0.99, GFI = 0.98; Minnesota: χ2df=7 = 11.78, RMSEA = 0.04, NNFI = 0.99, CFI = 0.99, GFI = 0.99), indicating that the measurement model was a good fit for the data from each sample. Scale Reliability and Convergent Validity After determining the adequacy of the measurement model via the initial CFAs, reliability and validity were assessed. The place-attachment dimensions had an acceptable internal consistency statistic (α ≥ 0.68) for place identity and place dependence across the samples. Following a suggestion by Williams and Vaske (2003), the subsamples’ descriptive statistics and a pair of t-tests designed to identify differences in mean place-dependence and place-identity scores between the subsamples (see Table 1) assessed convergent validity. As the subsamples had varying visitation frequencies and tenure lengths, past research (e.g., Hammitt et al. 2004; Williams and Vaske 2003) suggested the place-dependence and place-identity scores should also significantly differ from one another. Differences were observed for both the place-dependence and place-identity mean scores between the samples (place dependence: Austria, m = 3.45, sd = 0.88; Minnesota, m = 3.18, sd = 0.97, t = 3.93, p ≤ 0.001; place identity: Austria, m = 3.94, sd = 0.90; Minnesota, m = 3.80, sd = 0.96, t = 2.09, p ≤ 0.05). Combined, these analyses suggest that the place-attachment scale meets minimum reliability and convergent validity expectations. Scale Cross-Validation The next analysis step was to cross-validate the place-attachment scale between the two subsamples. Acceptable goodness-of-fit indices (χ2df=14 = 29.28, RMSEA = 0.05, NNFI = 0.99, CFI = 0.99, GFI = 0.99; Table 3) for the test of equal form indicated the number of factors and pattern of indicator-factor loadings were identical among the samples. However, a chi-square difference test indicated a significant difference in the pattern of factor loadings between the datasets (Δχ2df=20 = 27.80, p ≤ 0.001). To identify where the difference occurred in the factor-loading pattern, post hoc analyses utilizing the modification indices (i.e., substantive change in chi-squared values) calculated via LISREL 9.2 were used. These results suggested that while the place dependence pattern of factor loadings was equivalent between the two samples, the factor loading pattern for place identity was not. Lastly, we tested the place dependence measurement for scalar invariance between the groups. Fit indices worsened (χ2df=3 = 32.01, RMSEA = 0.15, NNFI = 0.94, CFI = 0.97, GFI = 0.93) when the intercepts were constrained equally across the groups. Post hoc analysis indicated that the intercept corresponding to item PD2, “I wouldn’t substitute any other recreation area for the type of recreation I can do here,” was the likely source of the variance. However, it is important to note that scalar invariance is an ideal not often achieved in practice by psychometric scales (Marsh et al. 2017). Table 3. CFA goodness of fit for the measurement model. Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 ***≤0.001 View Large Table 3. CFA goodness of fit for the measurement model. Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 Model χ2 Δχ2 df Δdf RMSEA NNFI CFI GFI Pooled 17.01 — 7 — 0.04 0.99 0.99 0.99 Individual models  Austria 15.26 — 7 — 0.06 0.97 0.99 0.98  Minnesota 11.78 — 7 0.04 0.99 0.99 0.99 Multiple Group Analysis  PD & PI equal form 29.28 — 14 — 0.05 0.99 0.99 0.99  PD & PI equal pattern of factor loadings 57.08 27.80*** 20 6 0.07 0.99 0.99 0.96  PD scalar invariance 32.01 — 3 — 0.15 0.94 0.97 0.93 ***≤0.001 View Large Discussion We continue to encourage the use of place attachment to aid in understanding forest stakeholders and inform decision-making process, especially when used for monitoring and making within site comparisons. However, in light of the current observations and similar results from previous studies (Budruk 2010; Wynveen et al. 2017), these results suggest place-attachment measurement varies among samples from differing contexts. One explanation may be linguistic differences, even though the current and past investigations have used standard methods of forward- and back-translation. Alternatively, several researchers have observed differences in place conceptualization among people with different experience and cultural backgrounds (e.g., Kyle and Johnson 2008, Trentelman 2009). Hence, it may be that existing place-identity items are sensitive to these differences and therefore result in the measurement variance. Future research should attempt to identify the source of the variance within each item. Possible areas to focus on may involve linguistic or behavioral experience issues. For example, concerning linguistics, there may be differences in how different groups of people interpret words and phrases when they are not clear about their exact meaning (conceptual meaning), differences in how groups have used specific words in the past (linguistic semantics), or differences in how context affects word meaning (pragmatics) (Akmajian et al. 2001). Alternatively, there could be slight variation in the way different groups experience places that have not been captured by common experience use history or activity preference variables. Future research should focus on an exhaustive list of activity preference, experience use history, mode of experience, and other variables that may influence how people experience a place, which in turn, may influence how they respond to particular place-identity measurement items. Given the current results, managers and researchers should take steps to mitigate measurement issues with the place-attachment scales, especially when making comparisons across contexts. Two recommendations can easily be implemented. First, the entire set of items hypothesized to measure place attachment should be used rather than just three or four. Using the breadth of items will allow investigators to remove items that do not statistically behave as expected. Second, as with all research, investigations should include measurement instrument pretests to identify measurement problems before data are collected. Hopefully, by following these guidelines and others regarding data collection, analysis, and interpretation, forest managers can feel confident in using data elicited from place-attachment scales to make decisions affecting recreational visitors and other stakeholders. Specifically, there are several examples in the literature where natural resource managers have used place attachment to help understand environmental concern, awareness of resource impacts, pro-environmental behavior adoption, attitudes toward fees, inter- and intra-stakeholder group conflict, and recreational visitor motivations (Eder and Arnberger 2012; Farnum et al. 2005). To use place attachment data to inform decisions about these issues and others, Williams and Stewart (1998) provide a four-step framework for managers to follow. First, identify and use place names that stakeholders use to identify spaces within the forest. Next, use these names and other place-specific language when communicating management plans to stakeholders. Third, recognize that the reasons for attachment may vary between, or even within, stakeholder groups. Lastly, plan for greater management attention among contested areas. In sum, this investigation indicated a need to assess the psychometric properties of the place-attachment scale and take appropriate action to address any abnormalities during every investigation to ensure that the data provide valid information to inform an understanding of people’s connection to place. Acknowledgements The authors would like to thank the USDA Forest Service Northern Research Station for providing funding that contributed to this project. They would also like to thank Dr. Gerard Kyle for his advice on the manuscript. Literature Cited Akmajian , A. , R.A. Demer , A.K. Farmer , and R.M. Harnish . 2001 . Linguistics: An introduction to language and communication . MIT Press , Cambridge, MA . Altman , I. and S. Low . 1992 . Place attachment, human behavior, and environment: Advances in theory and research Vol. 12 . Plenum Press , New York, NY . Arnberger , A . 2006 . Recreation use of urban forests: an inter-area comparison . Urban For. Urban Green . 4 ( 3–4 ): 135 – 144 Google Scholar CrossRef Search ADS Behling , O. , and K.S. Law . 2000 . Translating questionnaires and other research instruments: Problems and solutions . Sage , Thousand Oaks, CA . Google Scholar CrossRef Search ADS Bentler , P. , and D. Bonett . 1980 . Significance tests and goodness of fit in the analysis of covariance structures . Psychol. Bull . 88 : 588 – 606 . Google Scholar CrossRef Search ADS Bricker , K.S. , and D. Kerstetter . 2000 . Level of specialization and place attachment: an exploratory study of whitewater recreationists . Leis. Sci . 11 : 233 – 257 . Brown , T.A . 2006 . Confirmatory factor analysis for applied research . Guilford Press , New York, NY . Budruk , M . 2010 . Cross-cultural equivalency of the place attachment scale: a multigroup confirmatory factor analysis approach . J. Leisure Res . 42 ( 1 ): 25 – 42 . Google Scholar CrossRef Search ADS DeVellis , R.F . 2003 . Scale development: Theory and applications . Sage , Thousand Oaks, CA . Eder , R. , and A. Arnberger . 2012 . The influence of place attachment and experience use history on perceived depreciative visitor behavior and crowding in an urban national park . Environ. Manage . 50 ( 4 ): 566 – 580 . Google Scholar CrossRef Search ADS PubMed Farnum , J. , T. Hall , and L.E. Kruger . 2005 . Sense of place in natural resource recreation and tourism: an evaluation and assessment of research findings . General technical report PNW-GTR-660: US Department of Agriculture, Forest Service, Pacific Northwest Research Station . Groth-Marnat , G . 2009 . Handbook of psychological assessment . Wiley , Hoboken, NJ . Hammitt , W.E. , E.A. Backlund , and R.D. Bixler . 2004 . Experience use history, place bonding, and resource substitution of trout anglers during recreation engagements . J. Leisure Res . 36 : 356 – 378 . Google Scholar CrossRef Search ADS Harkness , J.A. , A. Villar , and B. Edwards . 2010 . Translation, adaptation, and design . P. 115 – 140 , in Survey methods in multinational, multiregional, and multicultural contexts , Harkness , J.A. , M. Braun , B. Edwards , T.P. Johnson , L. Lyberg , P. Mohler , B. Pennell , and T.W. Smith (eds). Wiley , Hoboken, NJ . Google Scholar CrossRef Search ADS Hu , L. , and P.M. Bentler . 1995 . Evaluating model fit . P. 76 – 99 , in Structural equation modeling: concepts issues and applications , R.H. Hoyle (ed.). Sage , Thousand Oaks, CA . Jorgensen , B.S. , and R. Stedman . 2001 . Sense of place as an attitude: lakeshore property owners’ attitudes toward their properties . J. Environ. Psychol . 21 : 233 – 248 . Google Scholar CrossRef Search ADS Kil , N. , T.V. Stein , S.M. Holland , and D.H. Anderson . 2012 . Understanding place meanings in planning and managing the wildland–urban interface: the case of Florida trail hikers . Landscape Urban Plan . 107 ( 4 ): 370 – 379 . Google Scholar CrossRef Search ADS Kozak , M. , E. Bigné , and L. Andreu . 2004 . Limitations of cross-cultural customer satisfaction research and recommending alternative methods . J. Qual. Assur. Hosp. Tour . 4 ( 3–4 ): 37 – 59 . Google Scholar CrossRef Search ADS Kyle , G.T. , and C.Y. Johnson . 2008 . Understanding cultural variation in place meaning . P. 109 – 134 , in Understanding concepts of place in recreation research and management , Kruger , L.E. , T.E. Hall , and M.C. Stiefel (eds). USDA For. Serv. Gen. Tech. Rep, Portland, OR . PNW-GTR-744. Kyle , G.T. , A.J. Mowen , and M. Tarrant . 2004 . Linking place preferences with place meaning: an examination of the relationship between place motivation and place attachment . J. Environ. Psychol . 24 : 439 – 454 . Google Scholar CrossRef Search ADS Marsh , H.W. , J. Guo , P.D. Parker , B. Nagengast , T. Asparouhov , B. Muthén , and T. Dicke . 2017 . What to do when scalar invariance fails: the extended alignment method for multi-group factor analysis comparison of latent means across many groups . Psychol. Methods . doi: 10.1037/met0000113 Mesch , G. , and O. Manor . 1998 . Social ties, environmental perception, and local attachment . Environ. Behav . 30 : 504 – 519 . Google Scholar CrossRef Search ADS Moore , R.L. , and D. Scott . 2003 . Place attachment and context: comparing a park and a trail within . For. Sci . 49 ( 6 ): 877 – 884 . Morgan , M. and B. Messenger . 2009 . Using an activity and place-based typology to explain visitor motivations . J. For . 107 ( 1 ): 44 – 49 . Potschin , M. , and R. Haines-Young . 2013 . Landscapes, sustainability and the place-based analysis of ecosystem services . Landsc. Ecol . 28 ( 6 ): 1053 – 1065 . Google Scholar CrossRef Search ADS Proshansky , H.M . 1978 . The city and self-identity . Environ. Behav . 10 ( 2 ): 147 – 169 . Google Scholar CrossRef Search ADS Proshansky , H.M. , A.K. Fabian , and R. Kaminoff . 1983 . Place identity: physical world socialization of the self . J. Environ. Psychol . 3 ( 1 ): 57 – 83 . Google Scholar CrossRef Search ADS Rickard , L.N. , and R.C. Stedman . 2015 . From ranger talks to radio stations: the role of communication in sense of place . J. Leisure Res . 47 ( 1 ): 15 – 33 . Google Scholar CrossRef Search ADS Raadik Cottrell , J. and S.P. Cottrell . 2015 . Sense-of-place influences on perceived environmental change effects on future holiday experiences to Saaremaa, Estonia . Scand. J. Hosp. Tour . 15 ( 4 ): 425 – 446 . Google Scholar CrossRef Search ADS Steiger , J.H. , and J.C. Lind . 1980 . Statistically-based tests for the number of common factors . Paper presented at the Annual Spring Meeting of the Psychometric Society , May, Iowa City, IA . Stewart , W.P. , D.R. Williams , and L. Kruger , eds. 2013 . Place-based conservation: Perspectives from the social sciences . Springer , Dordrecht, NL . Google Scholar CrossRef Search ADS Stokols , D. , and S.A. Shumaker . 1981 . People and places: a transactional view of settings . P. 441 – 488 , in Cognition, social behavior and the environment , J. Harvey (ed.). Erlbaum , Mahwah, NJ . Trentelman , C.K . 2009 . Place attachment and community attachment: a primer grounded in the lived experience of a community sociologist . Soc. Nat. Resour . 22 ( 3 ): 191 – 210 . Google Scholar CrossRef Search ADS Twigger-Ross , C.L. , and D.L. Uzzell . 1996 . Place and identity processes . J. Environ. Psychol . 16 : 205 – 220 . Google Scholar CrossRef Search ADS Williams , D.R. , and J.W. Roggenbuck . 1989 . Measuring place attachment: some preliminary results . P. 32 , in Abstracts: 1989 Leisure Research Symposium , eds. McAvoy , L.H. and D. Howard , National Recreation and Park Association , Arlington, VA . Williams , D.R. , and S.I. Stewart . 1998 . Sense of place: an elusive concept that is finding a home in ecosystem management . J. For . 96 ( 5 ): 18 – 23 . Williams , D.R. , and J.J. Vaske . 2003 . The measurement of place attachment: validity and generalizability of a psychometric approach . For. Sci . 49 ( 6 ): 830 – 840 . Wynveen , C.J. , G.T. Kyle , J.D. Absher , and G.L. Theodori . 2011 . The meanings associated with varying degrees of attachment to a natural landscape . J. Leisure Res . 43 ( 2 ): 290 – 311 . Google Scholar CrossRef Search ADS Wynveen , C.J. , G.T. Kyle , and S.G. Sutton . 2010 . Place meanings ascribed to marine settings: the case of the Great Barrier Reef Marine Park . Leisure Sci . 32 ( 3 ): 270 – 287 . Google Scholar CrossRef Search ADS Wynveen , C.J. , G.T. Kyle , and G.L. Theodori . 2010 . Place bonding and trust: the case of feral hog management surrounding Big Thicket National Preserve . J. Rural Soc. Sci . 25 ( 2 ): 58 – 80 . Wynveen , C.J. , I.E. Schneider , S. Cottrell , A. Arnberger , A.C. Schlueter , and E. Von Ruschkowski . 2017 . Comparing the validity and reliability of place attachment across cultures . Soc. Nat. Resour . doi: 10.1080/08941920.2017.1295499 Wynveen , C.J. , and S.G. Sutton . 2015 . Engaging the public in climate change-related pro-environmental behaviors to protect coral reefs: the role of public trust in the management agency . Mar. Policy . 53 : 131 – 140 . Google Scholar CrossRef Search ADS Wynveen , C.J. , B.J. Wynveen , and S.G. sutton . 2015 . Applying the value-belief-norm theory to marine contexts: implications for encouraging pro-environmental behavior . Coastal Manage . 43 ( 1 ): 84 – 103 . Google Scholar CrossRef Search ADS © 2018 Society of American Foresters This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of ForestryOxford University Press

Published: May 24, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off