Laserfiche WebLink
1252 <br />nm = ()2 <br />where: <br />n. = minimum sample size <br />s 2 = Sample variance (n-1 degrees of freedom) <br />d = Precision (0.10 for cover and production; 0.15 for woody <br />plant density) <br />x = Sample arithmetic mean <br />t = The (a=.10) t -table value for a 1 tail t-test, n-1 degrees of <br />freedom (infinite degrees of freedom may be used if n > 30) <br />n = Sample size <br />(b) If the reclaimed area sample mean is less than 90 percent of the relevant standard, a one <br />sided t-test with alpha error probability of 0.10 shall be used to demonstrate that the reclaimed <br />area mean is not less than 90 percent of the applicable success standard with 90 percent <br />statistical confidence. If the basic assumption, or null hypothesis, of equality between the <br />reclaimed area and 90 percent of the standard is not rejected by the test, then reclamation is <br />considered successful for the parameter tested. This is the "standard null " hypothesis approach <br />to the t-test. Sample size adequacy must be demonstrated using the formula in (a) above. A <br />minimum of 15 sample observations shall be taken even if statistical sample adequacy is <br />achieved with fewer observations. <br />The general form of the "standard null" t-test is: <br />Page 38 May 2016 (TR -75) <br />