Laserfiche WebLink
demonstrate success, and for practical purposes the test demonstrates that the reclaimed area mean exceeds <br />the standard by an extremely wide margin, at the 90% level of statistical confidence. Unfortunately, Rule <br />4.15.11(2)(c) requires a minimum sample size of 30, if the reverse null demonstration is employed, and a <br />sample size of only 20 was obtained. <br />An alternative approach that Cedar Creek could have employed would be to use a direct comparison if <br />sample adequacy is obtained in the reclaimed area. Cedar Creek's results indicate that sample adequacy <br />was not obtained (Nmin of 24 indicated on Table 16). However, it appears that Cedar Creek used a <br />precision level of 0.10 in the sample adequacy formula, and in addition, they used a table "t" value for a <br />two-tailed test rather than the value for cone-tailed test. Both the precision level and the table t value used <br />by Cedar Creek were more stringent that required by regulation, and had the effect of inflating the required <br />sample size. Rule 4.15.11(2)(a) allows for use of the t-tail value for aone-tailed test, and also allows for <br />use of a precision level of 0. l5 in the sample adequacy formula, specifically for woody plant density. <br />Using the formula with a precision level of 0. l5, and table "t" value for aone-tailed test results in a <br />minimum required sample size of 6, rather than 24 (see spreadsheet calculations attached, "2006 Blue <br />Ribbon Riparian Woody Plant Density"). <br />Table 16 of the report will need to be revised, using the less stringent table "t" and 0.15 precision <br />level allowed by 4.15.11(2)(a), and success should be demonstrated by direct comparison of the <br />sample mean to (90% of) the approved standard. Summary Table 1 will need to be revised <br />accordingly. Section 3.5 narrative on page 16 of the report should be amended to reference the <br />direct comparison success demonstration, based on comparison of the sample mean of a statistically <br />adequate sample to the standard. For 2007 sampling, the operator will need to ensure that a <br />minimum sample size of 30 is obtained (if reverse null t-test is employed), or that sample adequacy <br />is demonstrated (with minimum sample size of 15), if direct comparison of sample mean to standard <br />is to be employed. <br />[f you have any questions, please call me. <br />Sincerely, <br />J seph J. Dash <br />nvironmental Protection Svecialist <br />Attachment <br />c:\word\blueribbon/yr9vegreportadeq I <br />