Laserfiche WebLink
Where: <br />x = Bond Release Block Sample Mean <br />Q = 90 % of Standard <br />S- = Standard error of mean [ s / ,%fn- ] <br />S = Sample standard deviation <br />n = Sample size <br />tc = Calculated t value <br />t, = Table t value <br />The null hypothesis being tested is that the bond release block mean (u) is greater than or equal <br />to 90% of the standard, stated as Ho : u > Q. <br />If c t is less than or equal to the ]-tailed t table value for alpha error probability of .10, at (n- <br />1) degrees offreedom (infinite degrees offreedom may be used if n>30), then Ho is not <br />rejected, and revegetation is deemed successful for the parameter tested. <br />(c) If the reclaimed area sample mean is greater than 90 percent of the standard, and sample <br />adequacy is not demonstrated using the formula in (a), success may be demonstrated by use of <br />the "reverse null" hypothesis. Under the reverse null approach, a one sided t-test with alpha <br />error probability of 0.20, is used to demonstrate that the reclaimed area mean is greater than 90 <br />percent of the relevant success standard with 80 percent statistical confidence. The basic <br />assumption, or null hypothesis, is that the reclaimed area mean is less than or equal to 90 <br />percent of the standard. If the null hypothesis of equality is rejected by the test, then reclamation <br />is considered successful for the parameter tested. A minimum sample size of thirty is required. <br />The general form of the "reverse null " t-test is: <br />Page 39 May 2016 (TR -75) <br />