| PRIVACY ACT MATERIAL REMOVE: is to sum the arithmetic means, are summing distributions reliably, and also, no matter what the distribution is, that you can sum the variances; equal jis sum the variance of the that to the sum of the individual Once we have done that, we then have an arithmetic mean and an variances. arithmetic standard deviation. If we know that we can, indeed, calculate back to a geometric mean and a geometric standard deviation to provide a distribution of values. The next viewgraph these parameters. (LRA-35) shows the relationships between all of This is how we calculate We've already seen that one. 10 an arithmetic standard deviation where a variance is shown here, if we know ll the geometric standard deviation and also the geometric mean. 12 two 13 deviation, 14 standard deviation. 15 Dunning and Schwarz (published) in Health Physics. show how, Well, 16 if we know can how we the an arithmetic mean, calculate the or standard arithmetic an mean geometric And these and geometric the These results, by the way, are taken from a paper by next four viewgraphs (LRA-36, -37, -38, show -39) was an _ (LRA-36) the 17 results of doing all of these calculations. 18 individual who had melanoma. 19 dose on the skin directly from the Los Alamos calculations. 20 the probability distribution. 21 percent confident that his dose was equal to or less than 590 rads, and so 22 forth. For his case we are looking now at the beta The most likely dose is 310 rads. The next viewgraph (LRA-37), 23 This indicates We are 90 » he had the doses for — We have assumed that the organ of interest is the whole 24 Hodgkin's-disease. 25 body. 26 and 27 in utero exposure from external dose. 28 a fairly typical result that the dose from ingestion is much smaller, say, This indicates now that -- in his case we have an in-utero exposure, this is not the one that Yook 44 is largely This comes from Los Alamos. We see Ng calculated, but this