Laserfiche WebLink
<br />t= w t + w t ) <br />wl + wz ) <br />Where: <br />wl - SZiini <br />w2 = s2z/n2 <br />nl = number of samples in sample population 1 <br />n2 = number of samples in sample population 2 <br />si = sample variance, sample population 1 <br />s~ = sample variance, sample population 2 <br />tl = tabular value for (nl - 1) degrees of freedom <br />t2 = tabular value for (n2 - 1) degrees of freedom <br />The t statistic, t', was calculated according to the <br />following formula, (Snedecor and Cochran 1967): <br />t~ _ (X1 - x2 ) <br />(s~/ni + s2/nz~ <br />Where: <br />xl = mean of sample population 1 <br />xZ = mean of sample population 2 <br /> <br />The null hypothesis tested by this comparison was that there <br />was no difference between the population means for the <br />disturbance area and reference area sample. <br />The same statistical procedures were also used to test the <br />null hypothesis that there was no difference between the <br />population means for exclosed and unexclosed cover plots. <br />This analysis was required because the maximum number of <br />exclosures put out (25) in each type was insufficient to <br />meet sampling adequacy for canopy cover in the Aspen, <br />Sagebrush, and Mixed brush types. Therefore, additional <br />17 <br />