Laserfiche WebLink
HYPOTHESIS TESTING APPROACHES <br />For formal hypothesis testing relative to Phase Ill Bond Release the following procedures were followed: <br />Cover and Production <br />Phase III Revegetation success for cover and production will be demonstrated by one of following <br />statistical approaches. <br />If sample adequacy has been achieved (minimum of 15 samples) in BRB -5 and the reference areas, and <br />the reclaimed area mean total allowable herbaceous cover and mean allowable herbaceous production <br />are above 90% of the reference area - derived standard, All criteria are met for a direct comparison and <br />revegetation would be deemed successful (CDMG 2010 proposed rules revisions 4.15.1 (2)(a)). <br />Were the mean allowable herbaceous cover and /or production of the reclaimed area above 90% of the <br />reference area - derived standard, and sample adequacy was met for the reference areas but not for BRB - <br />5 (minimum of 30 samples, technical standards not subject to sample adequacy), then the hypothesis of <br />reclamation success would be tested using a parametric test of the "reverse null" hypothesis. This would <br />be conducted using the following expression (CDMG 2010 proposed rules revisions, 4.15.11 (2)(c) and as <br />set forth in McDonald et al. 2003): <br />x —Q <br />t _ <br />S- <br />x <br />Where: x = Bond Release Block Sample Mean <br />Q = 90% of Standard <br />S- = standard error of mean [s/ <br />] <br />S = sample standard deviation <br />n = sample size <br />t = calculated t value <br />t = Table t value (alpha = 0.2) <br />The (reverse) null hypothesis being tested would be that the bond release block mean (t) was less than <br />or equal to 90% of the standard, stated as H µ _< Q. If t was greater than the 1- tailed t table value for <br />alpha error probability of 0.20, with (n -1) degrees of freedom then H would be rejected, and revegetation <br />would be deemed successful. <br />Under the previous scenario if sample adequacy was not met for the reference areas nor the BRB -5 <br />(minimum of 30 samples), then the hypothesis of reclamation success would be tested using a two - <br />sample t -test of Reverse Null Hypothesis with Satterthwaite's Adjustment (CDMG 2010 proposed rules <br />6 <br />