Laserfiche WebLink
D. Magna Western Fuels 007 10-19-04 1240 <br />Program run on 10-26-2004 at 13:48:07 <br />~ULTS CALCULATED USING THE BINOMINAL METHOD <br />CONC. NUMBER NUMBER PERCENT BINOMIAL <br />EXPOSED DEAD DEAD PROB. (PERCENT) <br />-- <br />- <br />---- <br />------ <br />----------------------------------------------------- <br />100 20 1 5.00 0.0020 -------------- <br />75 20 0 0.00 LESS THAN 0.001 <br />50 20 0 0.00 LESS THAN 0.001 <br />25 20 0 0.00 LESS THAN 0.001 <br />12.5 20 0 0.00 LESS THAN 0.001 <br />AT A CONFIDENCE LEVEL OF 95 PERCENT, THE BINOMIAL TEST <br />SHOWS THAT THE LC50 IS ABOVE 100 <br />THE USEFULNESS OF ANY LC50 CALCULATED FROM THIS SET OF DATA <br />IS QUESTIONABLE BECAUSE A CONCENTRATION-EFFECT RELATIONSHIP <br />HAS NOT BEEN DEMONSTRATED OVER A REASONABLE RANGE (e.g. <br /><37 TO>63 OF PERCENT DEAD <br />.NEITHER THE APPROXIMATE LC50 CALCULATION NOR THE MOVING <br />AVERAGE METHOD CAN BE USED WITH THIS SET OF DATA. <br />EITHER THE HIGHEST CONCENTRATION KILLED LESS THAN 50 <br />PERCENT OR THE LOWEST KILLED MORE THAN 50 PERCENT. <br />IF THE PROBIT SLOPE IS NEGATIVE, ENTER THE DATA AGAIN <br />USING THE NUMBER ALIVE INSTEAD OF THE NUMBER DEAD. <br /> <br />W._~N THERE ARE LESS THAN TWO DIFFERENT CONCENTRATIONS AT <br />WHICH THE PERCENT DEAD IS BETWEEN 0 AND 100, THE PROBIT <br />'METHOD CANNOT GIVE ANY STATISTICALLY SOUND RESULTS. <br />COMPARE RESULTS WITH ORIGINAL DATA TO SEE IF THEY ARE <br />REASONABLE. <br /> <br />