2011 Supporting_Statement_Part_B 0596-0208 12 13 2010

Part B - Supporting Statement for OMB 0596-0208

Trends in Use and Users in the Boundary Waters Canoe Area Wilderness, Minnesota

2011

Supporting Statement for OMB 0596-0208

Trends in Use and Users in the Boundary Waters Canoe Area Wilderness, Minnesota

2011

NOTE: This request is for extension of OMB 0596-0208 for an additional two years to complete a small subset of the original survey questions. The data will be used to sample and monitor use patterns as input to quota system evaluation model for the Boundary Waters Canoe Area Wilderness.

B. Collections of Information Employing Statistical Methods

1. Describe (including a numerical estimate) the potential respondent universe and any sam­pling or other respondent selection method to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corre­sponding sample are to be provided in tabular form for the uni­verse as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

The following plan provides the details of sampling for the Boundary Waters Canoe Area Wilderness social science research to be conducted by the U.S. Forest Service Aldo Leopold Wilderness Research Institute, with cooperators at the University of Montana in 2011 & 2012. This design is based on previous studies conducted in 1969, 1991 and 2007, as well as by current knowledge about distribution of recreation use in the Boundary Water Canoe Area Wilderness, state-of-the-art methods, and input from study cooperators.

The population of interest for the trend/change/management study includes current adult visitors (> 15 years old) to the Boundary Water Canoe Area Wilderness during the peak season of May to September 30. Total visitation per year is estimated at over 200,000; while modeling of the permit data suggest that at least 130,000 permitted (both allocated and self-issue) day and multi-day visits occur during the peak period. The population of interest in 1969 and 1991 included only overnight visitors during the peak season, while in 2007 it included all permitted recreation visitors, including self-permitted day users (permits are available for self-issue at launch points for day use, non-motorized). The sample of visitors will be sub-divided according to the type of trip they were on when contacted for this study (either day use or overnight use), and a separate survey instrument will be developed for each of these trip types. Sufficiently large samples of day and overnight users will be required for each of the survey instruments.

The current front-end information will come from permit receipts, completed before the trip and includes several questions about trip characteristics of the party. This form provides limited ability to test for non-response bias, but it is easy to administer and is completed by only one member of the group.

In 1969, visitors were contacted on-site as they finished their Boundary Water Canoe Area Wilderness trip and asked to either complete a questionnaire at that time or provide contact information for later mailing of a questionnaire. In 1991, visitors were contacted on-site as they began their trip and asked to complete a short on-site interview to collect the information on the front-end form for later mailing of a questionnaire.

In 1991, approximately 400 people were intercepted at entry points to the Boundary Water Canoe Area Wilderness and at permit distribution locations. They were asked to provide their contact information for a mail-back questionnaire. Contacts were made on-site at the busiest entry points as visitors began their trips, and low use sites were targeted through the central distribution locations. The sampling was partitioned by sample day, with a different location chosen for each sample day. There were 36 sampling days that were determined according to how they were distributed in 1969 – 18 weekday/ 18 weekend distributed during specific weeks across the peak season. The entry locations were distributed across sampling days to roughly correspond with their estimated distribution of use. This intercept method, using a mail-back questionnaire, obtained a 74 percent response rate. There are two mail back questionnaires – one for overnight visitors and another for day visitors (with only some mention of camping removed). In 2007, visitor contacts followed the same pattern as in 1991.

2. Describe the procedures for the collection of information including:

• Statistical methodology for stratification and sample selection,

• Estimation procedure,

• Degree of accuracy needed for the pur­pose described in the justification,

• Unusual problems requiring specialized sampling procedures, and

• Any use of periodic (less frequent than annual) data collection cycles to reduce burden.

The proposed sampling design for the current study is based on an example laid out in 1991 and used again in 2007. Visitor population estimates by entry point and type of use have been made using the most recent self-issue permit data and allocated permit data. This model of the population distribution was used to develop a sampling schedule that includes interviews with parties at the busiest 17 entry/exit points that account for more than 70 percent of use by the population of interest for the trend study. The overall sampling goal is to obtain two representative samples from relatively large populations of visitors – (1) overnight users and (2) day users. It is generally desirable to obtain a sample of at least 250 from a large population to provide the appropriate power for statistical analysis, and assuming a 75 percent response rate, this requires 666 visitor intercepts for the two samples.

Allocated permits are picked up by group leaders or their designees on the day before or the day of the trip. The schedule calendar is constructed to provide two independent sample schedules at the 17 primary sampling points with each occurring randomly across 25 percent of the days in the peak season, for a total of 76 sample days, to reach the target of 666 total contacts.

The following table shows the 17 entry points that will be sampled along with estimates of their types and levels of use during the peak season.

Analysis of quantitative data will begin with descriptive statistics to display the current responses from visitors. For selected variables, parametric and nonparametric tests of comparability for categories of subjects, such as party leaders and party members or for outfitted and non-outfitted, will be presented and discussed.

To accomplish the stated objectives of determining trends, the data is subjected to a series of comparative analyses. The major question to be addressed is whether or not there are differences in user travel behavior, or visitor response to conditions encountered. We will be interested in understanding day use visitors and how visitors at low use entry points differ in travel patterns from those at high use entry points. We will examine those variables that exhibit some degree of change. Through cross-tabulations by selected independent variables such as mode of travel, length of stay, experience, and socio-demographic characteristics.

Specific hypotheses can be tested for each item in the survey. The following hypotheses are the focus of continued monitoring:

• Hypothesis 1: The percentage distribution of visitors across the primary methods of wilderness travel and location is not different across additional study years. Variables to be tested: Study year and method of travel. Appropriate analysis method: Chi-square.

• Hypothesis 2: The amount of experience the visitors have at the study area is not different for the study years. Variables to be tested: Independent variable – study year; Dependent variable – number of previous visits to the site. Appropriate analysis method: ANOVA with adjustment for non-normality if needed.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sam­pling, a special justification must be provid­ed for any collection that will not yield "reli­able" data that can be generalized to the universe studied.

A multi-stage cluster sampling design has been suggested for this type of social research. The primary sampling unit is actually blocks of time (essentially visitors to the area during that block of time). Before the blocks of time are selected, a stratification scheme is employed to define weekend clusters (Friday through Sunday) and weekday clusters (Monday through Thursday). The first stage cluster sample draws random clusters from each strata per month of sampling. This is 2 weekend and 2 weekday clusters chosen randomly from the possible clusters for each month. In a 7-month use period, the sample is 28 total clusters, 7 pairs of weekdays and 7 pairs of weekends. The second stage of this sampling procedure is to select smaller clusters within each of the 14 pairs of clusters. Visitors to a specific trailhead on a particular day are a subdivision of a cluster of days.

In the selected sampling method, clusters are chosen through simple random sampling, but a ratio estimator is used as a measure of central tendency. Ratio estimators are quotients of two variables, each of which varies randomly from cluster to cluster. Ratio estimation is considered to be an efficient technique. The ratio estimator equals the sum of the cluster totals divided by the sum of the cluster sizes, where the sums range over all clusters in the sample. Ratio estimators may be biased and variances can only be approximated. However, the degree of bias is usually negligible for sample sizes likely to be encountered in practice. The ratio estimator is consistent. As is the ratio estimator of the population mean, this estimation is biased. The bias of the estimated variance is inversely proportional to the sample size, n, and a serious problem only for small sample sizes. Jaeger provides a method of approximation of the bias of the ratio estimator: the estimator is inversely proportional to the number of clusters sampled. The ratio estimator is unbiased if the mean per element within clusters is uncorrelated with sample size.

Response to the mailback questionnaire is expected to be high. It is not uncommon for 70 - 80 percent of a sample of visitors contacted through visitor permit receipts to agree to participate in a study. There are some who will not return the mail-back questionnaire. It is believed that the primary reason that some do not mail the questionnaire back is due to a belief that since visitors may not participate in recreation very often at that particular place, their opinions may not be very important. Follow-up mailings are used to convince them otherwise. Past response rate examples for similar surveys include the Boundary Waters Canoe Area Wilderness (74 percent response), Shining Rock Wilderness (75 percent response), Desolation Wilderness (83 percent response), and Gates of the Arctic National Park and Preserve (95 percent response).

Don A. Dillman, of Washington State University, published a book entitled Mail and Internet Surveys: The Tailored Design Method in 2000, which precisely documents the appropriate ways to assure high response rates in mail-back surveys in social research. Dillman’s methods have been used in many dispersed recreation visitor studies and have produced consistently high response rates. Dillman provided guidelines for writing initial and subsequent cover letters in which a justification of the information collection effort appears along with an appeal for response based upon the importance of each individual sampled to respond for a larger population of people represented. Following this approach, there would typically be an initial mailing of information, a postcard reminder, and two follow-up mailings of the questionnaire and appropriate cover letter.

Whether or not this minimum response rate of 70 percent is obtained using these methods, permit date for respondents and non-respondents will be compared. Enough basic information is being collected from all people to help us understand whether the respondents and non-respondents differ to a significant degree on basic demographic factors and area visitation patterns.

The chief statistical consultant for this study will be David Turner, Station Statistician, Rocky Mountain Research Station, 860 North 1200 East, Logan, UT 84321 (801) 755-3560. All sampling and surveys will conform to guidelines established by Watson, Cole, Turner, and Reynolds (2000).

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of tests may be submitted for approval separate­ly or in combination with the main collection of information.

The test instrument is largely a replication of a portion of the survey used in 2007. Some Boundary Waters Canoe Area Wilderness visitors have already reviewed and completed the draft survey, and an additional 10 pilot test respondents will be identified at the beginning of the use season and will be asked to complete the survey on-site to allow analysis prior to actual initiating mailings to the larger sample.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Robert G. Dvorak, Ph.D.

Department of Recreation, Parks, and Leisure Services Administration

Finch 106A

Central Michigan University

Mount Pleasant, MI 48859

Dr. William T. Borrie

College of Forestry & Conservation

32 Campus Drive

University of Montana

Missoula, MT 59812

Dr. Neal Christensen

Christensen Research

1626 S. 6th Street W.

Missoula, MT 59801

Carolyn Swan

Mathematical Statistician

USDA/NASS

Statistics Division/Methods Branch

Phone: (202) 690-8639