Response to comment received from Outdoor Alliance

02/26/2013

0596-0110 response to Outdoor Alliance NVUM FRN comment.

This letter comes from a set of recreation interest groups, and presents several concerns with the NVUM program and method. The concerns will be restated and discussed below.

1. The NVUM system should a) collect a more representative visitor use data sample, b) more precisely categorize the data, and c) provide guidance on how to analyze the data in the relevant context.

The sample structure for the NVUM system is designed to distribute the times and locations of field data collection in proportion with the overall patterns of visitation for each forest. That is, sampling occurs primarily at the places where visitors exit from the forest; the times and locations selected are distributed through both spatially and temporally for where and when visitors are exiting. In this way, the structure yields an unbiased picture of recreation visits. Wider representation of the visit population in any one sample year is likely only through an expanded sample size.

Categories used to describe visit characteristics, including activity, age, race/ethnicity, and others are as detailed as possible given the limits of constraints such as budgets, public burden, average survey time. At present the average duration for a complete recreation survey is about 12 minutes. Adding more detailed categories would increase that time, and likely lead to incomplete responses as visitors ‘time out’. One goal is to have a nationally consistent survey form, which often precludes customized categories that would be of greater interest to individual forest units.

Agency guidance on using NVUM data is adjusting to conform with application of the 2012 Planning Rule. At present, NVUM is suggested as one of several recreation data sources for examining activity participation on National Forest System lands.

2. Concerns about the validity of the method and the assumptions used. The groups specify concerns with representation of human-powered activities, sampling locations/times, the need to incorporate near-term weather and resource conditions to better reflect recreation patterns. In addition, the groups suggest adding questions on an experiential approach to understanding recreation visitor motivations.

The NVUM sample of locations and days is a stratified random sample where the strata are defined by types of sites and expected exiting volumes of visitation. The exiting volumes are provided by forest staff who are knowledgeable about overall visitation patterns to their forest. For the set of activities the groups are concerned with, the sampling locations are typically at entry/exit portals to the undeveloped portion of the forest. Most often these portals are where visitors to could be exiting the forest on roads and trails. For each portal, every day of the year that forest staff expects some volume of recreation traffic is included in the population to be sampled. Consequently, the traffic exiting from these portals includes a complete array of activity participants, and all of these users are potentially included in the sample. In order to ensure that a representative sample of visitors is obtained, the algorithm used to select the field schedule distributes sample locations for each stratum across seasons and areas of the forest in approximately the same proportion as the population of location/day combinations to be sampled. It may be that a larger sample size would alleviate the concerns the groups have with the sampling approach.

Setting the NVUM sampling frame is done according to expected visitation patterns for typical years. We recognize that day-to-day recreation patterns are undoubtedly influenced by short term conditions of resources, weather, and other factors. In fact, that sort of variation highlights the degree of uncertainty inherent in managing recreation resources, and one of the key challenges to accurately estimating the volume and character of recreation visits. However, it is not logistically possible from a

Adding a set of experiential questions may be of value to forest managers. Review of the list of activities as well as other questions is done prior to the start of each national survey cycle. Our goal is to have a consistent survey instrument for an entire 5-year data collection cycle. The current national cycle for NVUM ends with FY2014; the next national cycle begins in FY2015. During this collection period, we will be evaluating the existing activity list and considering other adjustments to the survey. Including a set of experiential questions into the survey will be evaluated, but would not be incorporated into survey instruments until 2015.

3. Suggestions to incorporate social media from interest groups into sampling plans.

It is not clear what the sampling frame for these would be nor how we could incorporate these results into the overall NVUM framework. It seems that for completeness, it would be necessary to incorporate all social media outlets from all activity or interest groups. It appears that the volume of data or responses would be a function of number and size of social media networks. These may or may not be directly related to visit volume for any given forest. Given the negative relationships between age and use of social media, results from these outlets could greatly underrepresent older users of National Forests. However, social media have the potential to provide valuable information. But rather that attempt to incorporate them into the NVUM process, it seems that another use would be to suggest that forest recreation staff be aware of these and consult with them alternative data sources to augment the data and results from the NVUM program.

4. Add disclaimers and caveats about the results and their limitations, and provide guidance for locating other sources of visitation and activity participation data.

This is not an issue with the NVUM method, but rather with how field units choose to employ the results. However, agency guidance does recommend use of both NVUM results as well as results from other sources, such as SCORP documents, in recreation planning.

5. The groups suggest additional methods to the intercept survey to augment the data collection – websites of OA member organizations, and partnering with such outdoor organizations to obtain data on forest visitation.

The current NVUM approach ensures that each person leaving the forest, regardless of where they went, what activity(ies) they participated in or how long they stayed has a chance of being surveyed. Moreover, the likelihood of being surveyed is not dependent on any of those aspects of their visit. It is not likely that same result could be achieved with the methods proposed. Nor is it clear right now how to incorporate the data obtained through such methods into the NVUM approach, because the sampling frame is so different. Neither is it clear whether any significant biases would result, or how they could be minimized. However, we are beginning to discuss how such contact methods might be used to provide limits to some of the NVUM results.