IX « MJMBER OF TREES REqUIBED IN THE CHECK SCALE (BASIS FOR FIGURE 3) 



nn 

 n 



^ 2 - 2 r 



C 



AFTTTF 



m 



where 



m 

 n 



ratio of number of trees in check scale 

 to number of trees in first sample 



Xo 



USE OF RAKDOM MJlffiERS 



The use of random numbers is the quickest and best method of select- 

 ing sample trees at random. Pages of random numbers can be found 

 in "Converting Factors and Tables of Equivalents Used in Forestry", 

 U. S. Dept, Agr. Misc. Pub. 225, and in many statistical books. 



A random starting point is selected on a page of the random numbers, 

 and either one or two columns of digits are used depending upon the 

 sampling rate,, If the rate is one in three, four, or up to tenj then 

 only one column is required. The digits are merely noted in consecu- 

 tive order, and those that are inappropriate discarded. For example, 

 if a rate of one in five is in use, the digits 0, 6, 7, 8, and 9 

 would be disregarded » As numbers from one to five are encountered, 

 the appropriate numbers on the sample tree measurement tally sheet 

 would be encircled o The first number might be 4, so number 4 in the 

 first group of five numbers vjould be encircled » The next number 

 might be 1, so number 6 (the "1" in the next series of five) would 

 be encircled o In this example, only 20 of the sample tree designation 

 numbers in each section of the tally sheet v;ould be used and the 

 remaining 20 crossed outo 



If the sampling rate is one in ten, one column of digits can be used 

 if zero is considered IC For sampling rates less than one in ten, 

 two columns of digits will be needed. For example, with a rate of 

 one in 20, all numbers from 01 to 20 would be used vdiile numbers 

 00, and 21 to 99 would be disregarded. 



-47- 



