Big data spatial analysis of campers’ landscape preferences: Examining demand for amenities
Introduction
Understanding what drives demand for outdoor recreation within protected areas and how recreationists make decisions about recreational opportunities are queries that have generated a significant body of theory and literature in park and recreation research (Loomis and Walsh, 1997). Of late, this area of research has received considerable interest to address management issues emerging from the increased use of recreational facilities in U.S. national parks (i.e., Richardson et al., 2017; Sessions et al., 2016; Timmons, 2019). However, most of the previous research relies upon stated choice measures of demand. That is, recreationists’ actual behavior is not observed. Instead, stated responses are collected regarding hypothetical scenarios concerning the conditions of the ecological, managerial, or social environments of park settings (Tietenberg and Lewis, 2015).
It is thus the aim of this study to analyze revealed preference data for recreational facilities that have experienced record demand in recent years—campgrounds (Rice et al., 2019). This research strives to understand how recreationists reach decisions on the selection of campsites and what aspects of the recreational setting drive this demand. Specifically, we aim to understand what managerial, social, and ecological aspects of the setting most influence decision-making prior to visiting a campground. As a case study, we examine the campsites of Watchman Campground in Zion National Park, USA using reservation data extracted from the Recreation Information Database (RIDB). If successful, similar site-specific analyses will enable park managers to be more informed in future alteration of campsite distribution, access to recreational amenities, and the management of campsite allocation based on relative demand.
This research seeks to build on an already extensive body of literature concerning recreation decision-making by incorporating revealed-preference data from campsite reservations to examine what aspects of the setting are most impactful in driving demand for particular campsites. In building from previous research, this study thus relies upon existing means of modeling and measuring recreation decision-making and camping recreation, as a whole.
Driver and Brown (1978) provided an early framework for how recreationists make decisions concerning their participation in recreation activities through their Recreation Opportunity Demand Hierarchy. This framework posits that recreationists' awareness of their demand for the various components of the recreational experience decreases across a hierarchy of demand that flows from activity to setting to motivations to benefits (Driver and Brown, 1978). Important to the present study—and consistent with other recreation decision-making models—the authors’ conceptualization of setting is divided between the managerial, social, and ecological. Krumpe and McLaughlin (1982) presented another early model—the Recreation Decision Model—in which decisions are reached via an elimination process of recreation options. This model was later applied by Brunson and Shelby (1990) to demonstrate how campsites are chosen. The authors presented three stages of decision-making in which options are systematically eliminated by recreationists over three stages—based on their necessity, experience, and amenity attributes (Brunson and Shelby, 1990).
Additionally, valuable models for recreation decision-making have been proposed in the wake of Krumpe and McLaughlin (1982). Clark and Downing (1985) provided an alternative framework in which decisions are reached based on substitutability. More recently, Park et al. (2017) presented an extended structural model of goal-directed recreation choices. While these models were developed specifically to examine the decision-making of recreationists, it is important to note that other models and theories were simultaneously developed and have been applied within recreation research, such as the Theory of Reasoned Action (Fishbein and Ajzen, 1975) and the Theory of Planned Behavior (Ajzen, 1991).
No matter the chosen model, measurement must follow. There are two overarching methods of measuring recreationist demand for various factors that influence decision-making: stated and revealed preference measures (Adamowicz et al., 1994). Principally, stated preference methods rely on survey data, while revealed preference methods rely on actual observed behaviors (Kaval and Baskaran, 2013). For each overarching method, a variety of specific methodologies exist.
Of the stated preference methodologies, contingent valuation and choice modeling are most common in eliciting amenity values (Kaval and Baskaran, 2013). Contingent valuation involves asking recreationists what they would be willing-to-pay, or willing-to-accept, to have a particular experience or enact some change in their experience (Kaval and Baskaran, 2013). A depth of protected area and outdoor recreation management research has utilized this valuation method, as evidenced through reviews of the literature (e.g., Brander and Koetse, 2011). Recreation research employing choice modeling has been similarly prolific (e.g., Newman et al., 2005; Pettebone et al., 2011). Choice modeling requires that recreationists be “faced with a variety of alternatives and may be asked to select their most preferred alternative from a choice set (choice experiment), group their preferences (contingent grouping), rate their preferences (contingent rating), or rank their preferences (contingent ranking)” (Kaval and Baskaran, 2013, p. 32).
No matter the stated preference method employed, a variety of assumptions and limitations potentially impair reliability. As reported by Bartkowski and Lienhoop (2018), three core assumptions underlie stated preference research: 1) respondents hold full information concerning the good or service of interest, 2) respondents are self-interested (largely disregarding the interests of others), and 3) respondents hold pre-defined preferences. Overlapping these assumptions is the larger assumption of rationality and the presumption that respondents can translate their preferences into monetary or otherwise stated terms (Bartkowski and Lienhoop, 2018). Given these assumptions, revealed preferences—when available—are usually preferred to stated preferences (Cameron, 2011).
In contrast with stated preferences, revealed preference methodologies are based on observations of actual behavior (Kaval and Baskaran, 2013). In protected area management and outdoor recreation research, a number of revealed preference methodologies are common, including the travel cost method (e.g. Amoako-Tuffour and Martínez-Espiñeira, 2012), hedonic pricing (e.g. Nelson, 2010), and big data analysis (e.g. Keeler et al., 2015).
In recent years, revealed preference measures of recreation behavior using big data have become increasingly popular (see Tenkanen et al., 2017). Big data is broadly defined as a large volume of information that is created in real-time in a variety of formats such as text, audio, and video (Gandomi and Haider, 2015). Given significant computational power, high value can be extracted from these large datasets (Gandomi and Haider, 2015). In outdoor recreation contexts, big data has been used to address a number of management issues (see Pickering et al., 2018). Rice et al. (2020) similarly used protected area data to examine how park conservation status, ownership, and access impacts the provisioning of ecosystem services such as noise abatement and exclusion. Moreover, using 13,600 Instagram and Flickr photographs, Hausmann et al. (2018) were able to identify recreationist biodiversity preferences. Other studies have used big data specifically to measure outdoor recreation demand. Using Twitter and Flickr posts, Donahue et al. (2018) were able to access demand for urban parks. Utilizing data from tens of thousands of campground reservations, Rice et al. (2019) forecasted demand for national park campgrounds to allow managers to allocate resources more effectively. Moving forward, this method of data collection is projected to grow as a means of informing visitor use management (Pickering et al., 2018).
In addition to big data, emerging spatial analysis methods have also been applied to examine recreation demand. Regression-based spatial analysis has the ability to harness the power of large datasets to create predictive models, while accounting for spatial autocorrelation (Chi and Zhu, 2019). Spatial autocorrelation is a phenomenon described by Tobler (1970) in which near items are likely to more similar than faraway items. Lee and Schuett (2014) used geographically-weighed regression to assess what socio-economic attributes predict national park visitation among Texas (U.S.) residents. Rice and Pan (2021) used a similar model to assess factors influencing changes in park visitation during the COVID-19 pandemic in the Western U.S.
Specific to camping, a variety of aspatial models and measures of decision-making have been put forward over the past forty years. Stated preference methods include semi-structured interviews (Lime, 1971), relevance-determinance analysis of campsite amenities (Mikulić et al., 2017), exploratory factor analysis of campsite amenities (Gursoy and Chen, 2012), choice modeling of campsite setting attributes (Oh et al., 2007), and multinomial logit modeling of camper characteristics (McFarlane, 2004) and campsite setting (Stewart et al., 2003). A number of studies have also examined specific setting characteristics or amenities and their importance to the camping experience or campsite selection, including campfires (Lillywhite et al., 2013), price (Bamford et al., 1988), shading from forest overstory (James and Cordell, 1970), ecological impacts (White et al., 2001), and proximity to other campers (Twight et al., 1981). However, spatial and revealed preferences methods have yet to be employed to examine campers’ observed behavior in choosing a campsite. Therefore, this study seeks to harness a big dataset of campsite choice behavior to spatially examine what aspects of the managerial, social, and ecological setting are most influential in decision-making among campers in Zion National Park.
Section snippets
Study overview and purpose
We selected Watchman Campground in Utah's Zion National Park as the study area for this analysis, due to its increasing demand (see Timmons, 2018) and superior data availability and consistency. To examine how various amenities and costs influence demand, we gathered data for various components of the campground setting. Each of the setting attributes was grouped within a larger setting category derived from the facets of recreational settings presented by Driver and Brown (1978). To this end,
Study site
Zion National Park (Zion) is located in Southwestern Utah, U.S (Fig. 1). Over-crowding in Zion has been well-documented in recent years, leading to visitor use issues including ecological impacts, strains on infrastructure, and safety concerns (Timmons, 2019). Watchman Campground is the largest campground in Zion, boasting 179 campsites. It sits near the park's southern entrance, just north of the town of Springdale. It contains access to the following amenities for all campers: an interpretive
Results
Descriptive statistics for the independent variables are listed in Table 1. The descriptive results of the OLS model and the SLM are listed in Table 3. The SLM exhibits a superior R2 and lower AIC and BIC statistics, and therefore provides a better fitting model. Table 4 contains full results of the SLM. The multicollinearity condition number (23.850) does not indicate unacceptable levels of multicollinearity within the model (Chi and Zhu, 2019; Dormann et al., 2013). However, this number does
Examining significant predictors of demand
Overall, it appears that the characteristics of a given campsite are more predictive of its demand than the characteristics of its surroundings. Price, which also serves as an indicator of private access to electricity, appears strongly predictive of demand. Similarly, the designation of a campsite as walk-in was strongly predictive of demand, albeit negatively. The positive relationship between price and average booking window indicates that campers are more than willing to pay the $10 premium
Conclusion
This research sought to understand how recreationists make decisions on the selection of campsites and what aspects of the recreational setting drive demand in this decision-making process through an examination of revealed preference data using spatial regression. Specifically, we examined which managerial, social, and ecological aspects of the setting influence demand for campsites in Zion's Watchman Campground using reservation data from the RIDB. Results indicate that price, access to
Credit author statement
William L. Rice: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization. Soyoung Park: Formal analysis, Investigation, Writing - Original Draft, Writing - Review & Editing.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (67)
- et al.
Combining revealed and stated preference methods for valuing environmental amenities
J. Environ. Econ. Manag.
(1994) - et al.
The choice of forest site for recreation: a revealed preference analysis using spatial data
Ecosyst. Serv.
(2018) The theory of planned behavior
Organ. Behav. Hum. Dec.
(1991)- et al.
Leisure and the net opportunity cost of travel time in recreation demand analysis: an application to Gros Morne National Park
J. Appl. Econ.
(2012) - et al.
Beyond rationality, towards reasonableness: enriching the theoretical foundation of deliberative monetary valuation
Ecol. Econ.
(2018) - et al.
The value of urban open space: meta-analyses of contingent valuation and hedonic pricing results
J. Environ. Manag.
(2011) - et al.
In-migration to remote rural regions: the relative impacts of natural amenities and land developability
Landsc. Urban Plann.
(2013) - et al.
Using social media to understand drivers of urban park visitation in the Twin Cities, MN
Landsc. Urban Plann.
(2018) - et al.
Beyond the hype: big data concepts, methods, and analytics
Int. J. Inf. Manag.
(2015) Resort camping in America
Ann. Tourism Res.
(1990)
Exploring spatial variations in the relationships between residents' recreation demand and associated factors: a case study in Texas
Appl. Geogr.
Campsite choice and the camping tourism experience: investigating decisive campsite attributes using relevance-determinance analysis
Tourism Manag.
Iowa wetlands outdoor recreation visitors' decision-making process: an extended model of goal-directed behavior
J. Outdoor Recreat. Tour.
An investigation of perceptions of social equity and price acceptability judgments for campers in the U.S. national forest
Tourism Manag.
Estimating visitors' travel mode choices along the bear lake road in rocky mountain national park
J. Transport Geogr.
Current knowledge and future research directions for the monitoring and management of visitors in recreational and protected areas
J. Outdoor Recreat. Tour.
Protected areas and noise abatement: a spatial approach
Landsc. Urban Plann.
Understanding changes in park visitation during the COVID-19 pandemic: A spatial application of big data
Wellbeing, Space, and Society
Forecasting campground demand in US national parks
Ann. Tourism Res.
Mapping recreation as an ecosystem service: considering scale, interregional differences and the influence of physical attributes
Landsc. Urban Plann.
Measuring recreational visitation at U.S. National Parks with crowd-sourced photographs
J. Environ. Manag.
Boater preferences for beach characteristics downstream from Glen Canyon Dam, Arizona
J. Environ. Manag.
Differential campsite pricing: an experiment
J. Leisure Res.
Exploring spatial data analysis techniques using R: the case of observations with no neighbours
Camp 4 Daily Lottery. Recreation.Gov
A hierarchy of campsite attributes in dispersed recreation settings
Leisure Sci.
Foreword
Spatial Regression Models for the Social Sciences
Why here and not there: the conditional nature of recreation choice
Recreation.gov Is Your Gateway to Explore America’s Outdoor Destinations and National Treasures
Collinearity: a review of methods to deal with it and a simulation study evaluating their performance
Ecography
The opportunity spectrum concept and behavioural information in outdoor recreation resource supply inventories: a rationale
Belief, Attitude, Intention, and Behavior: an Introduction to Theory and Research
Cited by (7)
What is “big data” and how should we use it? The role of large datasets, secondary data, and associated analysis techniques in outdoor recreation research
2023, Journal of Outdoor Recreation and TourismNature dependent tourism – Combining big data and local knowledge
2023, Journal of Environmental ManagementIntegrating social values with GPS tracks through Denali National Park and Preserve
2023, Applied GeographyWeather and recreational vehicle camping businesses
2022, Annals of Tourism Research Empirical InsightsCitation Excerpt :The businesses offer a range of comparable amenities and services that are of particular interest to RV campers including: (1) utilities (e.g., water, sewer, electric, cable, laundry, and WiFi), (2) services (e.g., general stores, restaurants), (3) recreation (e.g., swimming pools, dog parks, close proximity to national parks), and (4) extended pull-through sites that can accommodate larger RVs. Extended and innovative offerings are reinventing what is expected while RV camping, and contribute to traveler willingness to pay higher accommodation rates at for-profit campgrounds than non-profit campgrounds such as those at United States National Parks (Rice & Park, 2021; Walls, Wichman, & Ankney, 2018). The Moab and Shepherdsville camping businesses represent the Southwest and Central climate regions (Karl & Koss, 1984).
Explorations of preferred and maximum booking windows among U.S. national park campers: Implications for improved fairness
2024, Tourism and Hospitality Research