Is Alaska Really Different? A Review of CUSTOMER Recreation Visitor Survey Data1

Many believe that Alaska is unique and that its location, resources, and population influence the use patterns and attitudes of its National Forest recreation visitors so that they seem notably different from visitors to other National Forests outside Alaska. Data from a recreation visitor survey called CUSTOMER were analyzed for the years 1991 to 1993 to identify signs of differences between recreation visitors to the Chugach National Forest and other selected National Forests outside Alaska. Although some significant differences do appear, a definitive conclusion may not be drawn from the existing CUSTOMER data. Alaska is popularly regarded as different from other States of the union, a view that encompasses beliefs about its geography, natural resources, cultural heritage, population, and lifestyle. Many assert that such characteristics influence or create different use patterns and attitudes among its recreation visitors (State of Alaska 1993, International Tourism and Resort Advisors 1993). In lieu of comprehensive and substantive data to the contrary, this basic assertion may have influenced the management plans and the activities of State and Federal recreation managers in Alaska, including those of the Chugach National Forest (CNF). But is outdoor recreation in CNF really different, and more importantly for its recreation managers, are recreation visitors there really different from those of National Forests outside Alaska? Addressing the latter issue is important to better meet the needs and desires of recreation visitors to CNF. Clearly, CNF is distinguishable from other forests in the National Forest. Two hundred miles across and the size of Massachusetts and Rhode Island combined, the 5.6 million-acre CNF is second in size only to Alaska’s other National Forest, the Tongass. CNF is bounded to the north by the rock and ice of the Chugach Mountains and to the south by the 3,500 mile coastline of fjords and islands of the Prince William Sound. Despite its size, the vast majority of CNF is accessible only by small aircraft, boat, or foot. The natural and cultural resources of CNF seem unmatched elsewhere. Only 500 miles below the Arctic Circle, the climate supports dozens of active glaciers. Probably more wolves, bears, and bald eagles can be found in CNF than in any National Forest outside Alaska. Prince William Sound is itself a haven for a rich and diverse marine life, including several species of whale, sea lions, otters, as well An abbreviated version of this paper was presented at the Second Symposium on Social Aspects and Recreation Research, February 23-25, 1994, San Diego, California. Social Scientist, USDA Forest Service, Chugach National Forest, 3301 C. Street Suite 300, Anchorage, AK 99501. as its renowned salmon and shellfish. The 700,000 acre Copper River Delta in the eastern half of CNF is the largest contiguous wetland ecosystem on the west coast of the Nation and the yearly destination for millions of migratory birds. CNF is also the aboriginal and contemporary home of populations of Eskimos and Indians, whose ancestors have inhabited the coastal areas for more than 10,000 years. Recreation visitors to CNF can choose from a full spectrum of recreation opportunities. Developed facilities range from the modern Begich, Boggs Visitor Center (the most visited tourism site in Alaska) to developed overnight campgrounds equipped with flush toilets. Those seeking more primitive recreation experiences can choose from backpacking in de facto wilderness to cross-country skiing and sea kayaking. Still others may choose to see glaciers and wildlife from automobiles, aircraft, ships, or on foot. Thus, CNF does indeed have many special and unique characteristics that may influence its recreational visitors. However, whether the physical and social settings of CNF are different from those of other National Forests outside Alaska is a moot point. The more relevant issue is whether the settings influence the attitudes and behavior of recreational visitors so that they are different from those of visitors to other National Forests outside Alaska. Unfortunately, research has not been done that definitively establishes a cause and effect relationship. Perhaps the best that can be done at this time is to sort through existing data for clues to the following question: do recreation visitors to CNF exhibit characteristics, preferences, reasons for participating, and levels of satisfaction significantly different from those of National Forest visitors elsewhere? This paper addresses these questions through an examination of one recent recreation visitor survey applied to CNF and other sites outside Alaska. CUSTOMER Recreation Visitor Survey To examine the question of “difference,” both relevant and commensurate information about the recreation visitors to CNF and other sites outside Alaska is needed. Although location-specific visitor and tourism surveys have been administered for a number of years across the Nation, only one survey using a consistent questionnaire and sampling procedure—the Customer Use Survey Techniques for Oper­ ation, Management, Evaluation and Research (CUSTOMER) recreation visitor survey--has been applied to National Forests and other locations both within and outside Alaska. Developed and implemented by the USDA Forest Service’s Southeastern Forest Experiment Station, CUSTOMER was designed to provide a range of general USDA Forest Service Gen. Tech. Rep. PSW-156. 1995. 79 and site-specific recreation visitor data useful in the management of individual National Forests as well as in the preparation of the Forest Service’s national Renewable Forest and Rangeland Resources Planning Act (RPA) for the analysis of recreation and wilderness (English and others 1993; Cordell and others 1990). By compiling the results of individual National Forests and other locations, CUSTOMER may also be used to develop a basic national database on visitor demographics, trip profiles, reasons for choosing a site, activity participation, and visitor importance and satisfaction with general site attributes. Between 1990 and 1993, CUSTOMER was implemented on a U.S. Office of Management and Budget-approved pilot test basis in over 35 “sites” (defined for purposes of this paper as a unique combination of an administrative unit and a recreation season) across the Nation. More than 19,850 recreation visitors nationwide were interviewed on-site and also asked to complete one or more additional mailback surveys. Three CNF sites were sampled in 1991 and 1992. With more than 2,800 recreation visitors interviewed during summer and winter seasons, CNF accounted for about 14 percent of the national total. Because of its scope, consistency, and the number of interviews completed, CUSTOMER should be an ideal data source for comparing recreation visitors to Alaskan and sites outside Alaska, including sites administered by the Forest Service, USDI Bureau of Land Management, and USDI National Park Service. Nevertheless, several problems arise from the CUSTOMER sampling method, and thus efforts to make robust comparisons among individual sites are complicated. First, the sample of national sites was actually self-selecting because participation in CUSTOMER was either administratively mandated or voluntary. As a consequence, the sites did not represent a true random sample of National Forest visitors nationwide. Second, each site was responsible for defining the activity groups of interest. Thus, a description of “developed camping” may actually differ considerably according to a remote or poorly funded site, in comparison to an urbanized and well-funded site’s description of “developed camping.” Another problem was the weighting of the collected data. Most sites had a poor idea of the actual amount of annual visitor use, as well as the relative proportions of the various component activity groups. Therefore, the basis to properly weight data is low for developing summary means and frequencies for any single site and for all sites combined. Because of the relatively high cost of on-site survey sampling, small sample sizes--especially from mailback surveys--were inadequate for reliable statistical analysis for many of the activity groups. Last, but certainly not least, compilations of CUSTOMER data are not readily accessible. An independent working database, limited to summary means and frequencies contained in individual CUSTOMER site reports, was therefore created for this paper. The summary nature of the database severely limits extensive data analysis. Comparison Methodology Despite these difficulties, CUSTOMER data are used as the basis for comparing CNF recreation visitors with those of other sites outside Alaska in the attempt to detect visitor differences. Specific activity groups, representing the visitors’ stated primary recreation activity during a visit to a site, are chosen as the focus of comparison. Some preliminary manipulation of the data was necessary, however. The original self-reported activity groupings in the individual CUSTOMER survey reports are re-classified in the working database into groups that are assumed to be the most closely analogous in actuality. This reclassification results in a total of 27 possible activity groups. Of these, it is found that 19 cannot be used for comparison for one of the reasons: (1) a particular activity group has no representation in both CNF and non-Alaskan sites; (2) the total sample size for an activity group is determined to be too small for purposes of analysis; or (3) the activity group could only be classified as miscellaneous. This reduced the available overall sample sizes by 35 percent for CNF and by 47 percent for the non-Alaskan sites. The eight activity groups suitable for comparison are: • Angling� • Developed overnight use� • Dispersed day use� • Motorized boating� • Nonmotorized boating� • Roaded sightseeing� • Trail use� • Visitor information service (VIS) activities� The working database of summary CUSTOMER data contains more than 200 separate variable means and frequencies which describe the use patterns and attitudes of recreation visitors. Attempting to compare CNF and non-Alaskan recreation visitors across all of these variables is beyond the scope of this paper. Thus, five of the most relevant subsets of the larger dataset are chosen as the basis of comparison. The five data subsets include: • Selected demographic characteristics (age and income). • Selected visitation characteristics (primary vs. secondary destination; repeat vs. first-time visitation; in-state vs. out-of-state residence). • Reasons for choosing site (scenic beauty; good facilities; prior knowledge; convenient location; personal reasons; crowding at other areas; seeing new attractions; traveling with a group; trying a new site; and other). • Importance of general site attributes (reasonable fees; location of site; barrier-free accessibility; quality of scenery; cleanliness of facilities; clear directional signs; good roads and parking; helpfulness of personnel; information on site history; nearby shopping and supplies; presence of agency personnel; safety and security; site maps and information; and trip planning information). • Satisfaction with general site attributes (same as importance attributes). 80 USDA Forest Service Gen. Tech. Rep. PSW-156. 1995. Measurement of difference within the 5 data subsets for the 8 activity groups is dependent upon the nature of the data being compared. At least three measures of difference are possible given the data limitations: • Range of data subset variable means and frequencies. • Rank correlation of data subset variable means and frequencies. • Overall mean of data subset variable means and frequencies. The first and simplest measure of difference involves examining the data subset variables for CNF recreation visitors to find means and frequencies that are outside the range of means and frequencies in the other sites outside Alaska. Thus, if data subset variable means and frequencies for CNF recreation visitors are either higher or lower than the highest or lowest mean or frequency recorded among non-Alaskan sites, it is considered an indication of a possibly important difference between the Alaskan and non-Alaskan sites. This measure of difference is applied throughout the comparisons because it does not rely upon any statistical assumptions and can be used with the categorical data in the first two data subsets (demographic and visitation characteristics). A second measure of difference involves the rank ordering of data subset variables rated by CNF recreation visitors to determine if the order differs from the order of visitors to non-Alaskan sites. This second measure of difference is usable on lists of items that can be rank-ordered, including the third, fourth and fifth data subsets (reasons for choosing site, site attribute importance, and satisfaction with site attributes). Within each data subset, variables are rank-ordered for both CNF and non-Alaskan sites and compared to see if there is a positive correlation in the rankings. Because the data under analysis were nonparametric in nature, the Friedman two-way analysis of variance is chosen to test whether the rank-orders agreed (Wilkinson 1990). A third measure of difference considers whether CNF visitors rated general site attributes higher or lower than visitors to the non-Alaskan sites. This measure is applied to the fourth and fifth data subsets (importance of, and satisfaction with, general site attributes). A t-test is applied to the variable means of data subsets for CNF and non-Alaskan sites in order to determine if visitors rated the variables higher or lower. Assumptions different than those used in the selection of the activity groups, data subsets, and measures of difference used in this analysis could produce different results and conclusions in the comparisons.