Laserfiche WebLink
-9- <br />5. Confidence levels need to be provided for all comparisons made betv,een <br />reference areas and affected areas. <br />10 Vegetation Baseline Information - Appendix H, 2H <br />1. None of the figures 3.1-1, 3.2.1-1, 3.2.2-1, 3.2.3-1, 3.2.4-1 is present <br />in Appendix H. These should be provided. <br />2, l•Jhy was the reference area for mountain - shrubland (dry phase) selected <br />after analysis of the affected area for only one parameter (cover)? <br />3. Snowmass' formula for the students t-test is inappropriate. Sokal and Rolhf <br />(1969) state that ~•,hen sample sizes are unequal but large enough so that lnl - <br />(nl - 1) is a very small value the formula Snowmass has presented becomes <br />adequate representation. Snowmass needs to reevaluate their statistical com- <br />parisons utilizing a formula such as (Zar, 1974): <br />t= X1 - X~ <br />S_ _ <br />xl x2 <br />Where S- - = S2 + S2 <br />X1-x2 ~- ~ <br />and SP = <br />and nl -sample size for sample 1 <br />n2 -sample size for sample 2 <br />S12 - variance for sample 1 <br />S22 - variance for sample 2 <br />X - mean value for sample 1 <br />1 <br />X - mean value for sample 2 <br />2 <br />(nl - 1) S12 + (n2 - 1) S2 <br />(nl - 1) + (n2 - 1) <br />5. In their sample adequacy formula, Snowmass should utilize the z value for <br />a two-tailed test at the 80`,6 confidence level (1.282) in the case of v,oody <br />vegetative communities (1.646, in the case of herbaceous communities). <br />6. Snowmass needs to discuss the appropriateness of selecting "nearest" indi- <br />viduals rather than plotting an additional transect relative to their sampling <br />methodology. <br />7. The utilization of one square meter plots for sampling production in the <br />mountain shrubland (dry phase) community is not an appropriate technique as is <br />evidenced by the number of samples containing no vegetation. The presence of <br />no vegetation in a large number of the samples increases variance greatly, <br />consequently increasing minimum sample size required for sample adequacy. <br />Snowmass must either resample this community type utilizing an appropriate <br />