Laserfiche WebLink
<br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br /> <br />~rts decided the type of data that should be 'USE'Jd for analysis. nus <br />approach ~ used instecrl of discussing each individual study to insure that a <br />standard set of cri teria was applied to all datasets, and to reduce the anuunt <br />of time that it w:ro.ld take to discuss each study in the 23-year span of data. <br /> <br />Results. consensus of the five species e~rts in workshop #1 resulted in <br />a total of 291 data partitions that were the basis for data ana.lyses tor <br />p:>ssible category 2 curves. The eJ<PE!rts also identified 48 partitions for <br />category 1 curve developnent (See Species Dendrograms, Figures B-1 through B-3 <br />and Table B-1 in this Draft Report). A summa.ry of the catagory 1 and 2 curves <br />identified at ~rkshop #1 is presented in Table B-2. <br /> <br />A list of 21 itans were identified during W=lrkshop #1 as missing informa- <br />tion fran FWS/DATABASE. A list of these i terns was pre];Bred for the workshop <br />}Bl'ticipants, and assignments were ma:1e to secure missing data and correct <br />inconsistencies in the database. <br /> <br />Objective 2: Stratify, Pool, Analvze Data <br /> <br />Methods. In most cases, the five e~rts reached consensus on the type of <br />data that should be included in analyses for possible c:urve developnent. Where <br />the ~rts were nncertain, BID/WEST was asked to 'test' or 'decision/test' the <br />data. If a 'test' ~ specified, the data trere to be pooled if the signifi- <br />cance level did not exceed 95% (alpha = 0.05), b.1t the partitions \Ere to be <br />retained if significantly diffel"'ent. HOTrEVer, if the number of sanples (fish <br />wi th associated habitat measuranents) in one or both partitions of data was <br />less than 20, the partitions trere autanatically pooled ewer a 'test I . No <br />pooling was to be performed under a I decision/test I, b.1t the statistics were to <br />be reported in the Draft Report as a tool for the deCision-making process <br />during Workshop #2. <br /> <br />Results. The 291 data pa.rtitions identified by the e~rts in Workshop #1 <br />trere tested by BID/WEST in the manner specified, and pooled or partitioned <br />according to the predetenninad test guidelines. The results of those tests and <br />the resul tarrt preliminary SI curves are presented in the following section of <br />this Draft Report. <br /> <br />1] <br />