References - SPEC114220010025

DOE-SPEC-1142-2001

B.10.2. Select the maximum Ln(SI) for each BeLPT in the reference data set (see column

8 of Table B-V).

B.10.3. Calculate M the median of the maximum Ln (SI) values. In the example data set

above M = 0.081.

B.10.4. Calculate the absolute deviates for each BeLPT by taking the absolute value of

difference between the maximum Ln(SI) and the median maximum Ln(SI)].

B.10.5. Calculate the median absolute deviate (MAD). In the example above

MAD = 0.229

B.10.6. Calculate the outlier resistant (i.e., MAD) estimate of the standard deviation (SD)

for the Ln(maximum SI) values using the formula

SD = 1.48 x MAD x √(n/n-1) where n = the number of BeLPTs in the reference

data set.

In the example above n = 33 and MAD = 0.229 so

SD = 1.48 x 0.229 x √(33/32) = 0.34

B.10.7. The values of M and SD for each lot of serum are used to calculate the

standardized maximum Ln(SI) for each patient tested in that serum. The

standardized maximum Ln(SI) is obtained by subtracting the value of the median

of reference data set from the maximum Ln(SI) and then dividing by the standard

deviation from the reference data set (see section 8.3). Using the SIs from Table

B-IV and the reference data set in Table B-V:

maximum SI = 2.65, maximum Ln(SI) = 0.98, M = 0.081, and SD = 0.34;

standardized maximum Ln(SI) = (0.98 - 0.081)/0.34 = 2.64;

this is below the cut point of 3.1 (see section 9.3).

B.11

References

B.11.1 Draper, N.R. and Smith, H., (1998), Applied Regression Analysis. Third Edition,

NY, J. Wiley & Sons.

B.11.2 Frome, E., Cragle, D., Littlefield, L.G., and Colyer, S. (2001), Results of the

Analysis of the Blood Lymphocyte Proliferation Test Data From the Oak Ridge

Y-12 Study, DRAFT ORNL Report (under revision).