REPOS!ITONY PN NL . COLLECTION Marsha | DOCUMENT DOES NOT CONTAIN ECI eg . Reviewed byEX BAS yy, Date SL20sA BOX No. FOLDER tIslands IG FS le Battelle Pacific Northwest Laboratories : Chewetak Banelle Boulevard Richland, Washington 99352 19 7 a. Telephone(509) Telex 36921 August 28, 1972 409771 Dr. Don Hendricks Director of Radiological Operations Division Nevada Operations Office USAEC Box 14100 Las Vegas, Nevada 89114 Dear Don: This letter is to sunmarize our understanding of the sampling aspects Of the forthcoming Eniwetok survey. I assume there will be some further discussion of this, and this letter is intended to serve as a basis for such discussions. I have had several long talks with Oliver Lynch by telephone, and Dick Gilbert spent a day with him last week. We very much appreciate his cooperative spirit and have been much impressed with his understanding of the problems and thorough-going approach to the survey. Probably Ollie will not agree with all of our suggestions, since he has to implement them in the field. Having had occasion to modify my own survey designs in the field, I am aware of the kinds of things that come up. However, we do need to try to spel] out some of the consequences of particular modifications where this is possible. Mostly these have to do with what random sampling amounts to, and the effects of deviations from a truly random sample. There is ample evidence that a representative set of sampling locations cannot be selected by "eyebail". There are quite a few practical- minded people who don't believe that statement, but we've got enough to do at the moment that I don't want to spend any time on that old argument. Nor can we say exactly what the consequences of departures from random sampling scheiwes wil] be. We will suggest some possibilities, and perhaps it is sufficient to note that if the results of this study have to be somehow defended in "public" they had better be based on a defensible scheme. The balance of this letter will deal with some specific topics. Some of these we haven't looked ini in the-detail we feel advisable, but time is pressing on. In what follows we assume that there are reasons for wanting to make specific statements either about specific parts of an island, a single island, or about groups of islands. Our purpose is to try to provide some