Laserfiche WebLink
92 HANDBOOK OF ENVIRONMENTAL ANALYSIS I <br /> Table 1-31. One-tailed t-statistic at 99% confidence level for a variety of <br /> repetitions. <br /> Repetitions t0.99 <br /> Repetitions to.99 <br /> 3 <br /> 6.965 15 2.624 <br /> 4 <br /> 4.541 16 2.602 <br /> 3.747 17 2.583 <br /> 5 <br /> g 3.365 18 2.567i <br /> 7 <br /> 3.143 19 2.552 <br /> 8 <br /> 2.998 20 2.539 <br /> 2.896 21 2.528 <br /> 9 <br /> 1 22 2.518 <br /> 0 2.821 <br /> 11 <br /> 2.764 23 2.508 <br /> 12 <br /> 2 718 24 2.500 <br /> 13 <br /> 2.681 25 2.492 <br /> 14 <br /> 2.650 26 2.485 <br /> In practice,determination of MDLs is quite challenging• If the spike level is too <br /> high,the resulting standard deviation will be so small,that the calculated MDL can <br /> actually be below the instrument detection ighethal. the calculated MDL n the other hand,if tiseuseless for spike is <br /> too low,the standard deviation can be so trig <br /> most regulatory purposes. Due to the considerable time that can be spent on these <br /> procedures,most analysts seek only to verify the MDLs listed in most of the published <br /> regulatory methods. This practice in itself causes problems for end users who have <br /> exacting data requirements. <br /> The determination of the MDL for total suspended solids is illustrated in Table 1-32. <br /> In the first trial,the spike level was 12 mg/L while for the second trial the spike was <br /> increased to 16 mg/L. The EPA expected MDL fowhile the second trial t is 5 mg/L. Trial I the EPA MDL, <br /> es an <br /> MDL that is not usable for the most part, <br /> yet there is only a slight difference in the spike level. <br /> Table 1-32. MDL example for Total Suspended Solids in mg/L <br /> Trial 1 Trial <br /> 12.5 14.5 <br /> 17 16 <br /> 15.5 18.5 <br /> 13 17 <br /> 8 15.5 <br /> 16 16 <br /> 13.5 16 <br /> 17 15 <br /> 16 16 <br /> LiMlean 14.28 16.06 <br /> 2.89 1.16 <br /> 8.373.36 <br /> Genium Publishing Corporation <br />