Laserfiche WebLink
<br />24 <br /> <br />potential use values (option values) were beyoud tl1e scope of tilis researcb. However, tile task force did discuss <br />tl1ese otIier components of public benefits. The economic researcb to measure tilese values is evolving and <br />taking on growing importance in tlIe formulation of environmental policy. (See for example, Desvousges, et <br />al., Mitcbell and Carson, and NOAA.) The reliability and validity of these measures is sllown to be enbanced by <br />application to a particular policy so tl1at people can respood to a survey referendum in an incentive-compatible <br />fasllion. Because tilere was no specific policy cboice for respondents to "vote on" in Iltis case, and given tI1at <br />Iltis was a baseline study, emphasis was placed on use value for tlIis researcb. <br /> <br />Furthermore, as a baseline study tile metilodology employed here does not utilize some of tile more sopbisticated <br />recreational survey fonnats, in part due to concern about demands placed on respondents and to pursue a <br />representative profile. Due to the intensity of questionnaire formats wbich measure "experiences" and relate to <br />"Benefits-Based Management (BBM)", tiley are typically implemented tI1roogh a screening process wbere <br />potential respondents can self-select tlIemselves for participation. Again given tlIe emphasis on establisbing a <br />baseline for recreation, portions of tlIe upper eud of tile informational bierarcby, such as BBM, were deemed to <br />be beyond tile scope of tlIis study. Witilin tlIe limits of tilis study, an effort was made to collect tlIe most <br />comprehensive information possible, including the demand for activities and settings, and user satisfaction. The <br />resulting survey design is discussed below in section V and the questionnaire is sbown in Appendix B. <br /> <br />Sampling Design <br /> <br />As mentioned above, tile sample was comprised of four sub- samples. This design was deemed to be most <br />appropriate and cost-effective for gatilering tile higbest quality data available given the constraints uf the research <br />project. The sub-samples enabled the collection of data that woold be botll comparable and distinguisbable <br />between locals, visitors, property owners wbo use tile river and property owners wbo do not <br /> <br />A. Interviews Of River Users <br /> <br />Interviews of river users enconntered on tile river were deemed essential for tile research objectives. To establish <br />a representative baseline of river uses, tilis approacb wonld yield the most reliable and comprehensive results. In <br />the absence of information regarding tile proportions of uses by property owners versus non-owners, uses on tile <br />water versus uses from tile river bank, bridges, etc. and the relative importance of public accesses versus private <br />or undeveloped accesses, on-the-water interviews served as tile fonndation of the sampling design. <br /> <br />As mentioned above tile various segments of tile river were broken down into 15 stretches. (See Appendix A.) <br />Tbe stretcbes were delineated based on logical terminus points, sucb as dams, lakes, public accesses, etc. to tile <br />degnee possible, along witll considerations for tile amount of water tI1at interviewers could cover during an <br />interviewing day. For tlIe most part, tilese stretches were from 20 to 30 miles long. Wben water conditions <br />allowed, tile interviewers covered a stretcb by motoring slowly upstream to maximize tlIe cbance of intercepting <br />river users wbo were stationary or travelling downstream. <br /> <br />The sampling plan for determining whicb stretches would be covered on whicb days was based on information <br />on patterns and timing of use provided by experts (agency staff, etc.) combined witil general infonnation on <br />water use patterns by day of tile week, week and montil. This information was used to construct an a priori <br />probability distribution regarding wbere and when the beaviest use would occur during tile four peak montils, <br />June tIirougb September. Advice from experts was used exclusively to set tile sampling plan during tile sbonlder <br />montlls of May, October and November. Wbile tilis was an imperfect metllod, random sampling from tilis a <br />priori probability distribution enabled tile construction of a sampling plan wliicb would be rougbl y proportiunal <br />to the patterns of use. That is to say, the likelihood of being sampled increased witil tlIe expected level of use by <br />stretch, day of the week, week, and montll. As seen in Appendix A. the busiest stretcbes were sampled <br />proportionally more often and weekends were sampled more tilan weekdays. Daily logs were kept by <br />interviewers to determine if use patterns reflected the a priori expectations. Extenuating circumstances sucb as <br />variation in the wealller, etc. were also recorded. <br /> <br />American River Management Society <br />