EXAMPLES 



The first example shows the output from sampling the same population described by 

 Hartley (1966). 



ID 



10 

 8 

 6 

 A 

 2 

 9 

 7 

 5 

 3 

 I 



EST. SIZE 



NO. OF UMTS TO BE 

 MEASURED = 3 



130. 

 109. 

 105. 



93. 



32. 

 115. 

 109. 



95. 



93. 



30. 



OCCCO 

 OOCCO 

 00000 

 00000 



ocooo 



00000 

 00000 

 00000 

 OCCCO 

 00000 



SAMPLES DRAWN 



( NO. IN STRATUM = 10) 



( NO. OF SAMPLES = 3) 



( TX = 0. lOUOOOE OA) 



( TS = 0. 3359580E-03 ) 



( TP = 0.8516406E 00) 



ID PROBABILITY 



3 0.27596A4E 00 

 6 0.3115721E 00 

 9 0.3A12457E 00 



From the population of 10 units, three are to be drawn for measurement. The third, 

 sixth, and ninth units comprise the sample. Table 1 shows the calculation of the 

 estimates of the population total and its variance based on the measurements (J^) of 

 these three sample units. The estimate of the pop ulation .total is the sum of Y-i/Pi = 

 578. The sample standard error of this value is /642 .61 -' or ± 25.3. 



^ This value differs from Hartley's calculated variance of 653.2 only because the 

 exact value of (1 - Z s^^) has replaced his (^N-l)/N approximation. 



