Laserfiche WebLink
analysis of a sample in a given matrix type containing <br /> the analyte. <br /> For operational purposes, when it is necessary to <br /> determine the MDL in the matrix, the MDL should be <br /> determined by multiplying the appropriate one-sided 99% t- <br /> statistic by the standard deviation obtained from a <br /> minimum of three analyses of a matrix spike containing the <br /> analyte of interest at a concentration three to five times <br /> the estimated MDL, where the t-statistic is obtained from <br /> standard references or the table below. <br /> No. of samples: t-statistic <br /> 3 6.96 <br /> 4 4.54 <br /> 5 3.75 <br /> 6 3.36 <br /> 7 3.14 <br /> 8 3.00 <br /> 9 2.90 <br /> 10 2.82 <br /> Estimate the MDL as follows: <br /> Obtain the concentration value that corresponds to: <br /> a) an instrument signal/noise ratio within the range of <br /> 2.5 to 5.0, or <br /> b) the region of the standard curve where there is a <br /> significant change in sensitivity (i .e. , a break in the <br /> slope of the standard curve) . <br /> Determine the variance (S2) for each analyte as follows: <br /> n I <br /> S2= 1 E (X1 i—X) <br /> n-1 i= <br /> where x; = the ith measurement of the variable x <br /> and x = the average value of x; <br /> ONE - 26 Revision 1 <br /> July 1992 <br />