Laserfiche WebLink
~.. - f/^~_ yiaTE CF [C G43."')C c,: ..:.e• D '. a~.v --vc~,... <br />r. ~ DEPAR7 :'ENT Or ':+iUR;I F.ESOUACES <br />ya• T D. \tonu Pauoe, E.rcut rvr O•r ecror <br />r ~ <br />~-~a <br />• x !,7^ ;~ e23 Centennial Building. 7373 Snerman Street <br />- ~-~ ~ ~' ; 'r "~ Denver, Colorado 80203 7e1. (303) 839-3567 <br />'~ .~.~' <br />David C. Shelton <br />July ?7, 1981 Director <br />TO: Brian Munson <br />FROM: M.S. Savage SpN^~~ <br />RE: R-G Vegetation/Sampling Adequacy <br />D. Duba's Correspondence of 7/17/81 and 7/20/81 <br />File No. C-003-80 <br />Brian, <br />Dave Duba has submitted information addressing the specific levels of confidence <br />achieved during vegetative sampling for each community and parameter. In <br />addition, he has submitted an F-test analysis for parameters which were found <br />to be below acceptable confidence Zevels. Mr. Duba's (and Mr. Botts) argument <br />is that if an F-test determines that the variances are not significantly <br />different between two samples, one can utilize a students t-test to determine <br />• if there is a significant difference between the means of the samples. It is <br />my contention (as well as that of Larry Larson, staff statistician for OSM) that <br />a positive F-test determination does not necessarily indicate that a students <br />t-test may be justified. The F-test tests the variances of two samples and <br />determines if the samples are from the same population, we do not wish this <br />information. Rather, we wish to determine if the sample means of two different <br />populations are si~ificantly different from each other. Additionally, we <br />must have accurate estimates of each population parameter. By analogy: If <br />carpenters built your home utilizing measuring devices to the nearest 1 meter, <br />would ane be justified in trying to estimate the area of the floor plan in <br />square centimeters? <br />I have determined, with tha aid of the materials just received, that B-G <br />should re-sample the following crommunity parameters: <br />A. Aspen community <br />Z. Reference area productivity and woody plant Qensity~• <br />2. Pre-mine productivity and woody plant density <br />B. Sagebrush community <br />1. Pre-mine productivity, rnver, and woody plant densttg <br />• 2. Reference area cover <br />C. Oak cor:mrunity <br />1. Pre-mine woody plant density <br />~ ?. Reference area woody plane density~• <br />